View LAN9221-ABZJ_9065029.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

smsc lan9221/lan9221i datasheet revision 2.9 (03-01-12) datasheet product features lan9221/lan9221i high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o highlights ? optimized for high performance applications ? efficient architecture with low cpu overhead ? easily interfaces to most 16-bit embedded cpus ? 1.8v to 3.3v variable voltage i/o accommodates wide range of i/o signalling without voltage level shifters ? integrated phy with hp auto-mdix support ? integrated checksum offl oad engine helps reduce cpu load ? low pin count and small body size package for small form factor system designs target applications ? cable, satellite, and ip set-top boxes ? digital video recorders and dvd recorder/players ? digital tv ? digital media clients/servers and home gateways ? video-over ip solutions, ip pbx & video phones ? wireless routers & access points ? high-end audio di stribution systems key benefits ? non-pci ethernet controller for high performance applications 16-bit interface with fast bus cycle times burst-mode read support ? minimizes dropped packets internal buffer memory can store over 200 packets automatic pause and back-pressure flow control ? minimizes cpu overhead supports slave-dma interrupt pin with programmable hold-off timer ? reduces system cost and in creases design flexibility ? sram-like interface easily interfaces to most embedded cpus or socs ? reduced power modes numerous power management modes wake on lan magic packet wakeup wakeup indicator event signal link status change ? single chip ethernet controller fully compliant with ieee 802.3/802.3u standards integrated ethernet mac and phy 10base-t and 100base-tx support full- and half-duplex support full-duplex flow control backpressure for half-duplex flow control preamble generation and removal automatic 32-bit crc generation and checking automatic payload padding and pad removal loop-back modes ? flexible address filtering modes one 48-bit perfect address 64 hash-filtered multicast addresses pass all multicast promiscuous mode inverse filtering pass all incoming with status report disable reception of broadcast packets ? integrated 10/100 ethernet phy supports hp auto-mdix auto-negotiation supports energy-detect power down ? host bus interface simple, sram-like interface 16-bit data bus 16kbyte fifo with flexible tx/rx allocation one configurable host interrupt ? miscellaneous features small form factor, 56-pin qfn lead-free rohs compliant package integrated 1.8v regulator integrated checksum offload engine mixed endian support general purpose timer optional eeprom interface support for 3 status leds multiplexed with programmable gpio signals ? single 3.3v power supply with variable voltage i/o ? commercial and industrial temperature support downloaded from: http:///
order number(s): LAN9221-ABZJ for 56-pin, qfn lead-free rohs compliant package (0 to +70 c temp range) lan9221i-abzj for 56-pin, qfn lead-free rohs compliant package (-40 to +85 c temp range) this product meets the halogen maximum concentration values per iec61249-2-21 for rohs compliance and environmen tal information, please visit www.smsc.com/rohs high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 2 smsc lan9221/lan9221i datasheet copyright ? 2012 smsc or its subsidiaries. all rights reserved. circuit diagrams and other information relating to smsc produc ts are included as a means of illustrating typical applications. consequently, complete information sufficient for construction purposes is not necessarily given. although the information has been checked and is believed to b e accurate, no re sponsibility is assumed for inaccuracies. smsc reserves the right to make changes to specifications and produc t descriptions at any time without notice. contact your local sm sc sales office to obtain the latest specifications before placing your product order. the provision of this inform ation does not convey to the purchaser of the described semicond uctor devices any licenses under any patent rights or other intellectual property rights of smsc or others. all sales are expr essly conditional on your agreement to the te rms and conditions of the most recently dated version of smsc's standard terms of sale agreement dated before the date of your order (the "terms of sale agreement"). the pro duct may contain design defects or errors known as anomalies which may cause the product's functions to deviate from published specifications . anomaly sheets are availab le upon request. smsc products are not designed, intended, authorized or warranted for use in any life support or other applicati on where product failure could cause or contribute to personal injury or severe property damage. any and all such uses without prior written approval of an officer of smsc and furthe r testing and/or modification will be fully at the risk of the customer. copies of this document or other smsc literature, as well as the terms of sale agreement, may be obtai ned by visiting smscs website at h ttp://www.smsc.com. smsc is a registered trademark of standard microsystems corporation (smsc). product names and company names are the trademarks of their respective holders. smsc disclaims and excludes any and all warrant ies, including without limi tation any and all implied warranties of merchantabil ity, fitness for a particular purpose, title, a nd against infringement and the like, and any and all warranties arising from any cou rse of dealing or usage of trade. in no event shall smsc be liable for any direct, incidental, indi rect, special, punitive, or cons equential damages; or for lost data, profits, savings or revenues of any kind; regardless of the form of action, whether based on contrac t; tort; negligence of smsc or others; strict liability; breach of wa rranty; or otherwise; whether or not any remedy of buyer is h eld to have failed of its essential purpose, and whether or not smsc has been advised of the possibility of such damages. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 3 revision 2.9 (03-01-12) datasheet table of contents chapter 1 general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.1 block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.2 internal block overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.3 10/100 ethernet phy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.4 10/100 ethernet mac . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.5 receive and transmit fifos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.6 interrupt controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.7 gpio interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.8 serial eeprom interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.9 power management controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.10 general purpose timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.11 host bus interface (sram interface) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 chapter 2 pin description and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.1 pin list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.2 external pull-up/pull-do wn resistors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.3 buffer types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 chapter 3 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.1 10/100 ethernet mac . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.2 flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.2.1 full-duplex flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.2.2 half-duplex flow control (backpressure) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.2.3 virtual local area network (vlan) support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.3 address filtering functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 3.4 filtering modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.4.1 perfect filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.4.2 hash only filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5 wake-up frame detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 3.5.1 magic packet detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 3.6 checksum offload engines (coe) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.6.1 receive checksum offload engine (rxcoe). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.6.2 transmit checksum offload engine (txcoe) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.7 host bus operations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.7.1 bus writes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.7.2 bus reads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.7.3 mixed endian support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.7.4 word swap function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.8 general purpose timer (gp timer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.9 eeprom interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3.9.1 mac address auto-load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3.9.2 eeprom host operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3.10 power management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.10.1 system description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.10.2 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.10.3 internal phy power-down modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 3.11 detailed reset description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3.11.1 hardware reset input (nr eset) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 3.11.2 resume reset timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 3.11.3 soft reset (srst) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 3.11.4 phy reset timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 4 smsc lan9221/lan9221i datasheet 3.12 tx data path operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 3.12.1 tx buffer format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 3.12.2 tx command format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1 3.12.3 tx data format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3.12.4 tx status format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 3.12.5 calculating actual tx data fifo usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.12.6 transmit examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 3.12.7 transmitter errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.12.8 stopping and starting the transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.13 rx data path operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.13.1 rx slave pio operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.13.2 rx packet format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 3.13.3 rx status format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 3.13.4 stopping and starting the receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 3.13.5 receiver errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 chapter 4 internal ethernet phy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.1 top level functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.2 100base-tx transmit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.2.1 4b/5b encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.2.2 scrambling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.2.3 nrzi and mlt3 encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.2.4 100m transmit driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.2.5 100m phase lock loop (pll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.3 100base-tx receive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 4.3.1 100m receive input. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 4.3.2 equalizer, baseline wander correction and clock and data recovery . . . . . . . . . . . . . 70 4.3.3 nrzi and mlt-3 decoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 4.3.4 descrambling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.3.5 alignment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.3.6 5b/4b decoding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.4 10base-t transmit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.4.1 10m transmit data across the internal mii bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.4.2 manchester encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.4.3 10m transmit drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.5 10base-t receive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.5.1 10m receive input and squelch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.5.2 manchester decoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.5.3 jabber detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.6 auto-negotiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.7 parallel detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 4.7.1 re-starting auto-negotiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 4 4.7.2 disabling auto-negotiation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 4.7.3 half vs. full-duplex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 4.8 hp auto-mdix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 chapter 5 register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 5.1 register nomenclature and access attribut es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 8 5.2 rx and tx fifo ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 5.2.1 rx fifo ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 5.2.2 tx fifo ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 5.3 system control and status registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 5.3.1 id_revchip id and revision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 5.3.2 irq_cfginterrupt configuration register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 5 revision 2.9 (03-01-12) datasheet 5.3.3 int_stsinterrupt status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 5.3.4 int_eninterrupt enable register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 5.3.5 byte_testbyte order test register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 5.3.6 fifo_intfifo level interr upts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 5.3.7 rx_cfgreceive configuration register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 5.3.8 tx_cfgtransmit configur ation register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 5.3.9 hw_cfghardware configuration register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 5.3.10 rx_dp_ctrlreceive datapath co ntrol register . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 5.3.11 rx_fifo_infreceive fifo inform ation register. . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 5.3.12 tx_fifo_inftransmit fifo info rmation register . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 5.3.13 pmt_ctrl power management control register . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 5.3.14 gpio_cfggeneral purpose io c onfiguration register . . . . . . . . . . . . . . . . . . . . . . . 95 5.3.15 gpt_cfg-general purpose timer configuration regi ster . . . . . . . . . . . . . . . . . . . . . . 96 5.3.16 gpt_cnt-general purpose timer current count re gister . . . . . . . . . . . . . . . . . . . . . . 97 5.3.17 word_swapword swap control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 5.3.18 free_runfree-run 25mhz counter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 5.3.19 rx_dropC receiver dropped frames counter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 5.3.20 mac_csr_cmd C mac csr synchronizer command register . . . . . . . . . . . . . . . . . 99 5.3.21 mac_csr_data C mac csr synchronizer data re gister . . . . . . . . . . . . . . . . . . . . . 99 5.3.22 afc_cfg C automatic flow control configuration register . . . . . . . . . . . . . . . . . . . . 100 5.3.23 e2p_cmd C eeprom command register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 5.3.24 e2p_data C eeprom data register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 5.4 mac control and status registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 5.4.1 mac_crmac control register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 5.4.2 addrhmac address high register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 5.4.3 addrlmac address low register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 5.4.4 hashhmulticast hash table high register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 5.4.5 hashlmulticast hash table low register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 5.4.6 mii_accmii access register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 5.4.7 mii_datamii data register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 5.4.8 flowflow control register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 5.4.9 vlan1vlan1 tag register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 5.4.10 vlan2vlan2 tag register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 5.4.11 wuffwake-up frame filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 5.4.12 wucsrwake-up control and status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 5.4.13 coe_crchecksum offload engine control register . . . . . . . . . . . . . . . . . . . . . . . . 116 5.5 phy registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 5.5.1 basic control register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 5.5.2 basic status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 5.5.3 phy identifier 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 5.5.4 phy identifier 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 5.5.5 auto-negotiation advertisement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 5.5.6 auto-negotiation link part ner ability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 5.5.7 auto-negotiation expansion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 5.5.8 mode control/status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 5.5.9 special modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 5.5.10 special control/status indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 5.5.11 interrupt source flag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 5.5.12 interrupt mask . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 5.5.13 phy special control/status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 chapter 6 timing diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 6.1 equivalent test load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 6.2 host interface timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 6 smsc lan9221/lan9221i datasheet 6.2.1 special restrictions on back-to-back write/read cycles . . . . . . . . . . . . . . . . . . . . . . . 128 6.2.2 special restrictions on back-to-back read cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 6.3 pio reads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 6.4 pio burst reads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 6.5 rx data fifo direct pio reads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 6.6 rx data fifo direct pio burst reads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 6.7 pio writes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 6.8 tx data fifo direct pio writes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 6.9 power sequence timing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 6.10 reset timing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 6.11 eeprom timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 chapter 7 operational characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 7.1 absolute maximum ratings*. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 7.2 operating conditions** . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 7.3 power consumption (device only). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 7.4 power consumption (device and system components) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 7.5 worst case current co nsumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 7.6 dc electrical specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 7.7 clock circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 chapter 8 package outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 8.1 56-qfn package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 chapter 9 datasheet revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 7 revision 2.9 (03-01-12) datasheet list of figures figure 1.1 system block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 figure 1.2 internal block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 figure 2.1 56-qfn pin configuration (top view) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 figure 3.1 vlan frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 figure 3.2 rxcoe checksum calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 figure 3.3 type ii ethernet frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 figure 3.4 ethernet frame with vlan tag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 figure 3.5 ethernet frame with length field and snap header . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 figure 3.6 ethernet frame with vlan tag and snap header. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 figure 3.7 ethernet frame with multiple vlan tags and snap header . . . . . . . . . . . . . . . . . . . . . . . . 31 figure 3.8 lan9221/lan9221i host data path diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 figure 3.9 fifo access byte ordering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 figure 3.10 eeprom access flow diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 figure 3.11 eeprom erase cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 figure 3.12 eeprom eral cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 figure 3.13 eeprom ewds cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 figure 3.14 eeprom ewen cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 figure 3.15 eeprom read cycle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 figure 3.16 eeprom write cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 figure 3.17 eeprom wral cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 figure 3.18 pme and pme_int signal generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 6 figure 3.19 simplified host tx flow diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 figure 3.20 tx buffer format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 figure 3.21 tx example 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 figure 3.22 tx example 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 figure 3.23 tx example 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 figure 3.24 host receive routine using interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 figure 3.25 host receive routine with polling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 figure 3.26 rx packet format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 figure 3.27 rx packet format with rx checks um . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 5 figure 4.1 100base-tx data path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 figure 4.2 receive data path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 figure 4.3 direct cable connection vs. cross-over cable conn ection . . . . . . . . . . . . . . . . . . . . . . . . . 76 figure 5.1 memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 figure 5.2 example addrl, addrh and eeprom setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 figure 6.1 equivalent test load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 figure 6.2 pio read cycle timing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 figure 6.3 pio burst read cycle timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 figure 6.4 rx data fifo direct pio read cycle timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 figure 6.5 rx data fifo direct pio burst read cycle timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 figure 6.6 pio write cycle timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 figure 6.7 tx data fifo direct pio write timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 35 figure 6.8 power sequence timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 figure 6.9 reset timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 figure 6.10 eeprom timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 figure 8.1 56 pin qfn package definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 figure 8.2 56 pin qfn recommended pcb land pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 8 smsc lan9221/lan9221i datasheet list of tables table 2.1 host bus interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 table 2.2 lan interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 table 2.3 serial eeprom interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 table 2.4 system and power signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 table 2.5 56-qfn package pin assignments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 table 2.6 external pull-up/pull-down resistor values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 table 2.7 buffer types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 table 3.1 address filtering modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 table 3.2 wake-up frame filter register stru cture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 table 3.3 filter i byte mask bit definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 table 3.4 filter i command bit definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 table 3.5 filter i offset bit definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 table 3.6 filter i crc-16 bit definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 table 3.7 tx checksum preamble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 table 3.8 endian ordering logic operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 table 3.9 required eeclk cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 table 3.10 power management states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 table 3.11 reset sources and affected circuitry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 table 3.12 tx command 'a' format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 table 3.13 tx command 'b' format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 table 3.14 tx data start offset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 table 4.1 4b/5b code table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 table 4.2 crs behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 table 5.1 direct address register map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 table 5.2 rx alignment bit definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 table 5.3 valid tx/rx fifo allocations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 table 5.4 eeprom enable bit definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 table 5.5 backpressure duration bit mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 table 5.6 mac csr register map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 table 5.7 addrl, addrh and eeprom byte ordering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 table 5.8 lan9221/lan9221i phy control and status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 table 5.9 mode control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 table 6.1 read after write timing rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 table 6.2 read after read timing rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 table 6.3 pio read timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 table 6.4 pio burst read timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 table 6.5 rx data fifo direct pio read timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2 table 6.6 rx data fifo direct pio burst read cycle timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 table 6.7 pio write cycle timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 table 6.8 tx data fifo direct pio write timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 table 6.9 power sequence timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 table 6.10 reset timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 table 6.11 eeprom timing values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 table 7.1 power consumption device only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 table 7.2 power consumption device and system components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 table 7.3 maximum supply current char acteristics vddvario=3.3v + 300mv. . . . . . . . . . . . . . . . . 142 table 7.4 maximum supply current characteristics vddvario=2.5v + 10%. . . . . . . . . . . . . . . . . . . 142 table 7.5 maximum supply current characteristics vddvario=1.8v + 10%. . . . . . . . . . . . . . . . . . . 142 table 7.6 i/o buffer characteristics vddvario=3.3v +/- 300mv . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 table 7.7 i/o buffer characteristics vddvario=2.5v +/- 10% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 table 7.8 i/o buffer characteristics vddvario=1.8v +/- 10% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 table 7.9 100base-tx transceiver ch aracteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 9 revision 2.9 (03-01-12) datasheet table 7.10 10base-t transceiver char acteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 table 7.11 lan9221/lan9221i crystal specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 7 table 8.1 56 pin qfn package parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 table 9.1 customer revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 10 smsc lan9221/lan9221i datasheet chapter 1 general description the lan9221/lan9221i is a full-featured, singl e-chip 10/100 ethernet controller designed for embedded applications where performance, flexibility, ease of integration and system cost control are required. the lan9221/lan9221i has been specific ally designed to provide high performance and throughput for 16-bit applications . the lan9221/lan9221i is fu lly ieee 802.3 10base-t and 802.3u 100base-tx compliant, and suppor ts hp auto-mdix. the variable voltage i/o signals of the lan9221/lan9221i accommo date lower voltage i/o si gnalling without the ne ed for voltage level shifters. the lan9221/lan9221i includes an integrated et hernet mac and phy with a high-performance sram-like slave interface. the simple, yet highly functional host bus interface provides a glue-less connection to most common 16-bit microprocesso rs and microcontrollers as well as 32-bit microprocessors with a 16-bit external bus. t he integrated checksum offload engines enable the automatic generation of t he 16-bit checksum for received and tr ansmitted ethernet frames, offloading the task from the cpu. the lan9221/lan9221i also includes large transmit and receive data fifos to accommodate high latency applications. in addition, the lan9221/lan9221i memory buffer architecture allows highly efficient use of me mory resources by optimizing packet granularity. applications the lan9221/lan9221i is well suited for many hi gh-performance embedded applications, including: ? cable, satellite and ip set-top boxes ? high-end audio distribution systems ? digital video recorders ? dvd recorders/players ? digital tv ? digital media clients/servers ? home gateways ? industrial and embedded systems wit h extended temperature support the lan9221/lan9221i also supports features which reduce or eliminate packet loss. its internal 16- kbyte sram can hold over 200 received packe ts. if the receive fifo gets too full, the lan9221/lan9221i can automatically generate flow c ontrol packets to the remote node, or assert back-pressure on the remote node by generating network collisions. the lan9221/lan9221i supports numerous power management and wakeup features. the lan9221/lan9221i can be placed in a reduced power mode and can be programmed to issue an external wake signal via several methods, includi ng magic packet, wake on lan and link status change. this signal is ideal for triggering system power-up using remote ethernet wakeup events. the device can be removed from the low po wer state via a host processor command. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 11 revision 2.9 (03-01-12) datasheet 1.1 block diagram the smsc lan9221/lan9221i integrated 10/100 ma c/phy controller is a peripheral chip that performs the function of translating parallel data from a host controller into ethernet packets. the lan9221/lan9221i ethernet mac/phy controller is designed and optimized to function in an embedded environment. all communication is perfo rmed with programmed i/o transactions using the simple sram-like host interface bus. the diagram shown above, describes a typical syst em configuration of th e lan9221/lan9221i in a typical embedded environment. the lan9221/lan9221i is a general purpose, pl atform independent, ethernet controller. the lan9221/lan9221i consists of four major functional blocks. the four blocks are: ? 10/100 ethernet phy ? 10/100 ethernet mac ? rx/tx fifos ? host bus interface (hbi) figure 1.1 system block diagram microprocessor/ microcontroller lan9221/ lan9221i magnetics ethernet system peripherals system memory eeprom (optional) leds/ gpio 25mhz xtal system bus downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 12 smsc lan9221/lan9221i datasheet 1.2 internal block overview this section provides an overview of each of these functional blocks as shown in figure 1.2, "internal block diagram" . figure 1.2 internal block diagram 1.3 10/100 ethernet phy the lan9221/lan9221i integrates an ieee 802.3 physica l layer for twisted pair ethernet applications. the phy can be configured for either 100 mbps (100base-tx) or 10 mbps (10base-t) ethernet operation in either full or half duplex configurat ions. the phy block supports hp auto-mdix and auto- negotiation. minimal external components are required fo r the utilization of the integrated phy. 1.4 10/100 ethernet mac the transmit and receive data paths are separate within the mac allowing the highest performance especially in full duplex mode. the data paths connect to the pio interface function via separate busses to increase performance. payload data as we ll as transmit and receive status is passed on these busses. a third internal bus is used to access the macs c ontrol and status register s (csrs). this bus is accessible from the host through the pio interface function. on the backend, the mac interfac es with the internal 10/100 phy through a mii (media independent interface) port internal to the lan9221/lan9221i . the mac csr's also provide a mechanism for accessing the phys internal registers through th e internal smi (serial management interface) bus. the mac interface layer (mil), within the mac, cont ains a 2k byte transmit and a 128 byte receive fifo which is separate from the tx and rx fi fos. the fifos within the mac are not directly accessible from the host interface. the different iation between the tx/rx fifo memory buffers and the mac buffers is that when the transmit or receive packets are in the mac buffers, the host no longer 10/100 ethernet phy 10/100 ethernet mac 2kb to 14kb configurable tx fifo 2kb to 14kb configurable rx fifo interrupt controller gp timer pio controller 3.3v to 1.8v core regulator pll 25mhz +3.3v eeprom controller eeprom (optional) rx status fifo tx status fifo mil - tx elastic buffer - 2k bytes mil - rx elastic buffer - 128 bytes power management irq fifo_sel pme wakup indicator host bus interface (hbi) 16-bit sram i/f rx checksum offload engine tx checksum offload engine lan downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 13 revision 2.9 (03-01-12) datasheet can control or access the tx or rx data. the mac buffers (both tx and rx) are in effect the working buffers of the ethernet mac logic. in the case of reception, the data must be moved first to the rx fifos for the host to access the data. for tx operat ions, the mil operates in store-and-forward mode and will queue an entire frame before beginning transmission. 1.5 receive and transmit fifos the receive and transmit fifos allow increased pa cket buffer storage to the mac. the fifos are a conduit between the host interface and the mac th rough which all transmitted and received data and status information is passed. deep fifos allow a high degree of latency tolerance relative to the various transport and os software stacks thus reducing or minimizing overrun conditions. like the mac, the fifos have separate receive and transmit data paths. in addition, the rx and tx fifos are configurable in size, allowing increased flexibility. 1.6 interrupt controller the lan9221/lan9221i supports a single programm able interrupt. the programmable nature of this interrupt allows the user the ability to optimize performance dependent upon the application requirement. both the polarity and buffer type of th e interrupt pin are configurable for the external interrupt processing. the interrupt line can be conf igured as an open-drain output to facilitate the sharing of interrupts with other devices. in addition, a programmable interrupt de-assertion interval is provided. 1.7 gpio interface a 3-bit gpio and 2-bit gpo (multiplexed on the eepr om and led pins) interf ace is included in the lan9221/lan9221i. it is accessible through the ho st bus interface via the csrs. the gpio signals can function as inputs, push- pull outputs and open drain outputs. the gpios (gpos are not configurable) can also be configured to tr igger interrupts with programmable polarity. 1.8 serial eeprom interface a serial eeprom interface is included in the lan9221/lan9221i. the serial eeprom is optional and can be programmed with the lan9221/lan9221i mac address. the lan9221/lan9221i can optionally load the mac address automatically after hardware reset, or soft reset. 1.9 power management controls the lan9221/lan9221i supports comprehensive arra y of power management modes to allow use in power sensitive applications. wake on lan, link status change and magic packet detection are supported by the lan9221/lan9221i. an external pme (power management event) interrupt is provided to indicate detec tion of a wakeup event. 1.10 general purpose timer the general-purpose timer has no dedicated f unction within the lan9221/lan9221i and may be programmed to issu e a timed interrupt. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 14 smsc lan9221/lan9221i datasheet 1.11 host bus interf ace (sram interface) the host bus interface provides a fifo interface fo r the transmit and receive data paths, as well as an interface for the lan9221/lan9221i c ontrol and status registers (csrs). the host bus interface is the primary bus for connection to the embedded host system. this interface models an asynchronous sram. tx fifo, rx fifo, and csrs are accessed through this interface. programmed i/o transactions are supported. the lan9221/lan9221i host bus interface supports 16-bit bus transfers. internally, all data paths are 32-bits wide. the lan9221/lan9221i can be inte rfaced to either big-endian or little-endian processors and includes mixed endian support for fifo accesses. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 15 revision 2.9 (03-01-12) datasheet chapter 2 pin description and configuration figure 2.1 56-qfn pin configuration (top view) vss note: exposed pad (vss) on bottom of package must be connected to ground **denotes a multifunction pin note: when hp auto-mdix is activated, the tpo+/- pins can function as tpi+/- and vice-versa smsc lan9221/lan9221i 56-qfn (top view) irq nreset d7 test 12 3 4 5 6 7 8 9 1011 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 4241 40 39 38 37 36 35 34 33 32 31 30 29 56 55 54 53 52 51 50 49 48 47 46 45 44 43 fifo_sel a1 a2 a3 a4 a5 a6 a7 gpio2/nled3** gpio1/nled2** gpio0/nled1** vdd18core vdd33reg pme eeclk/gpo4** eecs eedio/gpo3** vdd18core d0 d1 d2 d3 d4 d5 vddvario d6 tpo- tpo+ vdd33a tpi- tpi+ vdd33a exres vdd33a amdix_en vdd18a xtal2 xtal1/clkin** vddvario d8d9 d10 vddvario d11 d12 d13 d14 d15 vddvario ncs nwr nrd downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 16 smsc lan9221/lan9221i datasheet 2.1 pin list note: the pin names for the twisted pair pins shown ab ove apply to a normal connection. if hp auto- mdix is enabled and a reverse connection is detected, or a revers e connection is manually selected, the input pins become outputs, and vice-versa, as indicated in the descriptions. table 2.1 host bus interface signals name symbol buffer type # pins description host data d[15:0] vis/vo8 16 bi-directional data port. host address a[7:1] vis 7 7-bit address po rt. used to select internal csrs and tx and rx fifos. read strobe nrd vis 1 active low st robe to indicate a read cycle. write strobe nwr vis 1 active low strobe to indicate a write cycle. this signal, qualified with ncs, is also used to wakeup the lan9221/lan9221i when it is in a reduced power state. chip select ncs vis 1 active low signal used to qualify read and write operations. this signal qualified with nwr is also used to wakeup the lan9221/lan9221i when it is in a reduced power state. interrupt request irq vo8/ vod8 1 programmable interrupt request. programmable polarity, source and buffer types. fifo select fifo_sel vis 1 when driven high all accesses to the lan9221/lan9221i are to the rx or tx data fifos. in this mode, the a[7:3] upper address inputs are ignored. table 2.2 lan interface signals name symbol buffer type num pins description tpo+ tpo+ ao 1 transmit positive output (normal) receive positive input (reversed) tpo- tpo- ao 1 transmit negative output (normal) receive negative input (reversed) tpi+ tpi+ ai 1 receive positive input (normal) transmit positive input (reversed) tpi- tpi- ai 1 receive negative input (normal) transmit negative output (reversed) phy external bias resistor exres ai 1 must be connected to ground through a 12.4k ohm 1% resistor. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 17 revision 2.9 (03-01-12) datasheet table 2.3 serial eeprom interface signals name symbol buffer type num pins description eeprom data, gpo3, tx_en, tx_clk eedio/gpo3/ tx_en/tx_clk vis/vo8 1 eeprom data: this bi-directional pin can be connected to a serial eeprom dio. this is optional. general purpose output 3: this pin can also function as a general purpose output, or it can be configured to monitor the tx_en or tx_clk signals on the internal mii port. when configured as a gpo signal, or as a tx_en/tx_clk monitor, the eecs pin is deasserted so as to never unintentionally access the serial eeprom. this signal cannot function as a general-purpose input. eeprom chip select eecs vo8 1 serial eeprom chip select. eeprom clock, gpo4 rx_dv, rx_clk eeclk/gpo4/ rx_dv/rx_clk vo8 (pu) 1 eeprom clock: serial eeprom clock pin. general purpose output 4: this pin can also function as a general-purpose output, or it can be configured to monitor the rx_dv or rx_clk signals on the internal mii port. when configured as a gpo signal, or as an rx_dv/rx_clk monitor, the eecs pin is deasserted so as to never unintentionally access the serial eeprom. this signal cannot function as a general-purpose input. note: when the eeprom interface is not used, the eeclk pin must be left unconnected. note: when operating at reduced vddvario voltage levels (less than 3.0v), this pin must be pulled-up with an external resistor. refer to section 2.2, "external pull-up/pull-down resistors" for more information. note: this pin must not be pulled low by an external resistor or driven low externally under any conditions. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 18 smsc lan9221/lan9221i datasheet table 2.4 system and power signals name symbol buffer type num pins description crystal 1, clock in xtal1/clkin lclk 1 external 25mhz crystal input. this pin can also be connected to single-ended ttl oscillator (clkin). if this method is implemented, xtal2 should be left unconnected. crystal 2 xtal2 oclk 1 exter nal 25mhz crystal output. reset nreset vis (pu) 1 active-low reset input. resets all logic and registers within the lan9221/lan9221i. this signal is pulled high with a weak internal pull-up resistor. note: the lan9221/lan9221i must be reset on power-up via nreset or following power-up via a soft reset (srst). the lan9221/lan9221i must always be read at least once after reset, or upon return from a power-saving state or write operations wi ll not function. see section 3.11, "detailed reset description," on page 47 for additional information note: when operating at reduced vddvario voltage levels (less than 3.0v), this pin must be pulled-high to a valid level with an external resistor or must be driven as an input. refer to section 2.2, "external pull-up/pull- down resistors" for more information. wakeup indicator pme vo8/ vod8 1 when programmed to do so, is asserted when the lan9221/lan9221i detects a wake event and is requesting the system to wake up from the associated sleep state. the polarity and buffer type of this signal is programmable. note: detection of a power management event, and assertion of the pme signal will not wakeup the lan9221/lan9221i. the lan9221/lan9221iwill only wake up when it detects a host write cycle (assertion of ncs and nwr). although any write to the lan9221/lan9221i, regardless of the data written, will wake-up the device when it is in a power-saving mode, it is required that the byte_test register be used for this purpose. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 19 revision 2.9 (03-01-12) datasheet auto-mdix enable amdix_en vis (pu) 1 enables auto-mdix. pull high or leave unconnected to enable auto-mdix, pull low to disable auto-mdix. note: when operating at reduced vddvario voltage levels (less than 3.0v), this pin must be pulled to a valid level with an external resistor. refer to section 2.2, "external pull-up/pull- down resistors" for more information. test test vis (pd) 1 reserved for internal test purposes only. note: when operating at a reduced vddvario voltage (less than 3.0v), this pin must be connected to ground or pulled-low with an external resistor. when vddvario = 3.3v, this pin may be left unconnected. refer to section 2.2, "external pull-up/pull-down resistors" for more information. general purpose i/o data, nled1 (speed indicator), nled2 (link & activity indicator), nled3 (full- duplex indicator ) . gpio[2:0]/ nled[3:1] vis/ vo12/ vod12 3 general purpose i/o data: these three general-purpose signals are fully programmable as either push-pull outpu t, open-drain output or input by writing the gpio_cfg configuration register in the csrs. they are also multiplexed as gp led connections. gpio signals are schmitt-triggered inputs. when configured as led outputs these signals are open-drain. nled1 (speed indicator). this signal is driven low when the operating speed is 100mb. during auto-negotiation, when the cable is disconnected, and during 10mbs operation, this signal is driven high. nled2 (link & activity indicator). this signal is driven low (led on) when the lan9221/lan9221i detects a valid link. this signal is pulsed high (led off) for 80ms whenever transmit or receive activity is detected. this signal is then driven low again for a minimum of 80ms, after which time it will repeat the process if tx or rx activity is detected. effectively, led2 is activated solid for a link. when transmit or receive activity is sensed led2 will flash as an activity indicator. nled3 (full-duplex indicator). this signal is driven low when the link is operating in full- duplex mode. variable voltage i/o power vddvario p 4 variable voltage i/o logic power supply pins. refer to section 7.2, "operating conditions**," on page 139 for additional details. common ground vss p 1 pad common ground +3.3v regulator power supply vdd33reg p 1 +3.3v power supply for internal +1.8v regulator. table 2.4 system and power signals (continued) name symbol buffer type num pins description downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 20 smsc lan9221/lan9221i datasheet +3.3v analog power vdd33a p 3 +3.3v analog power supply pins. +1.8v analog power vdd18a p 1 +1.8v analog power supply pin. this pin must be connected externally to vdd18core. core voltage decoupling vdd18core p 2 +1.8 v from internal core regulator. both pins must be connected toge ther externally. each pin requires a 0.01uf decoupling capacitor. in addition, pin 2 requires a bulk 4.7uf capacitor (<2 ohm esr) in parallel. these pins must not be used to supply power to other external devices. table 2.5 56-qfn package pin assignments pin num pin name pin num pin name pin num pin name pin num pin name 1 vdd33reg 15 nrd 29 d6 43 irq 2 vdd18core 16 nwr 30 vddvario 44 tpo- 3 gpio0/nled1 17 ncs 31 d5 45 tpo+ 4 gpio1/nled2 18 vddvario 32 d4 46 vdd33a 5 gpio2/nled3 19 d15 33 d3 47 tpi- 6 a7 20 d14 34 d2 48 tpi+ 7 a6 21 d13 35 d1 49 vdd33a 8 a5 22 d12 36 d0 50 exres 9 a4 23 d11 37 vdd18core 51 vdd33a 10 a3 24 vddvario 38 eedio/gpo3 52 amdix_en 11 a2 25 d10 39 eecs 53 vdd18a 12 a1 26 d9 40 eeclk/gpo4 54 xtal2 13 fifo_sel 27 d8 41 pme 55 xtal1/clkin 14 test 28 d7 42 nreset 56 vddvario exposed pad must be connected to vss table 2.4 system and power signals (continued) name symbol buffer type num pins description downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 21 revision 2.9 (03-01-12) datasheet 2.2 external pull-up /pull-down resistors as detailed in ta b l e 2 . 6 , when using an external pull-up or pull-down resistor, the required value is dependant on the operating voltage. usage of re commended pull-up/pull-down resistor values is required to ensure proper operation. 2.3 buffer types note 2.1 sink and source capabilities are dependant on the vddvario voltage. refer to section 7.6, "dc electrical specifications" for additional information. note 2.2 when operating at reduced vddvario voltage levels (less than 3.0v), do not rely on internal pull-up and pull-down resistors to determine signal state. refer to section 2.2, "external pull-up/pull-down resistors" and the individual signal descriptions in section 2.1, "pin list" for more information. table 2.6 external pull-u p/pull-down resistor values i/o voltage pull-up/pull-do wn resistor value (ohms) 3.3v +/- 300mv 10k 2.5 +/- 10% 7.5k 1.8v +/- 10% 4.7k table 2.7 buffer types type description vis variable voltage input pin vo12 variable voltage output with 12ma sink and 12ma source ( note 2.1 ) vod12 variable voltage open-drain output with 12ma sink ( note 2.1 ) vo8 variable voltage output 8ma symmetrical drive ( note 2.1 ) vod8 variable voltage open-drain output with 8 ma sink ( note 2.1 ) pu 50ua (typical) internal pull-up ( note 2.2 ) pd 50ua (typical) internal pull-down ( note 2.2 ) ai analog input ao analog output aio analog bi-directional iclk crystal oscillator input pin oclk crystal oscillator output pin downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 22 smsc lan9221/lan9221i datasheet chapter 3 functional description 3.1 10/100 ethernet mac the ethernet media access controll er (mac) incorporates the essential protocol requirements for operating an ethernet/ieee 802.3-compliant node and provides an inte rface between the host subsystem and the internal ethernet phy. the mac can operate in either 100-mbps or 10-mbps mode. the mac operates in both half-duplex and full- duplex modes. when operating in half-duplex mode, the mac complies fully with section 4 of iso/iec 8802-3 (ansi/ieee standard) and ansi/ieee 802.3 standards. when operating in full-duplex mode, the mac complies with ieee 802.3x full-duplex operation standard. the mac provides programmable enhanced features designed to mi nimize host supervision, bus utilization, and pre- or post-message processing. these features incl ude the ability to disable retries after a collision, dynamic fcs (frame check sequence) generation on a frame-by-frame basis, automatic pad field insert ion and deletion to enforce minimum frame size attributes, layer 3 checksum calculation for transmit and receive operations, and automatic retransmission and detection of collision frames. the mac can sustain transmission or reception of minimally-sized ba ck-to-back packets at full line speed with an interpacket gap (ipg) of 9.6 microseconds for 10 mbps and 0.96 microseconds for 100 mbps. the primary attributes of the mac function are: ? transmit and receive message data encapsulation ? framing (frame boundary delimitation, frame synchronization) ? error detection (physical medium transmission errors) ? media access management ? medium allocation (collision detectio n, except in full-duplex operation) ? contention resolution (collision handling, except in full-duplex operation) ? flow control during full-duplex mode ? decoding of control frames (pause command) and disabling the transmitter ? generation of control frames ? interface to the internal phy. ? checksum offload engine for calculation of layer 3 transmit and receive checksum. the transmit and receive data paths are separate within the lan9221/lan9221i from the mac to host interface allowing the highest perf ormance, especially in full duplex mode. payload data as well as transmit and receive status are passed on these busses. a third internal bus is used to access the macs c ontrol and status registers (csrs). this bus is also accessible from the host. on the backend, the mac interfaces with the 10/100 phy through an mii (media independent interface) port which is internal to the lan922 1/lan9221i. the mac csr's also provide a mechanism for accessing the phys inte rnal registers through the internal smi (serial management interface) bus. the receive and transmit fifos allow increased pa cket buffer storage to the mac. the fifos are a conduit between the host interface and the mac th rough which all transmitted and received data and status information is passed. deep fifos allow a high degree of latency tolerance relative to the various transport and os software stacks reducing and minimizing overrun conditions. like the mac, the fifos have separate receive and transmit data paths. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 23 revision 2.9 (03-01-12) datasheet the lan9221/lan9221i can store up to 250 ethernet packets utilizing fifos, totaling 16k bytes, with a packet granularity of 4 bytes. this memory is s hared by the rx and tx blocks and is configurable in terms of allocation. this depth of buffer st orage minimizes or eliminates receive overruns. 3.2 flow control the lan9221/lan9221i ethernet mac supports full-dupl ex flow control using the pause operation and control frame. it also supports half-duplex flow control using back pressure. 3.2.1 full-duplex flow control the pause operation inhibits data transmission of data frames for a specified period of time. a pause operation consists of a frame containing the global ly assigned multicast address (01-80-c2-00-00-01), the pause opcode, and a parameter indicating the quantum of slot time (512 bit times) to inhibit data transmissions. the pause parameter may range from 0 to 65,535 slot times. the ethernet mac logic, on receiving a frame with the re served multicast address and paus e opcode, inhibits data frame transmissions for the length of ti me indicated. if a pause request is received while a transmission is in progress, then the pause will take effect after the transmission is complete. control frames are received and processed by the mac and are passed on. the mac also transmits control frames (pause comm and) via both hardware and software control. the software driver requests the mac to transmit a control frame and gives the value of the pause time to be used in the control frame. the mac functi on constructs a control fr ame with the appropriate values set in all the different fields (as defined in the 802.3x specification) and transmits the frame to the mii interface. the transmission of the control frame is not affected by the current state of the pause timer value that is set because of a recently received control frame. 3.2.2 half-duplex flow control (backpressure) in half-duplex mode, back pressure is used for flow control. whenever the receive buffer/fifo becomes full or crosses a certain threshold level, the mac starts sending a jam signal. the mac transmit logic enters a state at the end of current transmission (if an y), where it waits for the beginning of a received frame. once a new frame starts, the mac starts sending the jam signal, which will result in a collision. after sensing the collision, the remote station will back off its transmission. the mac continues sending the jam to make other stations defer transmission. the mac only generates this collision-based back pressure when it receives a new frame, in order to avoid any late collisions. 3.2.3 virtual local area network (vlan) support virtual local area networks or vlans, as defined within the ieee 802.3 standard, provide network administrators one means of grouping nodes within a larger netwo rk into broadcast domains. to implement a vlan, four extra bytes are added to the basic ethernet packet. as shown in figure 3.1, "vlan frame" , the four bytes are inserted after the sour ce address field and before the type/length field. the first two bytes of the vlan tag identif y the tag, and by convention are set to the value 0x8100. the last two bytes identify the specific vlan associated with the packet; they also provide a priority field. the lan9221/lan9221i supports vlan-tagged pa ckets. the lan9221/lan9221i provides two registers which are used to identify vlan-tagged packets. one register should normally be set to the conventional vlan id of 0x8100. the other register provides a way of identifying vlan frames tagged with a proprietary (not 0x8100) iden tifier. if a packet arrives bearing either of these tags in the two bytes succeeding the source address field, the controller will recognize the packet as a vlan-tagged packet. in this case, the controller increases th e maximum allowed packet size from 1518 to 1522 bytes (normally the controller f ilters packets larger than 1518 bytes) . this allows the packet to be received, and then processed by host software, or to be transmitted on the network. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 24 smsc lan9221/lan9221i datasheet 3.3 address filtering functional description the ethernet address fields of an ethernet packet, consists of two 6-byte fields: one for the destination address and one for the source address. the first bi t of the destination address signifies whether it is a physical address or a multicast address. the lan9221/lan9221i address check logic filters the frame based on the ethernet receive filter mode that has been enabled. filter m odes are specified based on the state of the control bits in ta b l e 3 . 1 , "address filtering modes" , which shows the various filtering modes used by the ethernet mac function. these bits are defined in more detail in the mac control register. please refer to section 5.4.1, "mac_crmac control register" for more information on this register. if the frame fails the filter, the ethernet mac function does not rece ive the packet. the host has the option of accepting or ignoring the packet. figure 3.1 vlan frame table 3.1 address filtering modes mcpas prms invfilt ho hpfilt description 0 0 0 0 0 mac address perfect filtering only for all addresses. 0 0 0 0 1 mac address perfect filtering for physical address and hash filtering for multicast addresses downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 25 revision 2.9 (03-01-12) datasheet 3.4 filtering modes 3.4.1 perfect filtering this filtering mode passes only in coming frames whose destination address field exactly matches the value programmed into the mac address high regi ster and the mac address low register. the mac address is formed by the concatenation of the above two registers in the mac csr function. 3.4.2 hash only filtering this type of filtering checks for incoming receive packets with either multicast or physical destination addresses, and executes an imperfect addr ess filtering against the hash table. during imperfect hash filtering, the destination address in the incoming frame is passed through the crc logic and the upper six bits of the crc register are used to index the contents of the hash table. the hash table is formed by merging the registers multicast hash table high and multicast hash table low in the mac csr function to form a 64-bit hash table. the most significant bit determines the register to be used (high/low), while the other five bits determine the bit within the register. a value of 00000 selects bit 0 of the multicast hash table low register and a value of 11111 selects bit 31 of the multicast hash table high register. 3.4.2.1 hash perfect filtering in hash perfect filtering, if the received frame is a physical address, the lan9221/lan9221i packet filter block perfect-filters the incoming frames de stination field with the value programmed into the mac address high register and the mac address low register. if the incoming frame is a multicast frame, however, the lan9221/lan9221i packet filter function performs an im perfect address filtering against the hash table. the imperfect filtering again st the hash table is the same imperfe ct filtering process described in the hash only filtering section above. 3.4.2.2 inverse filtering in inverse filtering, the packet filter block accepts incoming frames with a destination address not matching the perfect address (i.e., the value progr ammed into the mac address high register and the mac address low register in the crc block and re jects frames with destination addresses matching the perfect address). 0 0 0 1 1 hash filtering for physical and multicast addresses 0 0 1 0 0 inverse filtering x 1 0 x x promiscuous 1 0 0 0 x pass all multicast frames. frames with physical addresses are perfect-filtered 1 0 0 1 1 pass all multicast frames. frames with physical addresses are hash- filtered table 3.1 address filtering modes (continued) mcpas prms invfilt ho hpfilt description downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 26 smsc lan9221/lan9221i datasheet for all filtering modes, when the mcpas bit is set, all multicast frames are accepted. when the prms bit is set, all frames are accepted regardless of th eir destination address. this includes all broadcast frames as well. 3.5 wake-up frame detection setting the wake-up frame enable bit (wuen) in the wucsrwake-up control and status register, places the lan9221/la n9221i mac in the wake-up frame detection mode. in this mode, normal data reception is disabled, and detection lo gic within the mac examines receive data for the pre-programmed wake-up frame patterns. the lan9221/lan9221i can be programmed to notify the host of the wake-up frame detection with the assertio n of the host interrupt (irq) or assertion of the power management event signal (pme). upon dete ction, the wake-up frame received bit (wufr) in the wucsr is set. when the host clears the wu en bit the lan9221/lan9221i will resume normal receive operation. before putting the mac into the wake-up frame dete ction state, the host must provide the detection logic with a list of sample frames and their corresponding byte masks. this information is written into the wake-up frame filter register (wuff). please refer to section 5.4.11, "wuffwake-up frame filter," on page 115 for additional information on this register. the mac supports four programmable filters that su pport many different receive packet patterns. if remote wake-up mode is enabled, the remote wake-up function receives all frames addressed to the mac. it then checks each frame against the enabled filter and recognizes the frame as a remote wake- up frame if it passes the wakeup frame filter registers address filtering and crc value match. in order to determine which bytes of the frames should be checked by the crc module, the mac uses a programmable byte mask and a programmable pattern offset for each of the four supported filters. the patterns offset defines the location of the firs t byte that should be che cked in the frame. since the destination address is checked by the address filtering function, t he pattern offset is always greater than 12. the byte mask is a 31-bit field that specifies whet her or not each of the 31 contiguous bytes within the frame, beginning in the pattern offset, should be checked. if bit j in the byte mask is set, the detection logic checks byte offset +j in the frame. in order to load the wake-up frame filter register, the host lan driver software must perform eight wr ites to the wake-up frame filter register (wuff). the diagram shown in table 3.2, "wake-up frame filter register structure" below, shows the wake- up frame filter registers structure. note 3.1 wake-up frame detection can be performed when the lan9221/lan9221i is in the d0 or d1 power states. in the d0 state, wake-up frame detection is enabled when the wuen bit is set. note 3.2 wake-up frame detection, as well as magi c packet detection, is always enabled and cannot be disabled when the device enters the d1 state. note 3.3 when wake-up frame detection is enabled via the wuen bit of the wucsrwake-up control and status register , a broadcast wake-up frame wil l wake-up the device despite the state of the disable broadcast frame (bcast) bit in the mac_crmac control register . downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 27 revision 2.9 (03-01-12) datasheet the filter i byte mask defines which incoming fram e bytes filter i will examine to determine whether or not this is a wake-up frame. ta b l e 3 . 3 , describes the byte masks bit fields. the filter i command register controls filter i operation. table 3.4 shows the filter i command register. the filter i offset register defines the offset in the frames desti nation address field from which the frames are examined by filter i. ta b l e 3 . 5 describes the filter i offset bit fields. table 3.2 wake-up frame fi lter register structure filter 0 byte mask filter 1 byte mask filter 2 byte mask filter 3 byte mask reserved filter 3 command reserved filter 2 command reserved filter 1 command reserved filter 0 command filter 3 offset filter 2 offset filter 1offset filter 0 offset filter 1 crc-16 filter 0 crc-16 filter 3 crc-16 filter 2 crc-16 table 3.3 filter i byte mask bit definitions filter i byte mask description field description 31 must be zero (0) 30:0 byte mask: if bit j of the byte mask is set, the crc machine processes byte number pattern - (offset + j) of the incoming frame. otherwise, byte pattern - (offset + j) is ignored. table 3.4 filter i co mmand bit definitions filter i commands field description 3 address type: defines the destination address type of the pattern. when bit is set, the pattern applies only to multicast frames. when bit is cleared, the pattern applies only to unicast frames. 2:1 reserved 0 enable filter: when bit is set, filter i is enabled, otherwise, filter i is disabled. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 28 smsc lan9221/lan9221i datasheet the filter i crc-16 register cont ains the crc-16 result of the fr ame that should pass filter i. table 3.6 describes the filter i crc-16 bit fields. 3.5.1 magic packet detection setting the magic packet enable bit (mpen) in the wucsrwake-up control and status register, places the lan9221/lan9221i mac in the magic packet detection mode. in this mode, normal data reception is disabled, and detecti on logic within the mac examines receive data for a magic packet. the lan9221/lan9221i can be programmed to noti fy the host of the magic packet detection with the assertion of the host interr upt (irq) or assertion of the po wer management event signal (pme). upon detection, the magic packet received bit (mpr ) in the wucsr is set. when the host clears the mpen bit the lan9221/lan9221i will resume normal receive operation. please refer to section 5.4.12, "wucsrwake-up control and status register," on page 115 for additional information on this register in magic packet mode, the power management logic constantly monitors each frame addressed to the node for a specific magic packet pattern. it checks only packets with the macs address or a broadcast address to meet the magic packet requ irement. the power management logic checks each received frame for the pattern 48h ff_ff_ff_ff_ff_ff after the destination and source address field. then the function looks in the frame for 16 repetitions of the mac address without any breaks or interruptions. in case of a break in the 16 addr ess repetitions, the pmt function scans for the 48'hff_ff_ff_ff_ff_ff pattern again in the incoming frame. the 16 repetitions may be anywhere in the frame but must be preceded by the synchronization stream. the device will also accept a multicast frame, as l ong as it detects the 16 duplications of the mac address. if the mac address of a node is 00h 11h 22h 33h 44h 55h, then the mac scans for the following data sequence in an ethernet: frame. table 3.5 filter i offset bit definitions filter i offset description field description 7:0 pattern offset: the offset of the first byte in the fram e on which crc is checked for wake-up frame recognition. the minimum value of this field mu st be 12 since there should be no crc check for the destination address and the sour ce address fields. th e mac checks the first offset byte of the frame for crc and checks to determine whether the fr ame is a wake-up frame. offset 0 is the first byte of the incoming frame's destination address. table 3.6 filter i crc-16 bit definitions filter i crc-16 description field description 15:0 pattern crc-16: this field contains the 16-bit crc value from the pattern and the byte mask programmed to the wake-up filter register function. this val ue is compared against the crc calculated on the incoming frame, and a matc h indicates the reception of a wakeup frame. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 29 revision 2.9 (03-01-12) datasheet destination address source ad dress ff ff ff ff ff ff 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 00 11 22 33 44 55 crc it should be noted that magic packet detection can be performed when lan9221/lan9221i is in the d0 or d1 power states. in the d0 state, magic packet detection is enabled when the mpen bit is set. in the d1 state, magic pa cket detection, as well as wake-u p frame detection, are automatically enabled when the device enters the d1 state. 3.6 checksum offload engines (coe) the lan9221/lan9221i contains two checksum offload engines, which offload the calculation of the 16-bit checksum for transmitted and received ethe rnet frames. the functionality of the checksum offload engines is described in the following sections: ? receive checksum offload engine (rxcoe) ? transmit checksum offload engine (txcoe) 3.6.1 receive checksum offload engine (rxcoe) the receive checksum offload engine provides assistance to the cpu by calculati ng a 16-bit checksum for a received ethernet frame. the rxcoe readily supports the following ieee802.3 frame formats: ? type ii ethernet frames ? snap encapsulated frames ? support for up to 2, 802.1q vlan tags the resulting checksum value can also be modified by software to support other frame formats. the rxcoe has two modes of oper ation. in mode 0, the rxcoe calculates the checksum between the first 14 bytes of the ethernet frame and the fcs. this is illustrated in figure 3.2 . in mode 1, the rxcoe supports vlan tags and a snap header. in this mode the rxcoe calculates the checksum at the start of l3 packet. the vlan1 tag register is used by the rxcoe to indicate what protocol type is to be used to indicate the exis tence of a vlan tag. this value is typically 8100h. figure 3.2 rxcoe checksum calculation dst src ty p e frame data f cs calculate checksum downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 30 smsc lan9221/lan9221i datasheet example frame configurations: figure 3.3 type ii ethernet frame figure 3.4 ethernet frame with vlan tag figure 3.5 ethernet frame with length field and snap header dst src pr o t 0 1 2 3 l3 packet f c s calculate checksum 1dword dst src 81 0 0 vi d 0 1 2 3 ty p e 4 l3 packet f c s calculate checksum 1dword dst src l3 packet fc s sn a p 0 sn a p 1 le n 0 1 2 {dsap, ssap, ctrl, oui[23:16]} {oui[15:0], pid[15:0]} 3 4 5 calculate checksum 1dword downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 31 revision 2.9 (03-01-12) datasheet the rxcoe supports a maximum of two vlan tags. if there are more than two vlan tags, the vlan protocol identifier for the third tag is treated as an ethernet type field. the checksum calculation will begin immediately after the type field. the rxcoe resides in the rx path within the ma c. as the rxcoe receives an ethernet frame it calculates the 16-bit checksum. the rxcoe passes the ethernet frame to the rx data fifo with the checksum appended to the end of the frame. the rx coe inserts the checksum immediately after the last byte of the ethernet frame. the packet length field in the rx status word (re fer to section 3.13.3 ) will indicate that the frame size has increased by two bytes to accommodate the checksum. setting the rxcoe_en bit in the coe_crchecksum offload engine control register enables the rxcoe, while the rxcoe_mode bit selects the operating mode. when the rxcoe is disabled, the the received data is simply passed through the rxcoe unmodified. note: software applications must stop the receiver an d flush the rx data path before changing the state of the rxcoe_en or rxcoe_mode bits. note: when the rxcoe is enabled, automatic pad stripping must be disabled (bit 8 (padstr) of the mac_crmac control register ) and vice versa. these functions cannot be enabled simultaneously. figure 3.6 ethernet frame wi th vlan tag and snap header figure 3.7 ethernet frame with multiple vlan tags and snap header dst src 81 0 0 vi d l3 packet f c s sn a p 0 sn a p 1 3 le n 0 1 2 {dsap, ssap, ctrl, oui[23:16]} {oui[15:0], pid[15:0]} 4 5 6 calculate checksum 1dword dst src 81 0 0 vi d l3 packet fc s sn a p 0 sn a p 1 5 le n 0 1 2 {dsap, ssap, ctrl, oui[23:16]} {oui[15:0], pid[15:0]} 6 7 8 calculate checksum 81 0 0 vi d 4 1dword downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 32 smsc lan9221/lan9221i datasheet 3.6.1.1 rx checksum calculation the checksum is calculated 16 bits at a time. in t he case of an odd sized fram e, an extra byte of zero is used to pad up to 16 bits. consider the following packet: da, sa, type, b0, b1, b2 bn, fcs let [a, b] = a*256 + b; if the packet has an even number of octets then checksum = [b1, b0] + c0 + [b3, b2] + c1 + + [bn, bn-1] + cn-1 where c0, c1, ... cn-1 are the carry out results of the intermediate sums. if the packet has an odd number of octets then checksum = [b1, b0] + c0 + [b3, b2] + c1 + + [0, bn] + cn-1 3.6.2 transmit checksum of fload engine (txcoe) the transmit checksum offl oad engine provides assistance to the cpu by calculating a 16-bit checksum, typically for tcp, for a transmit ethernet fr ame. the txcoe calculates the checksum and inserts the results back into the data st ream as it is transferred to the mac. to activate the txcoe an d perform a checksum calc ulation, the host must first set the tx checksum offload engine enable bit (txcoe_en) in the coe_crchecksum offload engine control register . the host then pre-pends a 3 dword buffer to the data that will be transmitted. the pre-pended buffer includes a tx command a, tx command b, and a 32-bit tx checksum preamble. when bit 14 (ck) of the tx command b is set in conjunction with bit 13 (fs) of tx command a and bit 16 (txcoe_en) of the coe_cr register, the txcoe will perform a checksum calculation on the associated packet. when these thr ee bits are set, a 32-bit tx c hecksum preamble must be pre-pended to the beginning of th e tx packet (refer to table 3.7 ). the tx checksum preamble instructs the txcoe on the handling of the asso ciated packet. bits 11:0 of the tx ch ecksum preamble define the byte offset at which the data checksum calc ulation will begin (txcssp). the checksum calculation will begin at this offset and will cont inue until the end of the packet. the da ta checksum calculat ion must not begin in the mac header (first 14 bytes) or in the last 4 bytes of the tx packet. when the calculation is complete, the checksum will be inserted into the packet at the byte offset defined by bits 27:16 of the tx checksum preamble (txcsloc). the tx checksum cannot be inserted in the mac header (first 14 bytes) or in the last 4 bytes of the tx packet. if the ck bit is not set in the first tx command b of a packet, the packet is passed directly throu gh the txcoe without modification, regardless if the txcoe_en is set. an example of a tx packet with a pre-pend ed tx checksum preamble can be found in section 3.12.6.3, "tx example 3" . in this example the host writes the packet data to the ethernet controller in four fragments, the first containi ng the tx checksum preamble. figure 3.23 shows how these fragments are loaded into the tx data fifo. for more information on the tx command a and tx command b, refer to section 3.12.2, "tx command format" . if the tx packet already includes a partial checksum calculation (perhaps inserted by an upper layer protocol), this checksum can be included in the har dware checksum calculati on by setting the txcssp field in the tx checksum preamble to include t he partial checksum. the partial checksum can be replaced by the completed checksum calculation by setting the txcsloc pointer to point to the location of the partial checksum. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 33 revision 2.9 (03-01-12) datasheet note: when the txcoe is enabled, the third dword of the pre-pende d packet is not transmitted. however, 4 bytes must be added to the packet length field in tx command b. note: the tx checksum preamble must be dword-aligned (i.e., the two l east significant bits of the data start offset fields in tx command a must be zero). any valid buffer end alignment setting can be used. note: software applications must stop the transmitter and flush the tx data path before changing the state of the txcoe_en bit. however, the ck bit of tx command b can be set or cleared on a per-packet basis. 3.6.2.1 tx checksum calculation the tx checksum calculation is performed using the sa me operation as the rx checksum shown in section 3.6.1.1 , with the exception that the calculation star ts as indicated by the preamble, and the transmitted checksum is the ones-com pliment of the fi nal calculation. note: when the tx checksum offload feature is invoked, if the calculated checksum is 0000h, it is left unaltered. udp checksums are optional under ipv4, and a zero checksum calculated by the tx checksum offload feature will erroneously indicate to the re ceiver that no checksum was calculated, however, the packet will typically not be rejected by the receiver. under ipv6, however, according to rfc 24 60, the udp checksum is not optional. a calculated checksum that yields a result of zero must be changed to ffffh for insertion into the udp header. ipv6 receivers discard udp packets c ontaining a zero checksum. thus, this feature must not be used for udp checksum calculation under ipv6. 3.7 host bus operations 3.7.1 bus writes the host processor is required to perform two c ontiguous 16-bit writes to complete a single dword transfer. this dword must begin and end on a dword address boundary (a[2] and higher, cannot change during a sixteen bit write). no ordering r equirements exist. the processor can access either the low or high word first, as long as the next writ e is performed to the other word. if a write to the same word is performed, the lan9221/lan9221i disregards the transfer. table 3.7 tx checksum preamble field description 31:28 reserved 27:16 txcsloc - tx checksum location this field specifies the byte offset where the tx checksum will be inserted in the tx packet. the checksum will replace two bytes of data starting at this offset. note: the tx checksum cannot be inserted in the mac header (first 14 bytes) or in the last 4 bytes of the tx packet. 15:12 reserved 11:0 txcssp - tx checksum start pointer this field indicates start offset, in bytes, where the checksum calculation will begin in the associated tx packet. note: the data checksum calculation must not begin in the mac header (first 14 bytes) or in the last 4 bytes of the tx packet. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 34 smsc lan9221/lan9221i datasheet 3.7.2 bus reads the host processor is required to perform two cons ecutive 16-bit reads to complete a single dword transfer. this dword must begin and end on a dword address boundary (a[2] and higher, cannot change during a sixteen bit read). no ordering requ irements exist. the processor can access either the low or high word first, as long as the next re ad is performed from the other word. if a read to the same word is performed, the data read is invalid and should be re-read. this is not a fatal error. the lan9221/lan9221i will reset its read counters and restart a new cycle on the next read. 3.7.3 mixed endian support in order to allow flexibility with a range of desi gns, the lan9221/lan9221i supports mixed endian data fifo accesses. the lan9221/lan9221i provides the ability to select data fifo endianess separately for accesses through the data fifo ports (addresses 00h-3ch) or using the fifo_sel input signal. this is accomplished via the fportend and fselend bits of the hw_cfghardware configuration register , respectively. the fportend bit determines the endianess of rx and tx data fifo host accesses made through the data fifo port addresses ( 00h-3ch). when fportend is cleared, data fifo port accesses utilize little endian byte orderi ng. when fportend is set, data fifo port accesses utilize big endian byte ordering. the fselend bit determines the endianess of rx and tx data fifo host accesses when using the fifo_sel signal. when fselend is cleared, fifo_sel accesses utilize little endian byte ordering. when fselend is set, fifo_sel accesses utilize big endian byte ordering. in addition to mixed endian support, the lan92 21/lan9221i provides a word swap function, as described in section 3.7.4 . the word swap function combined with the endianess select bits described above determines how the data/status fifos and csr host access byte ordering is applied. ta b l e 3 . 8 describes the various operation modes of th e endianess and word swap ordering logic. figure 3.9 illustrates the fifo access byte ordering under various endianess and word swap settings. refer to section 3.7.4 for additional details. note: csr and status fifo accesses are not affect ed by the fportend and fselend endianess select bits. 3.7.4 word swap function in addition to mixed endian functionality, the lan9221/lan9221i supports a word swap function. this feature is controlled by the word swap register, which is described in section 5.3.17, "word_swapword swap control," on page 97 . this register affects how words on the data bus are written to or read from the control and status registers and the transmit and receive data/status fifos. both the word swap function and the mixed endian control bits contain the ability to change the byte ordering of host data path accesses. figure 3.8 illustrates the order in which the word swap and endianess select logic is applied within the lan92 21/lan9221i. logically, the endian control logic is applied after the word swap logic for write operations, and before the word swap logic for read operations. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 35 revision 2.9 (03-01-12) datasheet data path operations for the various supported endianess and word swap configurations are illustrated in figure 3.9 . table 3.8, "endian ordering logic operation" illustrates the byte ordering applied by the endian logic for each type of host access. this figure and table assume an internal byte ordering of 3- 2-1-0, where 3 is the most signific ant byte (data[31:24]) an d 0 is the least signif icant byte (data[7:0]). figure 3.8 lan9221/lan9221i host data path diagram "word swap" logic fifo port endian ordering logic rx/tx data fifo port access (addresses 00h to 3ch) direct fifo access endian ordering logic rx/tx data fifo direct access (fifo_sel = 1) csrs and status fifos fportend (hw_cfg[29]) fselend (hw_cfg[28]) d[15:0] (host data bus) word_swap downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 36 smsc lan9221/lan9221i datasheet figure 3.9 fifo access byte ordering big endian (fportend = 1 for data fifo port access on addresses 00h-3ch) and/or (fselend = 1 for data fifo direct access when fifo_sel=1) 0 1 2 3 0 78 15 16 23 31 24 3 2 1 0 0 78 15 a[1] = 0 a[1] = 1 msb lsb host data bus internal fifo order 0 1 2 3 0 1 2 3 0 78 15 0 78 15 16 23 31 24 a[1] = 0 a[1] = 1 msb lsb host data bus internal fifo order word_swap != ffff_ffffh little endian (fportend = 0 for data fifo port access on addresses 00h-3ch) and/or (fselend = 0 for data fifo direct access when fifo_sel=1) big endian (fportend = 1 for data fifo port access on addresses 00h-3ch) and/or (fselend = 1 for data fifo direct access when fifo_sel=1) 0 1 2 3 0 78 15 16 23 31 24 1 0 3 2 0 78 15 a[1] = 0 a[1] = 1 msb lsb host data bus internal fifo order 0 1 2 3 2 3 0 1 0 78 15 0 78 15 16 23 31 24 a[1] = 0 a[1] = 1 msb lsb host data bus internal fifo order word_swap = ffff_ffffh little endian (fportend = 0 for data fifo port access on addresses 00h-3ch) and/or (fselend = 0 for data fifo direct access when fifo_sel=1) downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 37 revision 2.9 (03-01-12) datasheet 3.8 general purpose timer (gp timer) the general purpose timer is a programmable block that can be used to generate periodic host interrupts. the resolution of this timer is 100us. the gp timer loads the gpt_cnt register with the value in the gpt_load field and begins counting down when the timer_en bit is set to a 1. on a reset, or when the timer_en bit changes from set 1 to cleared 0, the gpt_load field is initialized to ffffh. the gpt_cnt register is also initialized to ffffh on a reset. software can writ e the pre-load value into the gpt_load field at any time; e.g., before or after the timer_en bit is asserted. the gpt enable bit timer_en is located i n the gpt_cfg register. once enabled, the gpt counts down either until it reaches 0000h or until a new pre-load value is written to the gpt_load field. at 0000h, the counter wraps around to ffffh, asserts the gpt interrupt status bit and the irq signal if the gpt_int_en bit is set, and continues counting. the gpt interrupt status bit is in the int_sts register. the gpt_int hardware interrupt can only be set if the gpt_int_en bit is set. gpt_int is a sticky bit (r/wc); i.e., once t he gpt_int bit is set, it can only be cleared by writing a 1 to the bit. table 3.8 endian ordering logic operation fifo access via data fifo port (00h-3ch) direct fifo access via fifo_sel csr access host data bus host data bus host data bus d[15:8] d[7:0] d[15:8] d[7:0] d[15:8] d[7:0] word_swap != ffff_ffffh fportend=0 fselend=0 a1=1 323232 a1=0 101010 fportend=1 fselend=0 a1=1 0 13232 a1=0 2 31010 fportend=0 fselend=1 a1=1 32 0 132 a1=0 10 2 310 fportend=1 fselend=1 a1=1 0 1 0 132 a1=0 2 3 2 310 word_swap = ffff_ffffh fportend=0 fselend=0 a1=1 101010 a1=0 323232 fportend=1 fselend=0 a1=1 2 31010 a1=0 0 13232 portend=0 fselend=1 a1=1 10 2 310 a1=0 32 0 132 portend=1 fselend=1 a1=1 2 3 2 310 a1=0 0 1 0 132 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 38 smsc lan9221/lan9221i datasheet 3.9 eeprom interface the lan9221/lan9221i can optiona lly load its mac address from an external serial eeprom. if a properly configured eeprom is de tected by the lan9221/lan9221i at power-up, hard reset or soft reset, the addrh and addrl registers will be loaded with the contents of the eeprom. if a properly configured eeprom is not detected, it is the responsibility of the host lan driver to set the ieee addresses. the lan9221/lan9221i eeprom controller also allo ws the host system to r ead, write and erase the contents of the serial eeprom . the eeprom controller suppor ts most 93c46 type eeproms configured for 128 x 8-bit operation. 3.9.1 mac address auto-load on power-up, hard reset or soft reset, the eeprom controller attempts to read the first byte of data from the eeprom (address 00h). if the value a5h is read from the first address, then the eeprom controller will assume that an external serial e eprom is present. the eepr om controller will then access the next eeprom byte and send it to the mac address register by te 0 (addrl[7:0]). this process will be repeated for the next five bytes of the mac address, thus fully programming the 48- bit mac address. once all six bytes have been programmed, the mac address loaded bit is set in the e2p_cmd register. a detailed explanation of th e eeprom byte ordering with respect to the mac address is given in section 5.4.3, "addrlmac address low register," on page 110 . if an 0xa5h is not read from the first address, th e eeprom controller will end initialization. it is then the responsibility of the host lan driver software to set the ieee address by writing to the macs addrh and addrl registers. the host can initiate a reload of the mac address from t he eeprom by issuing the reload command via the e2p command (e2p_cmd) register. if the first byte read from the eeprom is not a5h, it is assumed that the eeprom is not pres ent, or not programmed, and the mac address reload will fail. the mac address loaded bit indicates a successful reload of the mac address. 3.9.2 eeprom host operations after the eeprom controller has finished reading (or attempting to read) the mac after power-on, hard reset or soft reset, the host is free to perform other eeprom op erations. eeprom operations are performed using the e2p_cmd and e2 p data (e2p_data) registers. section 5.3.23, "e2p_cmd C eeprom command register," on page 102 provides an explanatio n of the supported eeprom operations. if the eeprom operation is the write location (write) or write all (wral) commands, the host must first write the desired data into the e2p_data register. the host must then issue the write or wral command using the e2p_cmd register by setting the epc_cmd field appropriately. if the operation is a write, the epc_addr field in e2p_cmd must also be set to the desired location. the command is executed when the host sets the epc_ bsy bit high. the completion of the operation is indicated when the epc_bsy bit is cleared. if the eeprom operation is the read location (read) operation, the host must issue the read command using the e2p_cmd with the epc_addr se t to the desired location. the command is executed when the host sets the epc_bsy bit high. the completion of the operation is indicated when the epc_bsy bit is cleared, at which time t he data from the eeprom may be read from the e2p_data register. other eeprom operations are per formed by writing the appropr iate command to the epc_cmd register. the command is executed when the host se ts the epc_bsy bit high. the completion of the operation is indicated when the epc_bsy bit is clear ed. in all cases the host must wait for epc_bsy to clear before modifying the e2p_cmd register. note: the eeprom device powers-up in the erase/write disabled state. to modify the contents of the eeprom the host must fi rst issue the ewen command. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 39 revision 2.9 (03-01-12) datasheet if an operation is attempted, and an eeprom device does not respond within 30ms, the lan9221/lan9221i will timeout, and the epc timeou t bit (epc_to) in the e2p_cmd register will be set. figure 3.10, "eeprom access flow diagram" illustrates the host accesses required to perform an eeprom read or write operation. the host can disable the eeprom interface through the gpio_cfg register. when the interface is disabled, the eedio and eclk signals can be used as general-purpose outputs, or they may be used to monitor internal mii signals. 3.9.2.1 supported eeprom operations the eeprom controller supports the following eeprom operations under host control via the e2p_cmd register. the operations are commonly supported by 93c46 eeprom devices. a description and functional timing diagram is provided below for each operation. please refer to the e2p_cmd register description in section 5.3.23, "e2p_cmd C eeprom command register," on page 102 for e2p_cmd field settings for each command. figure 3.10 eeprom access flow diagram id le write data register write command register read command register busy bit = 0 id le write command regis ter read command regis ter read data regis ter busy bit = 0 eeprom write eeprom read downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 40 smsc lan9221/lan9221i datasheet erase (erase location): if erase/write operations are enabled in the eeprom, this command will erase the location selected by the epc address fi eld (epc_addr). the epc_to bit is set if the eeprom does not respond within 30ms. figure 3.11 eeprom erase cycle eral (erase all): if erase/write operations are enabled in the eeprom, this command will initiate a bulk erase of the entire eeprom .the epc_to bit is set if the eeprom does not respond within 30ms. figure 3.12 eeprom eral cycle 1 eeclk eedio (input) eedio (output) eecs 11 a6 a0 t csl 0 eeclk eedio (input) eedio (output) eecs 10 10 t csl downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 41 revision 2.9 (03-01-12) datasheet ewds (erase/write disable): after issued, the eeprom will ignore erase and write commands. to re-enable erase/write operatio ns issue the ewen command. figure 3.13 eeprom ewds cycle ewen (erase/write enable): enables the eeprom for erase and write operat ions. the eeprom will allow erase and write operations until the era se/write disable command is sent, or until power is cycled. note: the eeprom device will power-up in the erase/ write-disabled state. any erase or write operations will fail until an eras e/write enable command is issued. figure 3.14 eeprom ewen cycle 0 eeclk eedio (input) eedio (output) eecs 10 00 t csl 0 eeclk eedio (input) eedio (output) eecs 10 11 t csl downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 42 smsc lan9221/lan9221i datasheet read (read location): this command will cause a read of the eeprom location pointed to by epc address (epc_addr). the result of the read is available in the e2p_data register. figure 3.15 eeprom read cycle write (write location): if erase/write operatio ns are enabled in the eeprom, this command will cause the contents of the e2p_data register to be writ ten to the eeprom loca tion selected by the epc address field (epc_addr). the epc_to bit is set if the eeprom does not respond within 30ms. figure 3.16 eeprom write cycle wral (write all): if erase/write operations are enabled in the eeprom, this command will cause the contents of the e2p_data register to be written to every eepr om memory location. the epc_to bit is set if the eeprom does not respond within 30ms. figure 3.17 eeprom wral cycle 1 10 a6 eecs eeclk eedio (output) a0 d7 d0 eedio (input) t csl 0 eeclk eedio (input) eedio (output) eecs 11 a6 a0 d7 d0 t csl 0 eeclk eedio (input) eedio (output) eecs 1 d7 d0 0 01 t csl downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 43 revision 2.9 (03-01-12) datasheet table 3.9, "required eeclk cycles" , shown below, shows the number of eeclk cycles required for each eeprom operation. 3.9.2.2 mac address reload the mac address can be reloaded from the eeprom via a host command to the e2p_cmd register. if a value of 0xa5h is not found in the first address of the eeprom, the eeprom is assumed to be un-programmed and mac address reload operation will fail. the mac address loaded bit indicates a successful load of the mac address. the epc_load bit is set after a successful reload of the mac address. 3.9.2.3 eeprom command and data registers refer to section 5.3.23, "e2p_cmd C eeprom command register ," on page 102 and section 5.3.24, "e2p_data C eeprom data register," on page 104 for a detailed description of these registers. supported eeprom operat ions are described in these sections. 3.9.2.4 eeprom timing refer to section 6.11, "eeprom timing," on page 138 for detailed eeprom timing specifications. 3.10 power management the lan9221/lan9221i supports power-down modes to allow applications to minimize power consumption. the following sections describe these modes. 3.10.1 system description power is reduced to various modules by disabling the clocks as outlined in table 3.10, power management states, on page 45. all configuration data is saved when in either of the two low power states. register contents are not affected unless s pecifically indicated in the register description. 3.10.2 functional description there is one normal operating power state, d0 and there are two power saving states: d1, and d2. upon entry into either of the two power saving st ates, only the pmt_ctrl register is accessible for read operations. in either of the power saving st ates the ready bit in the pmt_ctrl register will be cleared. reads of any other addresses are forbidden until the ready bit is set. all writes, with the exception of the wakeup write to byte_test, are also forbidden until the ready bit is set. only when in the d0 (normal) state, when the ready bit is set, can the rest of the device be accessed. table 3.9 required eeclk cycles operation required eeclk cycles erase 10 eral 10 ewds 10 ewen 10 read 18 write 18 wral 18 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 44 smsc lan9221/lan9221i datasheet note 3.4 the lan9221/lan9221i must always be read at least once after power-up, reset, or upon return from a power-saving state, othe rwise write operati ons will not function. in system configurations where the pme signal is shared am ongst multiple devices, the wups field within the pmt_ctrl register can be read to determine which lan9221/lan9221i device is driving the pme signal. when the lan9221/lan9221i is in a power saving state (d1 or d2), a write cycle to the byte_test register will return the lan9221/lan9221i to the d0 state. table 7.2, power consumption device and system components, on page 141 and table 7.2, power consumption device and system components, on page 141 , shows the power consumption values for each power state. note 3.5 when the lan9221/lan9221i is in a power saving state, a write of any data to the byte_test register will wake-up the devi ce. do not perform writes to other addrresses while the ready bit in the pmt_ctrl register is cleared. 3.10.2.1 d1 sleep power consumption is reduced in this state by disabling clocks to portions of the internal logic as shown in table 3.10 . in this mode the clock to the intern al phy and portions of the mac are still operational. this state is enter ed when the host writes a '01' to the pm_mode bits in the power management (pmt_ctrl) register. the ready bit in pmt_ctrl is cleared when entering the d1 state. wake-up frame and magic packet dete ction are automatically enabled in the d1 state. if properly enabled via the wol_en and pme_en bits, the lan9221/lan9221i will assert the pme hardware signal upon the detection of the wake-up frame or magic packet. the lan9221/lan9221i can also assert the host interrupt (irq) on detection of a wake-up frame or magic packet. upon detection, the wups field in pmt_ctrl will be set to a 10b. note 3.6 the pme interrupt status bit (pme_int) in t he int_sts register is set regardless of the setting of pme_en. note 3.7 wake-up frame and magic packet detection is automatically enabled when entering the d1 state. for wake-up fr ame detection, the wake-up frame f ilter must be programmed before entering the d1 state (see section 3.5, "wake-up frame detection," on page 26 ). if used, the host interrupt and pme signal must be enabled prior to entering the d1 state. a write to the byte_test register, regardless of whether a wake-up frame or magic packet was detected, will return lan9221/lan9 221i to the d0 state and will reset the pm_mode field to the d0 state. as noted above, the host is required to c heck the ready bit and verify that it is set before attempting any other reads or writes of the device. note 3.8 the host must only perform read accesses prior to the ready bit being set. once the ready bit is set, the lan9221/lan9221i is ready to resume normal operation. at this time the wups field can be cleared. 3.10.2.2 d2 sleep in this state, as shown in ta b l e 3 . 1 0 , all clocks to the mac and host bus are disabled and the phy is placed in a reduced power state. to enter this stat e, the edpwrdown bit in register 17 of the phy (mode control/status register) must be set. this places the phy in the energy detect mode. the pm_mode bits in the pmt_ctrl register must t hen be set to 10b. upon setting the pm_mode bits, the lan9221/lan9221i will enter the d2 sleep state. the ready bit in pmt_ctrl is cleared when entering the d2 state. note 3.9 if carrier is present when this state is entered detection will occur immediately. if properly enabled via the ed_en and pme_en bits, the lan9221/lan9221i will assert the pme hardware signal upon detection of a valid carrier. upon detection, the wups field in pmt_ctrl will be set to a 01b. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 45 revision 2.9 (03-01-12) datasheet note 3.10 the pme interrupt status bit on the int_sts register (pme_int) is set regardless of the setting of pme_en. a write to the byte_test register, regardless of whether a carrier was detected, will return the lan9221/lan9221i to the d0 state and will reset the pm_mode field to the d0 state. as noted above, the host is required to check the ready bit and veri fy that it is set before attempting any other reads or writes of the device. before the lan9221/lan 9221i is fully awake from this state the edpwrdown bit in register 17 of the phy must be cleared in or der to wake the phy. do not attempt to clear the edpwrdown bit until the ready bit is se t. after clearing the edpwrdown bit the lan9221/lan9221i is ready to resume normal operatio n. at this time the wups field can be cleared. table 3.10 power management states device block d0 (normal operation) d1 (wol) d2 (energy detect) phy full on full on energy detect power-down mac power management full on rx power mgmt. block on off mac and host interface full on off off internal clock full on full on off key clock on block disabled C clock on full off downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 46 smsc lan9221/lan9221i datasheet 3.10.2.3 power management event indicators figure 3.18 is a simplified block diagram of the logic that controls the external pme, and internal pme_interrupt signals. the pme_inter rupt signal is used to set the pme_int status bit in the int_sts register, which, if enabled, will generate a host interr upt upon detection of a power management event. the pme_int status bit in int_sts will remain set un til the internal pme_interrupt signal is cleared by clearing the wups bits, or by clearing the corre sponding wol_en or ed_en bit. after clearing the internal pme_interrupt signal, the pme_int status bit may be cleared by writing a 1 to this bit in the int_sts register. it should be noted that the lan9221/lan9221i can generate a host interrupt regardless of the state of the pme_en bit, or the external pme signal. the external pme signal can be setup for pulsed, or static operation. when the pme_ind bit in the pmt_ctrl register is set to a 1, the external pme signal will be driven active for 50ms upon detection of a wake-up event. when the pme_ind bit is cleared, the pme signal will be driven continuously upon detection of a wake-up event. th e pme signal is deactivated by clearing the wups bits, or by clearing the corresponding wol_en or ed _en bit. the pme signal can also be deactivated by clearing the pme_en bit. figure 3.18 pme and pme_int signal generation 3.10.3 internal phy power-down modes there are 2 power-down modes for the internal phy: 3.10.3.1 general power-down this power-down is controlled by register 0, bit 11. in this mode the internal phy, except the management interface, is powered-down and stays in that condition as long as phy register bit 0.11 is high. when bit 0.11 is cleared, the phy powers up and is automatically reset. please refer to section 5.5.1, "basic control register," on page 118 for additional information on this register. pme ed_en wol_en 50ms pme_en pme_ind pme_pol pme_type logic wuen mpen phy_int wups wups wufr mpr denotes a level-triggered "sticky" status bit pme_int_en pme_int irq_en irq other system interrupts downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 47 revision 2.9 (03-01-12) datasheet 3.10.3.2 energy detect power-down this power-down mode is activated by setting the phy register bit 17.13 to 1. please refer to section 5.5.8, "mode control/status," on page 122 for additional information on this register. in this mode when no energy is present on the line, the phy is powered down, with the e xception of the management interface, the squelch circuit and the energyon logic. the energyon logic is used to detect the presence of valid energy from 100base- tx, 10base-t, or auto-negotiation signals in this mode, when the energyon signal is low, the phy is powered-down, and nothing is transmitted. when energy is received - link pulses or packets - the energyon signal goes high, and the phy powers-up. it automatically resets itself into the state it had prior to power-down, and asserts the int7.1 bit of the register defined in section 5.5.11, "interrupt source flag," on page 125 . if the energyon interrupt is enabled, this event will cause an interrupt to the host. the first and possibly the second packet to activate energyon may be lo st. when 17.13 is low, energy detect power-down is disabled. 3.11 detailed reset description the lan9221/lan9221i has four reset sources: ? hardware reset input pin (nreset) ? soft reset (srst) ? phy soft reset via pmt_ctrl bit 10 (phy_rst) ? phy soft reset via phy basic co ntrol register (phy reg 0.15) table 3.11 shows the effect of the various reset sources on the lan9221/lan9221i's circuitry. note: for proper operation, the lan9221/lan9221i must be reset on power-up via the hardware reset input (nreset) or soft reset (srst). to accomplish this, nreset should be asserted for the minimum period of 30ms at power-up. alternatively, a soft reset may be performed following power-up by setting the srst bit of t he hw_cfg register once the ready bit in the pmt_ctrl register has been set. refer to section 3.11.1, "hardware reset input (nreset)" and section 3.11.3, "soft reset (srst)" for additional information. note 3.11 after any phy reset, the application must wait until the link status bit in the phys basic status register (phy reg. 1.2) is set bef ore attempting to transmit or receive data. note 3.12 after a power-up, nreset or srst, the lan 9221/lan9221i will automatically check for the presence of an external eeprom. after any of these resets the a pplication must verify that the epc busy bit (e2p_cmd, bit 31) is cleared before attempting to access the eeprom, or change the function of the gpo/gpio signals, or before modifying the addrh or addrl registers in the mac. note 3.13 hbi - host bus interface, nasr - not affected by software reset. table 3.11 reset sources and affected circuitry reset source pll hbi note 3.13 nasr registers note 3.13 mil mac phy note 3.11 eeprom mac addr. reload note 3.12 config. straps latched nreset xx x x x x x x srst xx x x phy_rst x phy reg 0.15 x downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 48 smsc lan9221/lan9221i datasheet application note: under normal conditions, the ready bit in pmt_ ctrl will be set (high -1) after an internal reset (22ms). if the software driver polls this bit and it is not set within 100ms, then an error condition occurred. 3.11.1 hardware reset input (nreset) a hardware reset will occur when the nreset input signal is driven low. the ready bit in the pmt_ctrl register can be read fr om the host interface, and will read back a 0 until the hardware reset is complete. upon completion of the hardware reset, the ready bit in pmt_ctrl is set high. after the ready bit is set, the lan9221/lan9221i can be configured via its control registers. the nreset signal is pulled-high internally by the lan9221/lan9221i and can be left unconnected if unused. if used, nreset must be driven low for a minimum period as defined in section 6.10, "reset timing," on page 137 . if nreset is unused, the device mu st be reset following power-up via a soft reset (srst). application note: under normal conditions, the ready bit in pmt_ctrl will be set (high -1) immediately. if the software driver polls this bit and it is not set within 100ms, then an error condition occurred. 3.11.2 resume reset timing after issuing a write to the byte_test register to wake the lan9221/lan9221i from a power-down state, the ready bit in pmt_ctrl will assert (set high) within 2ms. application note: under normal conditions, the ready bit in pm t_ctrl will be set (high -1) within 2 ms. if the software driver polls this bit and it is not set within 100ms, then an error condition occurred. 3.11.3 soft reset (srst) soft reset is initiated by writing a 1 to bit 0 of the hw_cfg register (srst). this self-clearing bit will return to 0 after approximately 2 s, at which time the soft reset is complete. soft reset does not clear control register bits marked as nasr. follo wing power-on, a soft reset must not be performed until the ready bit in the pmt_ctrl register has been set. application note: under normal conditions, the ready bit in pmt_ctrl will be set (high -1) immediately, (within 2 s). if the software driver polls this bit and it is not set within 100ms, then an error condition occurred. 3.11.4 phy reset timing the following sections specify the operation and time required for the internal phy to become operational after various resets or when returning from the reduced power state. 3.11.4.1 phy soft reset via pmt_ctrl bit 10 (phy_rst) the phy soft reset is initiated by writing a 1 to bit 10 of the pmt_ctrl register (phy_rst). this self-clearing bit will return to 0 after approximately 100 s, at which time the phy reset is complete. 3.11.4.2 phy soft reset via phy basic control register (phy reg. 0.15) the phy reg. 0.15 soft reset is in itiated by writing a 1 to bit 15 of the phys basic control register. this self-clearing bit will return to 0 at which time the phy reset is complete. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 49 revision 2.9 (03-01-12) datasheet 3.12 tx data path operation data is queued for transmission by writing it into the tx data fifo. each packet to be transmitted may be divided among multiple buffers. each buffer st arts with a two dword tx command (tx command a and tx command b). the tx command instru cts the lan9221/lan9221i on the handling of the associated buffer. packet boundaries are delineated using control bits within the tx command. the host provides a 16-bit packet tag field in the tx command. the packet tag value is appended to the corresponding tx status dword. all packet tag fields must have the same value for all buffers in a given packet. if tags differ between buffers in the same packet the txe error will be asserted. any value may be chosen for a packet tag as long as all tags in the same packet are identical. packet tags also provide a method of synchronization between transmitted packets and their associated status. software can use unique packet tags to assist with validating matching status completions. note 3.14 the use of packet tags is not required by the hardware. this is a software lan driver only application example for use of this field. a packet length field in the tx command specifies th e number of bytes in the associated packet. all packet length fields must have the same value for all buffers in a given packet. hardware compares the packet length field and the actual amount of data received by the ethernet controller. if the actual packet length count does not match the packet length field as defined in the tx command, the transmitter error (txe) flag is asserted. the lan9221/lan9221i can be programmed to star t payload transmission of a buffer on a byte boundary by setting the data start offset field in the tx command. the data start offset field points to the actual start of the payload data within the fi rst 8 dwords of the buffer. data before the data start offset pointer will be ignored. when a packet is split into multiple buffers, each successive buffer may begin on any arbitrary byte. the lan9221/lan9221i can be programmed to strip padd ing from the end of a transmit packet in the event that the end of the packet does not align wit h the host burst boundary. this feature is necessary when the lan9221/lan9221i is operating in a system that always performs multi-word bursts. in such cases the lan9221/lan9221i must guar antee that it can accept data in multiples of the burst length regardless of the actual packet length. when co nfigured to do so, the lan9221/lan9221i will accept extra data at the end of the packet and will remove the extra padding before transmitting the packet. the lan9221/lan9221i automatically removes data up to the boundary specified in the buffer end alignment field specified in each tx command. the host can instruct the lan9221/lan9221i to issue an interrupt when the buffer has been fully loaded into the tx fifo contained in the lan9221/ lan9221i and transmitted. this feature is enabled through the tx command interrupt on completion field. upon completion of transmission, irrespective of success or failure, the status of the transmission is written to the tx status fifo. tx status is available to the host and may be read using pio operations. an interrupt can be optionally enabled by the host to indicate the availability of a programmable number tx status dwords. before writing the tx command and payload data to the tx fifo, the host mu st check the available tx fifo space by performing a pi o read of the tx_fifo_inf register. the host must ensure that it does not overfill the tx fifo or the tx error (txe) flag will be asserted. the host proceeds to write the tx command by first writing tx command a, then tx command b. after writing the command, the host can then move the payload data into the tx fifo. tx status dwords are stored in the tx status fifo to be read by the host at a later time upon completion of the data transmission onto the wire. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 50 smsc lan9221/lan9221i datasheet 3.12.1 tx buffer format tx buffers exist in the hosts memory in a given fo rmat. the host writes a tx command word into the tx data buffer before moving the ethernet pack et data. the tx command a and command b are 32- bit values that are used by the lan9221/lan9221i in the handling and proc essing of the associated ethernet packet data buffer. buffer alignment, segmentation and other packet processing parameters are included in the command structure. the following diagram illustrates the buffer format. figure 3.19 simplified host tx flow diagram idle check available fifo space init write tx command write buffer last buffer in packet not last buffer write start padding (optional) read tx status (optional) tx status available downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 51 revision 2.9 (03-01-12) datasheet figure 3.20, "tx buffer format" , shows the tx buffer as it is wr itten into the lan9221/lan9221i. it should be noted that not all of the data shown in this diagram is actually stored in the tx data fifo. this must be taken into account when calculating the actual tx data fifo usage. please refer to section 3.12.5, "calculating actual tx data fifo usage," on page 55 for a detailed explanation on calculating the actual tx data fifo usage. 3.12.2 tx command format the tx command instructs the tx fifo controlle r on handling the subsequent buffer. the command precedes the data to be transmitted. the tx command is divided into two, 32-bit words; tx command a and tx command b. there is a 16-bit packet tag in the tx command b command word. packet tags may, if host software desires, be unique for each packet (i.e., an incrementing count). the value of the tag will be returned in the rx status word for the associated packet. the packet tag can be used by host software to uniquely identify each status word as it is returned to the host. both tx command a and tx command b are r equired for each buffer in a given packet. tx command b must be identical for every buffer in a given packet. if the tx command b words do not match, the ethernet controller will as sert the transmitter error (txe) flag. figure 3.20 tx buffer format tx command 'a' offset + data dword0 .. . . . last data & pad 0 31 1st 2nd 3rd last host write order optional pad dword0 .. . optional pad dwordn tx command 'b' optional offset dword0 .. . optional offset dwordn downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 52 smsc lan9221/lan9221i datasheet tx command a table 3.12 tx command 'a' format bits description 31 interrupt on completion. when set, the txdone flag will be asserted when the current buffer has been fully loaded into the tx fifo. this flag may be optionally mapped to a host interrupt. 30:26 reserved. these bits are reserved. always write zeros to this field to guarantee future compatibility. 25:24 buffer end alignment. this field specifies the alignment that must be maintained on the last data transfer of a buffer. the host will add extra dwo rds of data up to the alignment specified in the table below. the lan9221/lan9221i will remove the extra dwords. this mechanism can be used to maintain cache line alignment on host processors. 23:21 reserved. these bits are reserved. always write zeros to this field to guarantee future compatibility 20:16 data start offset (bytes). this field specifies the offs et of the first byte of tx data. the offset value can be anywhere from 0 bytes to 31 a byte offset. 15:14 reserved. these bits are reserved. always write zeros to this field to guarantee future compatibility 13 first segment (fs). when set, this bit indicates that the asso ciated buffer is the first segment of the packet. 12 last segment. when set, this bit indicates that the asso ciated buffer is the last segment of the packet 11 reserved. these bits are reserved. always write zeros to this field to guarantee future compatibility. 10:0 buffer size (bytes). this field indicates the number of bytes contained in the buffer following this command. this value, along with the buffer end alignment field, is read and checked by the lan9221/lan9221i and used to determine how many extra dwords were added to the end of the buffer. a running count is also maintained in the lan9221/lan9221i of the cumulative buffer sizes for a given packet. this cumulative value is compared against the packet length field in the tx command b word and if they do not correlate, the txe flag is set. note: the buffer size specified does not include the buffer end alignment padding or data start offset added to a buffer. [25] [24] end alignment 0 0 4-byte alignment 0 1 16-byte alignment 1 0 32-byte alignment 11 r e s e r v e d downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 53 revision 2.9 (03-01-12) datasheet tx command b 3.12.3 tx data format the tx data section begins at the third dword in the tx buffer (after tx command a and tx command b). the location of the first byte of vali d buffer data to be transmitted is specified in the data start offset field of the tx command a word. table 3.14, "tx data start offset" , shows the correlation between the setting of the lsbs in the data start offset field and the byte location of the first valid data byte. additionally, transmit buffer data can be offset by up to 7 additional dwords as indicated by the upper three msbs (5:2) in the data start offset field. tx data is contiguous until the end of the buffer. the buffer may end on a byte boundary. unused bytes at the end of the packet will not be sent to the mil for transmission. the buffer end alignment field in tx command a specifies the alignment t hat must be maintained for the associated buffer. end alignment may be specified as 4-, 16-, or 32-byte. the host processor is responsible for adding the additional data to the end of the buffer. the hardware will automatically remove this extra data. 3.12.3.1 tx buffer fragmentation rules transmit buffers must adhere to the following rules: ? each buffer can start and end on any arbitrary byte alignment ? the first buffer of any transmit packet can be any length table 3.13 tx command 'b' format bits description 31:16 packet tag. the host should write a unique packet identifier to this field. this identifier is added to the corresponding tx status word and can be used by the host to correlate tx status words with their corresponding packets. note: the use of packet tags is not required by the hardware. this field can be used by the lan software driver for any application. packet tags is one application example. 15 reserved. this bit is reserved. always write zeros to this bit to guarantee future compatibility. 14 tx checksum enable (ck). when this bit is set in conjunction wi th the first segment (fs) bit in tx command a and the tx checksum offload engine enable bit (txcoe_en) in the coe_crchecksum offload engine control register , the tx checksum offload engine (txcoe) will calculate a l3 checksum for the associated frame. 13 add crc disable. when set, the automatic addi tion of the crc is disabled. 12 disable ethernet frame padding. when set, this bit prevents the automatic addition of padding to an ethernet frame of less than 64 bytes. the crc field is also added despite the state of the add crc disable field. 11 reserved. these bits are reserved. always write zeros to this field to guarantee future compatibility. 10:0 packet length (bytes). this field indicates the total number of bytes in the current packet. this length does not include the offset or padding. if the packet length field does not match the actual number of bytes in the packet the transmitter error (txe) flag will be set. table 3.14 tx data start offset data start offset [1:0]: 11 10 01 00 first tx data byte: d[31:24] d[23:16] d[15:8] d[7:0] downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 54 smsc lan9221/lan9221i datasheet ? middle buffers (i.e., those with first segment = la st segment = 0) must be greater than, or equal to 4 bytes in length ? the final buffer of any transmit packet can be any length the mil operates in store-and-forward mode and has specific rules with respect to fragmented packets. the total space consumed in the tx fifo (mil) must be limited to no more than 2kb - 3 dwords (2,036 bytes total). any transmit packet that is so highly fragmented that it takes more space than this must be un-fragmented (by copying to a driver-supplied buffer) before the transmit packet can be sent to the lan9221/lan9221i. one approach to determine whether a packet is too fragmented is to calculate the actual amount of space that it will consume, and check it against 2, 036 bytes. another approach is to check the number of buffers against a worst-case limit of 86 (see explanation below). 3.12.3.2 calculating worst-ca se tx fifo (mil) usage the actual space consumed by a buffer in the mil tx fifo consists only of any partial dword offsets in the first/last dword of the buffer, plus all of the whole dwords in between. any whole dword offsets and/or alignments are stripped off before the buffer is loaded into the tx data fifo, and tx command words are stripped off before the buffer is written to the mil tx fifo, so none of those dwords count as space consumed. the worst-case overhead for a tx buffer is 6 bytes, which assumes that it started on the high byte of a dword and ended on the low byte of a dword. a tx packet consisting of 86 such fragments would have an overhead of 516 bytes (6 * 86) which, when added to a 1514-byte max-size transmit packet (15 16 bytes, rounded up to the next whole dword), would give a total space consumption of 2,032 bytes, leaving 4 bytes to spare; this is the basis for the "86 fragment" rule mentioned above. 3.12.4 tx status format tx status is passed to the host cpu through a sepa rate fifo mechanism. a status word is returned for each packet transmitted. data transmission is suspended if the tx status fifo becomes full. data transmission will resume when the host reads the tx status and there is ro om in the fifo for more tx status data. the host can optionally choose to not read the tx st atus. the host can optionally ignore the tx status by setting the tx status discard allow overrun en able (txsao) bit in the tx configuration register (tx_cfg). if this option is chosen tx status will not be written to the fifo. setting this bit high allows the transmitter to continue operation with a full tx status fifo. in this mode the status information is still available in the tx status fifo, and tx status interrupts still function. in the case of an overrun, the txsused counter will stay at zero and no further tx status will be written to the tx status fifo until the host frees space by reading tx status. if txsao is enabled, a txe error will not be generated if the tx status fifo overruns. in this mode the host is responsible for re-synchronizing tx status in the case of an overrun. bits description 31:16 packet tag. unique identifier written by the host into the packet tag field of the tx command b word. this field can be used by the host to correlate tx status words with the associated tx packets. 15 error status (es). when set, this bit indicates that the ethernet controller has reported an error. this bit is the logical or of bits 11, 10 , 9, 8, 2, 1 in this status word. 14:12 reserved. these bits are reserved. always write zeros to this field to guarantee future compatibility. 11 loss of carrier. when set, this bit indicates the loss of carrier during transmission. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 55 revision 2.9 (03-01-12) datasheet 3.12.5 calculating actual tx data fifo usage the following rules are used to calculate the actual tx data fifo space consumed by a tx packet: ? tx command 'a' is stored in the tx data fifo for every tx buffer. ? tx command 'b' is written into the tx data fifo when the first segment (fs) bit is set in tx command 'a.' ? when tx checksum is enabled, t he 4-byte tx checksum preamble is written into tx data fifo. ? any dword-long data added as part of the data start offset is removed from each buffer before the data is written to the tx data fifo. any dat a that is less than 1 dword is passed to the tx data fifo. ? payload from each buffer within a packe t is written into the tx data fifo. ? any dword-long data added as part of the end p adding is removed from each buffer before the data is written to the tx data fifo. any end padd ing that is less than 1 dword is passed to the tx data fifo. 10 no carrier. when set, this bit indicates that the carrier signal from the transceiver was not present during transmission. note: during 10/100 mbps full-duplex transmission, the value of this bit is invalid and should be ignored. 9 late collision. when set, indicates that the packet transmission was aborted after the collision window of 64 bytes. 8 excessive collisions. when set, this bit indicates that th e transmission was aborted after 16 collisions while attempting to transmit the current packet. 7 reserved. this bit is reserved. always write zeros to this field to guarantee future compatibility. 6:3 collision count. this counter indicates the number of colli sions that occurred before the packet was transmitted. it is not valid when excessive collisions (bit 8) is also set. 2 excessive deferral. if the deferred bit is set in the contro l register, the setting of the excessive deferral bit indicates that the transmission has ended because of a deferral of over 24288 bit times during transmission. 1 reserved. this bit is reserved. always write zero to this bit to guarantee future compatibility. 0 deferred. when set, this bit indicates that the current packet transmission was deferred. bits description downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 56 smsc lan9221/lan9221i datasheet 3.12.6 transmit examples 3.12.6.1 tx example 1 in this example a single, 111-byte ethernet packet will be transmitted. this packet is divided into three buffers. the three buffers are as follows: buffer 0: ? 7-byte data start offset ? 79-bytes of payload data ? 16-byte buffer end alignment buffer 1: ? 0-byte data start offset ? 15-bytes of payload data ? 16-byte buffer end alignment buffer 2: ? 10-byte data start offset ? 17-bytes of payload data ? 16-byte buffer end alignment downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 57 revision 2.9 (03-01-12) datasheet figure 3.21, "tx example 1" illustrates the tx command structure for this example, and also shows how data is passed to the tx data fifo. figure 3.21 tx example 1 tx command 'a' 0 31 tx command 'b' pad dword 1 7-byte data start offset 10-byte end padding 79-byte payload buffer end alignment = 1 data start offset = 7 first segment = 1 last segment = 0 buf f er size = 79 packet length = 111 tx command 'a' 0 31 tx command 'b' 10-byte end offset padding 15-byte payload buffer end alignment = 1 data start offset = 0 first segment = 0 last segment = 0 buf f er size = 15 packet length = 111 tx command 'a' 0 31 tx command 'b' buffer end alignment = 1 data start of fset = 10 first segment = 0 last segment = 1 buf f er size = 17 packet length = 111 10-byte data start offset tx command 'a' tx command 'b' tx command 'a' tx command 'b' tx command 'a' tx command 'b' note: extra bytes betw een buff ers are not transmitted data written to the ethernet controller data passed to the tx data fifo 5-byte end padding 17-byte payload data tx command 'a' tx command 'b' tx command 'a' tx command 'a' 1b 79-byte payload 15-byte payload 17-byte payload downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 58 smsc lan9221/lan9221i datasheet 3.12.6.2 tx example 2 in this example, a single 183-byte ethernet packet will be transmitted. this packet is in a single buffer as follows: ? 2-byte data start offset ? 183-bytes of payload data ? 4-byte buffer end alignment figure 3.22, "tx example 2" illustrates the tx command structure for this example, and also shows how data is passed to the tx data fifo. note that the packet resides in a single tx buffer, therefore both the fs and ls bits are set in tx command a. figure 3.22 tx example 2 tx command 'a' 0 31 tx command 'b' data written to the ethernet controller tx command 'b' 183-byte payload data data start offset = 6 first segment = 1 last segment = 1 buffer size =183 packet length = 183 tx command 'a' tx command 'b' data passed to the tx data fifo buffer end alignment = 0 3b end padding tx command 'a' 6-byte data start offset 183-byte payload data note: extra bytes between buffers are not transmitted downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 59 revision 2.9 (03-01-12) datasheet 3.12.6.3 tx example 3 in this example a single, 111-byte ethernet packet wi ll be transmitted with a tx checksum. this packet is divided into four buffers. the four buffers are as follows: buffer 0: ? 4-byte data start offset ? 4-byte checksum preamble ? 16-byte buffer end alignment buffer 1: ? 7-byte data start offset ? 79-bytes of payload data ? 16-byte buffer end alignment buffer 2: ? 0-byte data start offset ? 15-bytes of payload data ? 16-byte buffer end alignment buffer 3: ? 10-byte data start offset ? 17-bytes of payload data ? 16-byte buffer end alignment figure 3.21, "tx example 1" illustrates the tx command structure for this example, and also shows how data is passed to the tx data fifo. note: in order to perform a tx checksu m calculation on the associated packet, bit 14 (ck) of the tx command b must be set in conjunction with bit 13 (fs) of tx command a and bit 16 (txcoe_en) of the coe_cr register . for more information, refer to section 3.6.2, "transmit checksum offload engine (txcoe)" . downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 60 smsc lan9221/lan9221i datasheet figure 3.23 tx example 3 0 31 tx command 'b' pad dword 1 7-byte data start offset 10-byte end padding buffer end alignment = 1 data start offset = 7 first segment = 0 last segment = 0 buffer size = 79 0 31 tx command 'b' 10-byte end offset padding buffer end alignment = 1 data start offset = 0 first segment = 0 last segment = 0 buffer size = 15 0 31 tx command 'b' buffer end alignment = 1 data start offset = 10 first segment = 0 last segment = 1 buffer size = 17 10-byte data start offset tx command 'a' tx command 'b' tx command 'a' tx command 'b' tx command 'a' tx command 'b' note: extra bytes between buffers are not transmitted data written to the ethernet controller data passed to the tx data fifo 5-byte end padding tx command 'a' tx command 'b' tx command 'a' tx command 'a' 79-byte payload 15-byte payload 17-byte payload note: when enabled, the tx checksum preamble is pre-pended to data to be transmitted. the fs bit in tx command 'a', the ck bit in tx command 'b' and the txcoe_en bit in the coe_cr register must all be set for the tx checksum to be generated. fs must not be set for subsequent fragments of the same packet. tx command 'a' 0 31 tx command 'b' 4-byte data start offset 8-byte end padding buffer end alignment = 1 data start offset = 4 first segment = 1 last segment = 0 buffer size = 4 packet length = 115 tx command 'a' tx command 'b' tx checksum location = 50 checksum preamble tx checksum start pointer = 14 tx command 'a' tx checksum preamble tx checksum preamble tx command 'a' 79-byte payload tx command 'a' 15-byte payload 1b tx command 'a' 17-byte payload data tx checksum enable = 1 packet length = 115 tx checksum enable = 1 packet length = 115 tx checksum enable = 1 packet length = 115 tx checksum enable = 1 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 61 revision 2.9 (03-01-12) datasheet 3.12.7 transmitter errors if the transmitter error (txe) flag is asserted fo r any reason, the transmit ter will continue operation. tx error (txe) will be asserted under the following conditions: ? if the actual packet length count does not matc h the packet length field as defined in the tx command. ? both tx command a and tx command b are required for each buffer in a given packet. tx command b must be identical for every buffer in a given packet. if the tx command b words do not match, the ethernet controller will assert the transmitter error (txe) flag. ? host overrun of the tx data fifo. ? overrun of the tx status fi fo (unless t xsao is enabled) 3.12.8 stopping and starting the transmitter to halt the transmitter, the host mu st set the tx_stop bit in the tx _cfg register. the transmitter will finish sending the current frame (if there is a fr ame transmission in progress). when the transmitter has received the tx status for this frame, it will clear the tx_stop and tx_on bits, and will pulse the txstop_int. once stopped, the host can optionally clear the tx status and tx data fifos. the host must re-enable the transmitter by setting the tx_o n bit. if the there ar e frames pending in the tx data fifo (i.e., tx data fifo was not purged), the transmission will resume with this data. 3.13 rx data path operation when an ethernet packet is received , the mil first begins to transfer the rx data. this data is loaded into the rx data fifo. the rx da ta fifo pointers are updated as data is written into the fifo. the last transfer from the mil is the rx status word. the lan9221/lan9221i implements a separate fifo for the rx status words. the total availabl e rx data and status queued in the rx fifo can be read from the rx_fifo_inf register. the host may read any number of available rx status words before reading the rx data fifo. the host must use caution when reading the rx data and status. the host must never read more data than what is available in the fifos. if this is attempted an underrun condition will occur. if this error occurs, the ethernet controller will assert the receiv er error (rxe) interrupt. if an underrun condition occurs, a soft reset is required to regain host synchronization. a configurable beginning offset is supported in the lan9221/lan9221i. the rx data offset field in the rx_cfg register controls the number of bytes that the beginning of the rx data buffer is shifted. the host can set an offset from 0-31 bytes. the offset may be changed in between rx packets, but it must not be changed during an rx packet read. the lan9221/lan9221i can be programmed to add padding at the end of a receive packet in the event that the end of the packet does not align wit h the host burst boundary. this feature is necessary when the lan9221/lan9221i is operating in a system t hat always performs multi-dword bursts. in such cases the lan9221/lan9221i must guarantee that it can transfer data in multiples of the burst length regardless of the actual packet length. when configured to do so, the lan9221/lan9221i will add extra data at the end of the packet to allo w the host to perform the necessary number of reads so that the burst length is not cut short. once a packet has been padded by the h/w, it is the responsibility of the host to interrogate the packe t length field in the rx status and determine how much padding to discard at the end of the packet. it is possible to read multiple packets out of t he rx data fifo in one c ontinuous stream. it should be noted that the programmed offset and padding will be added to each individual packet in the stream, since packet boundaries are maintained. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 62 smsc lan9221/lan9221i datasheet 3.13.1 rx slave pio operation using pio mode, the host can eith er implement a polling or interru pt scheme to empty the received packet out of the rx data fifo. the host will remain in the idle state until it receives an indication (interrupt or polling) that data is available in the rx data fifo. the host will then read the rx status fifo to get the packet status, which will contain the packet length and any other status information. the host should perform the proper number of reads, as indicated by the packet length plus the start offset and the amount of optional padding added to t he end of the frame, from the rx data fifo. figure 3.24 host receive routine using interrupts figure 3.25 host receive routine with polling not last packet idle read rx status dword init read rx packet last packet rx interrupt read rx_fifo_ inf read rx status dword init read rx packet last packet not last packet valid status dword downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 63 revision 2.9 (03-01-12) datasheet 3.13.1.1 receive data fifo fast forward the rx data path implements an automatic data discard function. using the rx data fifo fast forward bit (rx_ffwd) in the rx_dp_ctrl regist er, the host can instruct the lan9221/lan9221i to skip the packet at the head of the rx data fi fo. the rx data fifo po inters are automatically incremented to the beginning of the next rx packet. when performing a fast-forward, there must be at leas t 4 dwords of data in the rx data fifo for the packet being discarded. for less than 4 dwords do not use rx_ffwd. in this case data must be read from the rx data fifo and discard ed using standard pio read operations. after initiating a fast-forward operation, do not perform any reads of the rx data fifo until the rx_ffwd bit is cleared. other resources can be acce ssed during this time (i.e., any registers and/or the other three fifos). also note that the rx_ffw d will only fast-forward the rx data fifo, not the rx status fifo. the receiver does not have to be stopped to perform a fast-forward operation. 3.13.1.2 force receiver discard (receiver dump) in addition to the receive data fast forward featur e, lan9221/lan9221i also implements a receiver "dump" feature. this feat ure allows the host processor to flush the entire contents of the rx data and rx status fifos. when activated, the read and writ e pointers for the rx data and status fifos will be returned to their reset state. to perform a rece iver dump, the lan9221/lan9221i receiver must be halted. once the receiver stop completion is co nfirmed, the rx_dump bit can be set in the rx_cfg register. the rx_dump bit is cleared when the dump is complete. for more information on stopping the receiver, please refer to section 3.13.4, "stopping and starting the receiver," on page 66 . for more information on the rx_dump bit, please refer to section 5.3.7, "rx_cfgreceive configuration register," on page 86 . downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 64 smsc lan9221/lan9221i datasheet 3.13.2 rx packet format the rx status words can be read from the rx status fifo port, while the rx data packets can be read from the rx data fifo. rx data packets are formatted in a specific manner before the host can read them as shown in figure 3.26 . it is assumed that the host ha s previously read the associated status word from the rx status fifo, to ascertain the data size and any error conditions. figure 3.27 shows the rx packet format when the rx checksum is enabled. the rx checksum data appended to the data payload is treated just as an additional 4-bytes within the rx data fifo. the rx checksum is enabled by se tting the rxcoe_en bit in the coe_crchecksum offload engine control register . for more informati on on the rx ch ecksum, refer to section 3.6.1, "receive checksum offload engine (rxcoe)" . figure 3.26 rx packet format ofs + first data dword .. . . last data dword 0 31 host read order 1st 2nd last optional pad dword0 .. optional pad dwordn optional offset dword0 .. optional offset dwordn downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 65 revision 2.9 (03-01-12) datasheet 3.13.3 rx status format figure 3.27 rx packet format with rx checksum bits description 31 reserved. this bit is reserved. reads 0. 30 filtering fail. when set, this bit indicates that the associated frame failed the address recognizing filtering. 29:16 packet length. the size, in bytes, of the corresponding received frame. 15 error status (es). when set this bit indicates that the mil has reported an error. this bit is the internal logical or of bits 11,7,6 and 1. 14 reserved. these bits are reserved. reads 0. 13 broadcast frame. when set, this bit indicates that the received frame has a broadcast address. 12 length error (le). when set, this bit indicates that the actual length does not match with the length/type field of the received frame. 11 runt frame. when set, this bit indicates that frame was prematurely terminated before the collision window (64 bytes). runt frames are passed on to the host only if the pass bad frames bit mac_cr bit [16] is set. 10 multicast frame. when set, this bit indicates that t he received frame has a multicast address. 9:8 reserved. these bits are reserved. reads 0. ofs + first data dword .. . . last data dword 0 31 host read order 1st 2nd last optional pad dword0 .. optional pad dwordn optional offset dword0 .. optional offset dwordn rx checksum downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 66 smsc lan9221/lan9221i datasheet 3.13.4 stopping and starting the receiver to stop the receiver, the host must clear the rxen bit in the mac control register. when the receiver is halted, the rxstop_int will be pulsed. once st opped, the host can optiona lly clear the rx status and rx data fifos. the host must re-enable the receiver by setting the rxen bit. 3.13.5 receiver errors if the receiver error (rxe) flag is asserted for any reason, the receiver will continue operation. rx error (rxe) will be asserted under the following conditions: ? a host underrun of rx data fifo ? a host underrun of the rx status fifo ? an overrun of the rx status fifo it is the duty of the host to identi fy and resolve any error conditions. 7 frame too long. when set, this bit indicates that the frame length exceeds the maximum ethernet specification of 1518 bytes. this is only a frame too long indication and will not cause the frame reception to be truncated. 6 collision seen. when set, this bit indicates that the fram e has seen a collision after the collision window. this indicates that a late collision has occurred. 5 frame type. when set, this bit indicates that the frame is an ethernet-type frame (length/type field in the frame is greater than 1500). when reset, it indicates the incoming frame was an 802.3 type frame. this bit is not set for runt frames less than 14 bytes. 4 receive watchdog time-out. when set, this bit indicates that the incoming frame is greater than 2048 bytes through 2560 bytes, therefore expiring the receive watchdog timer. 3 mii error. when set, this bit indicates that a receive erro r (rx_er asserted) was detected during frame reception. 2 dribbling bit. when set, this bit indicates that the frame contained a non-integer multiple of 8 bits. this error is reported only if the number of dribbling bi ts in the last byte is 4 in the mii operating mode, or at least 3 in the 10 mbps oper ating mode. this bit will not be se t when the collision seen bit[6] is set. if set and the crc error bit[1] is cleared, then the packet is considered to be valid. 1 crc error. when set, this bit indicates that a crc error was detected. this bit is also set when the rx_er pin is asserted during the reception of a fr ame even though the crc may be correct. this bit is not valid if the received frame is a runt fram e, or a late collision was detected or when the watchdog time-out occurs. 0 reserved. these bits are reserved. reads 0 bits description downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 67 revision 2.9 (03-01-12) datasheet chapter 4 internal ethernet phy 4.1 top level functional description functionally, the internal phy can be divided into the following sections: ? 100base-tx transmit and receive ? 10base-t transmit and receive ? internal mii interface to the ethernet media access controller ? auto-negotiation to automatically dete rmine the best speed and duplex possible ? management control to read status r egisters and write control registers figure 4.1 100base-tx data path 4.2 100base-tx transmit the data path of the 100base-tx is shown in figure 4.1 . each major block is explained below. 4.2.1 4b/5b encoding the transmit data passes from the mii block to th e 4b/5b encoder. this block encodes the data from 4-bit nibbles to 5-bit symbols (known as code-groups) according to ta b l e 4 . 1 . each 4-bit data-nibble is mapped to 16 of the 32 possible code-groups. the remaining 16 code-groups are either used for control information or are not valid. the first 16 code-groups are referred to by the hexadecimal values of their corresponding data nibbles, 0 through f. the remaining code-groups are given le tter designations with slashes on either side. for example, an idle code-group is /i/, a tr ansmit error code-group is /h/, etc. the encoding process may be bypassed by clearing bit 6 of register 31. when the encoding is bypassed the 5 th transmit data bit is equivalent to tx_er. mac tx driver mlt-3 converter nrzi converter 4b/5b encoder magnetics cat-5 rj45 100m pll internal mii 25 mhz by 4 bits tx_clk 25mhz by 5 bits nrzi mlt-3 mlt-3 mlt-3 mlt-3 scrambler and piso 125 mbps serial mii 25mhz by 4 bits downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 68 smsc lan9221/lan9221i datasheet table 4.1 4b/5b code table code group sym receiver interpretation transmitter interpretation 11110 0 0 0000 data 0 0000 data 01001 1 1 0001 1 0001 10100 2 2 0010 2 0010 10101 3 3 0011 3 0011 01010 4 4 0100 4 0100 01011 5 5 0101 5 0101 01110 6 6 0110 6 0110 01111 7 7 0111 7 0111 10010 8 8 1000 8 1000 10011 9 9 1001 9 1001 10110 a a 1010 a 1010 10111 b b 1011 b 1011 11010 c c 1100 c 1100 11011 d d 1101 d 1101 11100 e e 1110 e 1110 11101 f f 1111 f 1111 11111 i idle sent after /t/r until tx_en 11000 j first nibble of ssd, translated to 0101 following idle, else rx_er sent for rising tx_en 10001 k second nibble of ssd, translated to 0101 following j, else rx_er sent for rising tx_en 01101 t first nibble of esd, causes de-assertion of crs if followed by /r/, else assertion of rx_er sent for falling tx_en 00111 r second nibble of esd, causes deassertion of crs if following /t/, else assertion of rx_er sent for falling tx_en 00100 h transmit error symbol sent for rising tx_er 00110 v invalid, rx_er if during rx_dv invalid 11001 v invalid, rx_er if during rx_dv invalid 00000 v invalid, rx_er if during rx_dv invalid 00001 v invalid, rx_er if during rx_dv invalid 00010 v invalid, rx_er if during rx_dv invalid 00011 v invalid, rx_er if during rx_dv invalid downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 69 revision 2.9 (03-01-12) datasheet 4.2.2 scrambling repeated data patterns (especially the idle code-grou p) can have power spectral densities with large narrow-band peaks. scrambling the data helps eliminate these peaks and spread the signal power more uniformly over the entire channel bandwidth. this uniform spectral density is required by fcc regulations to prevent excessive emi fr om being radiated by the physical wiring. the scrambler also performs the parallel in serial out conversion (piso) of the data. 4.2.3 nrzi and mlt3 encoding the scrambler block passes the 5-bit wide parallel data to the nrzi converter where it becomes a serial 125mhz nrzi data stream. the nrzi is encoded to mlt-3. mlt3 is a tri-level code where a change in the logic level represents a code bit 1 and the logic output remaining at the same level represents a code bit 0. 4.2.4 100m transmit driver the mlt3 data is then passed to the analog transmitte r, which launches the differential mlt-3 signal, on outputs txp and txn, to the t wisted pair media via a 1:1 ratio isolation transformer. the 10base- t and 100base-tx signals pass through the same transformer so that common magnetics can be used for both. the transmit ter drives into the 100 impedance of the cat-5 cable. cable termination and impedance matching require external components. 4.2.5 100m phase lock loop (pll) the 100m pll locks onto reference clock and genera tes the 125mhz clock used to drive the 125 mhz logic and the 100base-tx transmitter. 00101 v invalid, rx_er if during rx_dv invalid 01000 v invalid, rx_er if during rx_dv invalid 01100 v invalid, rx_er if during rx_dv invalid 10000 v invalid, rx_er if during rx_dv invalid table 4.1 4b/5b code table (continued) code group sym receiver interpretation transmitter interpretation downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 70 smsc lan9221/lan9221i datasheet figure 4.2 receive data path 4.3 100base-tx receive the receive data path is shown in figure 4.2 . detailed descriptions are given below. 4.3.1 100m receive input the mlt-3 from the cable is fed into the phy (on inputs rxp and rxn) via a 1:1 ratio transformer. the adc samples the incoming differential signal at a rate of 125m samples per second. using a 64- level quanitizer it generates 6 digital bits to repr esent each sample. the dsp adjusts the gain of the adc according to the observed signal levels such t hat the full dynamic range of the adc can be used. 4.3.2 equalizer, baseline wander corr ection and clock and data recovery the 6 bits from the adc are fed into the dsp bl ock. the equalizer in the dsp section compensates for phase and amplitude distortion caused by the ph ysical channel consisting of magnetics, connectors, and cat- 5 cable. the equalizer can restore the signal for any good-quality cat-5 cable between 1m and 150m. if the dc content of the signal is such that the low-frequency comp onents fall below the low frequency pole of the isolation transformer, then the dro op characteristics of t he transformer will become significant and baseline wander (blw) on the receiv ed signal will result. to prevent corruption of the received data, the phy corrects for blw and c an receive the ansi x3.263-1995 fddi tp-pmd defined killer packet with no bit errors. the 100m pll generates multiple phas es of the 125mhz clock. a mult iplexer, controlled by the timing unit of the dsp, selects the optimum phase for sa mpling the data. this is used as the received recovered clock. this clock is used to ex tract the serial data from the received signal. 4.3.3 nrzi and mlt-3 decoding the dsp generates the mlt-3 recovered levels that are fed to the mlt-3 converter. the mlt-3 is then converted to an nrzi data stream. mac a/d converter mlt-3 converter nrzi converter 4b/5b decoder magnetics cat-5 rj45 100m pll internal mii 25mhz by 4 bits rx_clk 25mhz by 5 bits nrzi mlt-3 mlt-3 mlt-3 6 bit data descrambler and sipo 125 mbps serial dsp: timing recovery, equalizer and blw correction mlt-3 mii 25mhz by 4 bits downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 71 revision 2.9 (03-01-12) datasheet 4.3.4 descrambling the descrambler performs an inverse function to the scrambler in the transmitter and also performs the serial in parallel out (sipo) conversion of the data. during reception of idle (/i/) symbols. the descrambler synchron izes its descrambler key to the incoming stream. once synchronizati on is achieved, the descrambler locks on this key and is able to descramble incoming data. special logic in the descrambler ensures synchronization with the remote phy by searching for idle symbols within a window of 4000 bytes (40us). th is window ensures that a maximum packet size of 1514 bytes, allowed by the ieee 802. 3 standard, can be received with no interference. if no idle- symbols are detected within this time-period, rece ive operation is aborted and the descrambler re-starts the synchronization process. the descrambler can be bypassed by setting bit 0 of register 31. 4.3.5 alignment the de-scrambled signal is then aligned into 5-bit code-groups by recognizing the /j/k/ start-of-stream delimiter (ssd) pair at the start of a packet. once the code-word alignment is determined, it is stored and utilized until the next start of frame. 4.3.6 5b/4b decoding the 5-bit code-groups are translated into 4-bit data nibbles according to the 4b/5b table. the ssd, /j/k/, is translated to 0101 0101 as the first 2 nibbles of the mac preamble. reception of the ssd causes the phy to assert the internal rx_dv signal , indicating that valid data is available on the internal rxd bus. successive valid code-groups are tran slated to data nibbles. reception of either the end of stream delimiter (esd) consisting of the /t /r/ symbols, or at least two /i/ symbols causes the phy to de-assert the internal carrier sense and rx_dv. these symbols are not translated into data. 4.4 10base-t transmit data to be transmitted comes from the mac layer controller. the 10base-t transmitter receives 4-bit nibbles from the mii at a rate of 2.5mhz and converts them to a 10mbps serial data stream. the data stream is then manchester-encoded and sent to the analog transmitter, which drives a signal onto the twisted pair via the external magnetics. the 10m transmitter uses the following blocks: ? mii (digital) ? tx 10m (digital) ? 10m transmitter (analog) ? 10m pll (analog) 4.4.1 10m transmit data across the internal mii bus the mac controller drives the transmit data onto the internal txd bus. when the controller has driven tx_en high to indicate valid data, the data is latched by the mii block on the rising edge of tx_clk. the data is in the form of 4-bit wide 2.5mhz data. 4.4.2 manchester encoding the 4-bit wide data is sent to the tx10m block. th e nibbles are converted to a 10mbps serial nrzi data stream. the 10m pll locks onto the external cl ock or internal oscillator and produces a 20mhz downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 72 smsc lan9221/lan9221i datasheet clock. this is used to manchester encode the nrz data stream. when no data is being transmitted (tx_en is low), the tx10m block outputs normal link pulses (nlps) to maintain communications with the remote link partner. 4.4.3 10m transmit drivers the manchester encoded data is sent to the anal og transmitter where it is shaped and filtered before being driven out as a differential signal across the txp and txn outputs. 4.5 10base-t receive the 10base-t receiver gets the manchester- encoded analog signal from the cable via the magnetics. it recovers the receive clock from the signal and uses this clock to recover the nrzi data stream. this 10m serial data is converted to 4-bit data nibbles which are passed to the controller across the mii at a rate of 2.5mhz. this 10m receiver uses the following blocks: ? filter and squelch (analog) ? 10m pll (analog) ? rx 10m (digital) ? mii (digital) 4.5.1 10m receive input and squelch the manchester signal from the cable is fed into the phy (on inputs rxp and rxn) via 1:1 ratio magnetics. it is first filtered to reduce any out-of-band noise. it then passes through a squelch circuit. the squelch is a set of amplitude and timi ng comparators that norma lly reject differential voltage levels below 300mv and detect and recognize differential voltages above 585mv. 4.5.2 manchester decoding the output of the squelch goes to the rx10m block where it is validated as manchester encoded data. the polarity of the signal is also checked. if the polarity is reversed (local rxp is connected to rxn of the remote partner and vice versa), then this is identified and corrected. the reversed condition is indicated by the flag xpol, bit 4 in register 27. the 10m pll is locked onto the received manchester signal and from this, generates the received 20mhz clock. using this clock, the manchester encoded data is extracted and conver ted to a 10mhz nrzi dat a stream. it is then converted from serial to 4-bit wide parallel data. the rx10m block also detects valid 10base-t idle signals - normal link pulses (nlps) - to maintain the link. 4.5.3 jabber detection jabber is a condition in which a station transmits for a period of time longer than the maximum permissible packet length, usually due to a fault condi tion, that results in ho lding the tx_en input for a long period. special logic is used to detect the jabber state and abort the transmission to the line, within 45ms. once tx_en is deasserted, the logic resets the jabber condition. 4.6 auto-negotiation the purpose of the auto-negotiation function is to automatically configure the phy to the optimum link parameters based on the capabilities of its li nk partner. auto-negotiation is a mechanism for exchanging configuration information between two lin k-partners and automatically selecting the highest downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 73 revision 2.9 (03-01-12) datasheet performance mode of operation supported by both si des. auto-negotiation is fully defined in clause 28 of the ieee 802.3 specification. once auto-negotiation has completed, information ab out the resolved link can be passed back to the controller via the internal serial management interf ace (smi). the results of the negotiation process are reflected in the speed indication bits in register 31, as well as the link partner ability register (register 5). the auto-negotiation protocol is a purely physical layer activity and proceeds independently of the mac controller. the advertised capabilities of the phy are stored in register 4 of the sm i registers. the default advertised by the phy is determined by user-defined on-chip signal options. the following blocks are activated during an auto-negotiation session: ? auto-negotiation (digital) ? 100m adc (analog) ? 100m pll (analog) ? 100m equalizer/blw/clock recovery (dsp) ? 10m squelch (analog) ? 10m pll (analog) ? 10m transmitter (analog) when enabled, auto-negotiation is started by t he occurrence of one of the following events: ? hardware reset ? software reset ? power-down reset ? link status down ? setting register 0, bit 9 high (auto-negotiation restart) on detection of one of these events, the phy begins auto-negotiation by transmitting bursts of fast link pulses (flp). these are bur sts of link pulses from the 10m transmitter. they are shaped as normal link pulses and can pass uncorrupted down cat-3 or cat-5 cable. a fast link pulse burst consists of up to 33 pulses. the 17 odd-number ed pulses, which are always present, frame the flp burst. the 16 even-numbered pulses, which may be present or absent, cont ain the data word being transmitted. presence of a data pulse repres ents a 1, while absence represents a 0. the data transmitted by an flp burst is known as a link code word. these are defined fully in ieee 802.3 clause 28. in summary, the phy advertises 8 02.3 compliance in its selector field (the first 5 bits of the link code word). it advertises its technology abi lity according to the bits set in register 4 of the smi registers. there are 4 possible matches of the technology ab ilities. in the order of priority these are: ? 100m full-duplex (highest priority) ? 100m half-duplex ? 10m full-duplex ? 10m half-duplex if the full capabilities of the phy are advertised (100 m, full-duplex), and if the link partner is capable of 10m and 100m, then auto-negotiation selects 1 00m as the highest performance mode. if the link partner is capable of half and full-duplex modes, th en auto-negotiation selects full-duplex as the highest performance operation. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 74 smsc lan9221/lan9221i datasheet once a capability match has been determined, the link code words are repeated with the acknowledge bit set. any difference in the main content of the link code words at this time will cause auto-negotiation to re-start. auto-negotiation will also re-start if not all of the required flp bursts are received. writing register 4 bits [8:5] allows software control of the capabilities advertised by the phy. writing register 4 does not automatically re -start auto-negotiation. register 0, bit 9 must be set before the new abilities will be advertised. auto-negotiation can also be disabled via software by clearing register 0, bit 12. the lan9221/lan9221i does not support next page" capability. 4.7 parallel detection if the lan9221/lan9221i is connected to a devic e lacking the ability to auto-negotiate (i.e. no flps are detected), it is able to determine the speed of the link based on either 100m mlt-3 symbols or 10m normal link pulses. in this case the link is presumed to be half-duplex per the ieee standard. this ability is known as parallel detection. this feature ensures inter operability with legacy link partners. if a link is formed via para llel detection, then bit 0 in register 6 is cleared to indicate that the link partner is not capable of auto-negotiation. the ethernet mac has access to this information via the management interface. if a fault occurs during parallel detection, bit 4 of register 6 is set. register 5 is used to store the link partner abilit y information, which is coded in the received flps. if the link partner is not auto-negotiation capable, then register 5 is updated after completion of parallel detection to reflect the speed capability of the link partner. 4.7.1 re-starting auto-negotiation auto-negotiation can be re-started at any time by sett ing register 0, bit 9. auto-negotiation will also re- start if the link is broken at any time. a broken link is caused by signal loss. this may occur because of a cable break, or because of an interruption in the signal transmitted by the link partner. auto- negotiation resumes in an attempt to determine the new link configuration. if the management entity re-starts au to-negotiation by writing to bi t 9 of the control register, the lan9221/lan9221i will respond by stopping al l transmission/receiving operations. once the break_link_timer is done, in the auto-negotiation state-machine (approximately 1200ms) the auto- negotiation will re-start. the link partner will have also dropped the link due to lack of a received signal, so it too will resume auto-negotiation. 4.7.2 disabling auto-negotiation auto-negotiation can be disabled by setting regist er 0, bit 12 to zero. the device will then force its speed of operation to reflect the information in regist er 0, bit 13 (speed) and register 0, bit 8 (duplex). the speed and duplex bits in register 0 should be ignored when auto-negotiation is enabled. 4.7.3 half vs. full-duplex half-duplex operation relies on the csma/cd (carrier sense multiple access / collision detect) protocol to handle network traffic and collisions. in this mode, the internal carrier sense signal, crs, responds to both transmit and receive activity. in this mode, if data is received while the phy is transmitting, a collision results. in full-duplex mode, the phy is able to transmit and receive data simultaneously. in this mode, the internal crs responds only to receive activity. the csma/cd protocol does not apply and collision detection is disabled. table 4.2 describes the behavior of the internal crs bit under all receive/transmit conditions. the internal crs signal is used to trigger bit 10 (no carrier) of the tx status word (see section 3.12.4, "tx status format" ). the crs value, and subsequ ently the no carrier value, are invalid during any downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 75 revision 2.9 (03-01-12) datasheet full-duplex transmission. therefore, these signals cannot be used as a verification method of transmitted packets when transmitting in 10/100 mbps full-duplex modes. note 4.1 the lan9221/lan9221i 10/100 phy internal crs signal operates in two modes: active and low. when in active mode, the internal crs will transition high and low upon line activity, where a high value indicates a carrier has been detected. in low mode, the internal crs stays low and does not indicate carrier detection. the internal crs signal and no carrier (bit 10 of the tx status word ) cannot be used as a verification method of transmitted packets when transmitting in 10/100 mbps full-duplex mode. 4.8 hp auto-mdix hp auto-mdix facilitates the use of cat-3 (10 base -t) or cat-5 (100 base-t ) media utp interconnect cable without consideration of inte rface wiring scheme. if a user plugs in either a direct connect lan cable, or a cross-over patch cable, as shown in figure 4.3 , the smsc lan9221/lan9221i auto-mdix phy is capable of configuring the tpo and tpi t wisted pair pins for correct transceiver operation. the internal logic of the device detects the tx and rx pins of the connecting device. since the rx and tx line pairs are interchangeable, special pcb design considerations are needed to accommodate the symmetrical magnetics and termination of an auto-mdix design. the auto-mdix function can be disabled through an inte rnal register 27.15, or the external control pins amdix_en. when disabled the tx and rx pins can be configured with the channel select (ch_select) pin as desired. table 4.2 crs behavior mode speed duplex activity crs behavior ( note 4.1 ) manual 10 mbps half-duplex transmitting active manual 10 mbps half-duplex receiving active manual 10 mbps full-duplex transmitting low manual 10 mbps full-duplex receiving active manual 100 mbps half-duplex transmitting active manual 100 mbps half-duplex receiving active manual 100 mbps full-duplex transmitting low manual 100 mbps full-duplex receiving active auto-negotiation 10 mbps half-duplex transmitting active auto-negotiation 10 mbps half-duplex receiving active auto-negotiation 10 mbps full-duplex transmitting low auto-negotiation 10 mbps full-duplex receiving active auto-negotiation 100 mbps half-duplex transmitting active auto-negotiation 100 mbps half-duplex receiving active auto-negotiation 100 mbps full-duplex transmitting low auto-negotiation 100 mbps full-duplex receiving active downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 76 smsc lan9221/lan9221i datasheet the figure below shows the signal names at the rj-45 connector, the mapping of these signals to the pins on the lan9221/lan9221i is as follows: txp = tpo+ txn = tpo- rxp = tpi+ rxn = tpi- figure 4.3 direct cable connection vs. cross-over cable connection downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 77 revision 2.9 (03-01-12) datasheet chapter 5 register description the following section describes all lan9 221/lan9221i registers and data ports. figure 5.1 memory map mac csr port a4h b0h base + 00h reserved b4h a0h rx data fifo port tx data fifo port rx status fifo por t 40h 20h 50h fch eeprom port 04h 1ch rx dat a fifo alias po r t s 24h 3ch tx data fifo alias por ts rx status fifo peek 44h tx status fifo port 48h tx status fifo peek 4ch a8h ach downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 78 smsc lan9221/lan9221i datasheet 5.1 register nomenclatu re and access attributes 5.2 rx and tx fifo ports the lan9221/lan9221i contains four host-accessible fifos: rx status, rx data, tx status, and tx data fifos. the sizes data fifos and the rx st atus fifo are configur able through the csrs. 5.2.1 rx fifo ports the rx data path contains two read-only fifos: rx status and rx data. the rx status fifo has two ports at different address locations. the rx stat us fifo port causes t he top of the rx status fifo to be popped, and is destructive. the rx status fifo peek port allows the top of the rx status fifo to be read without popping it. the rx data fifo has a single port; reading data from this port always causes the top of the rx data fifo to be popped. this port is aliased to 16 word locations. the host may access the top of the rx data fifo through any of these locations. 5.2.2 tx fifo ports the tx data path consists of two fifos, tx stat us and rx data. the tx st atus fifo also has two ports at different locations. when the tx status fifo port is read, the top of the tx status fifo is popped. when the tx status fifo peek port is read, the top of the tx status fifo is not popped. the tx data fifo is write only. it is aliased to 16 word locations . the host may access the top of the tx data fifo through any of these locations. symbol description ro read only: if a register is read only, writ es to this register have no effect. wo write only: if a register is write only, reads always return 0. r/w read/write: a register with this attri bute can be read and written r/wc read/write clear: a register bit with this attribute can be read and written. however, a write of a 1 clears (sets to 0) the corresponding bit and a write of a 0 has no effect. rc read to clear: a register bit with this attribute is cleared when read. ll latch low: clear on read of register lh latch high: clear on read of register sc self-clearing nasr not affected by software reset reserved bits certain bits within registers are listed as reser ved. unless stated otherwise, these bits must be written with zeros for futu re compatibility. the values of these bits are not guaranteed when read. reserved registers certain configuration registers within the lan9221/lan9221i ar e listed as reserved. these registers are not guaranteed to return any particular value when read. these registers must not be written to or modified by system failure; doing so could result in failure of the de vice and system. default states at reset - system reset or soft ware reset - internal registers are set to their default states. the default states provide a minimum level of fu nctionality needed to successfully bring up a system, but do not necessarily provide desired or optimal conf iguration of the device. it is the responsibility of the system initialization software to properly determine the operating parameters and optional system features that ar e applicable, and to program the lan9221/lan9221i registers accordingly. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 79 revision 2.9 (03-01-12) datasheet 5.3 system contro l and status registers table 5.1, "direct address register map" , lists the registers that are directly addressable by the host bus. table 5.1 direct address register map control and status registers base address + offset symbol register name default 50h id_rev chip id and revision. see page 80. 54h irq_cfg main interrupt configuration 00000000h 58h int_sts interrupt status 00000000h 5ch int_en interrupt enable register 00000000h 60h reserved reserved for future use - 64h byte_test read-only byte order testing register 87654321h 68h fifo_int fifo level interrupts 48000000h 6ch rx_cfg receive configuration 00000000h 70h tx_cfg transmit configuration 00000000h 74h hw_cfg hardware configuration 00050000h 78h rx_dp_ctl rx datapath control 00000000h 7ch rx_fifo_inf receive fifo information 00000000h 80h tx_fifo_inf transmit fifo information 00001200h 84h pmt_ctrl power management control 00000000h 88h gpio_cfg general purpose io configuration 00000000h 8ch gpt_cfg general purpose timer configuration 0000ffffh 90h gpt_cnt general purpose timer count 0000ffffh 94h reserved reserved for future use - 98h word_swap word swap register 00000000h 9ch free_run free run counter - a0h rx_drop rx dropped frames counter 00000000h a4h mac_csr_cmd mac csr synchronizer command (mac csrs are indexed through this register) 00000000h a8h mac_csr_data mac csr synchronizer data 00000000h ach afc_cfg automatic flow control configuration 00000000h b0h e2p_cmd eeprom command 00000000h b4h e2p_data eeprom data 00000000h b8h - fch reserved reserved for future use - downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 80 smsc lan9221/lan9221i datasheet 5.3.1 id_revchip id and revision this register contains the id and revision fields for this design. 5.3.2 irq_cfginterrupt configuration register this register configures and indi cates the state of the irq signal. offset: 50h size: 32 bits bits description type default 31-16 chip id. this read-only field identifies this design ro 9221h 15-0 chip revision ro 0000h offset: 54h size: 32 bits bits description type default 31:24 interrupt deassertion interval (int_deas). this field determines the interrupt request deassertion interval in multiples of 10 microseconds. setting this field to zero causes the device to disable the int_deas interval, reset the interval counter, and issue any pending interrupts. if a new, non-zero value is written to this field, any subsequent interrupts will obey the new setting. note: this field does not apply to the pme interrupt. r/w 0 23-15 reserved ro - 14 interrupt deassertion interval clear (int_deas_clr ). writing a one to this register clears the de-asserti on counter in the irq controller, thus causing a new de-assertion interval to begin (regardless of whether or not the irq controller is currently in an active de-assertion interval). sc 0 13 interrupt deassertion status (int_deas_sts). when set, this bit indicates that interrupts are currently in a deassertion interval, and will not be delivered to the irq pin. when this bit is clear, interrupts are not currently in a deassertion interval, and will be delivered to the irq pin. sc 0 12 master interrupt (irq_int). this read-only bit indicates the state of the internal irq line, regardless of the se tting of the irq_en bit, or the state of the interrupt de-assertion function. when this bit is high, one of the enabled interrupts is currently active. ro 0 11-9 reserved ro - 8 irq enable (irq_en) C this bit controls the final interrupt output to the irq pin. when clear, the irq output is disabled and permanently deasserted. this bit has no effect on any internal interrupt status bits. r/w 0 7-5 reserved ro - downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 81 revision 2.9 (03-01-12) datasheet 4 irq polarity (irq_pol) C when cleared, enables the irq line to function as an active low output. when set, the irq output is active high. when irq is configured as an open- drain output this field is ignored, and the interrupt output is always active low. r/w nasr 0 3-1 reserved ro - 0 irq buffer type (irq_type) C when cleared, enables irq to function as an open-drain buffer for use in a wired-or interrupt configuration. when set, the irq output is a push-pull driver. when configured as an open-drain output the irq_pol field is ignored, and the interrupt output is always active low. r/w nasr 0 bits description type default downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 82 smsc lan9221/lan9221i datasheet 5.3.3 int_stsinterrupt status register this register contains the current status of the generated interrupts. writing a 1 to the corresponding bits acknowledges and clears the interrupt. offset: 58h size: 32 bits bits description type default 31 software interrupt (sw_int). this interrupt is generated when the sw_int_en bit is set high. writing a one clears this interrupt. r/wc 0 30-26 reserved ro - 25 tx stopped (txstop_int). this interrupt is issued when stop_tx bit in tx_cfg is set, and the transmitter is halted. r/wc 0 24 rx stopped (rxstop_int). t his interrupt is issued when the receiver is halted. r/wc 0 23 rx dropped frame counter halfway (rxdfh_int). this interrupt is issued when the rx dropped frames counter counts past its halfway point (7fffffffh to 80000000h). r/wc 0 22 reserved ro 0 21 tx ioc interrupt (tx_ioc). when a buffer with the ioc flag set has finished being loaded into the tx fifo, this interrupt is generated. r/wc 0 20 rx dma interrupt (rxd_int). this interrupt is issued when the amount of data programmed in the rx dma count (rx_dma_cnt) field of the rx_cfg register has been transferred out of the rx fifo. r/wc 0 19 gp timer (gpt_int). this interrupt is issued when the general purpose timer wraps past zero to ffffh. r/wc 0 18 phy (phy_int). indicates a phy interrupt event. ro 0 17 power management even t interrupt (pme_int). this interrupt is issued when a power management event is detected as configured in the pmt_ctrl register. this interrupt functions independent of the pme signal, and will still function if the pme signal is disabled. writing a '1' clears this bit regardless of the state of the pme hardware signal. notes: ? detection of a power management event, and assertion of the pme signal will not wakeup the lan9221/lan9221i. the lan9221/lan9221i will only wake up when it detects a host write cycle of any data to the byte_test register. ? the interrupt deassertion interval does not apply to the pme interrupt. r/wc 0 16 tx status fifo overflow (txso). generated when the tx status fifo overflows. r/wc 0 15 receive watchdog time-out (rwt). interrupt is generated when a packet larger than 2048 bytes has been received. r/wc 0 14 receiver error (rxe). indicates that the receiver has encountered an error. please refer to section 3.13.5, "receiver errors," on page 66 for a description of the conditions that will cause an rxe. r/wc 0 13 transmitter error (txe). when generated, indicates that the transmitter has encountered an error. please refer to section 3.12.7, "transmitter errors," on page 61 , for a description of the conditions that will cause a txe. r/wc 0 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 83 revision 2.9 (03-01-12) datasheet 12:11 reserved ro - 10 tx data fifo overrun interrupt (tdfo). generated when the tx data fifo is full, and anothe r write is attempted. r/wc 0 9 tx data fifo available interrupt (tdfa). generated when the tx data fifo available space is greater than the programmed level. r/wc 0 8 tx status fifo full interrupt (tsff). generated when the tx status fifo is full. r/wc 0 7 tx status fifo level interrupt (tsfl). generated when the tx status fifo reaches the programmed level. r/wc 0 6 rx dropped frame interrupt (rxdf_int). this interrupt is issued whenever a receive frame is dropped. r/wc 0 5 reserved ro - 4 rx status fifo full interrupt (rsff). generated when the rx status fifo is full. r/wc 0 3 rx status fifo level interrupt (rsfl). generated when the rx status fifo reaches the programmed level. r/wc 0 2-0 gpio [2:0] (gpiox_int). interrupts are generated from the gpios. these interrupts are configured through the gpio_cfg register. r/wc 000 bits description type default downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 84 smsc lan9221/lan9221i datasheet 5.3.4 int_eninterrupt enable register this register cont ains the interrupt masks for irq. writin g 1 to any of the bits enables the corresponding interrupt as a source for irq. bits in the int_sts register will still reflect the status of the interrupt source regardless of whether the sour ce is enabled as an interrupt in this register. offset: 5ch size: 32 bits bits descriptio n type default 31 software interrupt (sw_int_en) r/w 0 30:26 reserved ro - 25 tx stopped interrupt enable (txstop_int_en) r/w 0 24 rx stopped interrupt enable (rxstop_int_en) r/w 0 23 rx dropped frame counter halfway in terrupt enable (rxdfh_int_en). r/w 0 22 reserved ro 0 21 tx ioc interrupt enable (tioc_int_en) r/w 0 20 rx dma interrupt (rxd_int). r/w 0 19 gp timer (gpt_int_en) r/w 0 18 phy (phy_int_en) r/w 0 17 power management event inte rrupt enable (pme_int_en) r/w 0 16 tx status fifo overflow (txso_en) r/w 0 15 receive watchdog time-out interrupt (rwt_int_en) r/w 0 14 receiver error interrupt (rxe_int_en) r/w 0 13 transmitter error inte rrupt (txe_int_en) r/w 0 12:11 reserved ro - 10 tx data fifo overrun in terrupt (tdfo_int_en) r/w 0 9 tx data fifo available interrupt (tdfa_int_en) r/w 0 8 tx status fifo full interrupt (tsff_int_en) r/w 0 7 tx status fifo level interrupt (tsfl_int_en) r/w 0 6 rx dropped frame interrupt enable (rxdf_int_en) r/w 0 5 reserved ro - 4 rx status fifo full interrupt (rsff_int_en) r/w 0 3 rx status fifo level interrupt (rsfl_int_en) r/w 0 2-0 gpio [2:0] (gpiox_int_en). r/w 000 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 85 revision 2.9 (03-01-12) datasheet 5.3.5 byte_testbyte order test register this register can be used to determine t he byte ordering of the current configuration 5.3.6 fifo_intfifo level interrupts this register configures the li mits where the fifo controllers will generate system interrupts. offset: 64h size: 32 bits bits description type default 31:0 byte test ro 87654321h offset: 68h size: 32 bits bits description type default 31-24 tx data available level. the value in this field sets the level, in number of 64 byte blocks, at wh ich the tx fifo available interrupt (tfda) will be generated. when the tx data fifo fr ee space is greater than this value a tx fifo available interrupt (tdfa) will be generated. r/w 48h 23-16 tx status level. the value in this field sets the level, in number of dwords, at which the tx status fi fo level interrupt (tsfl) will be generated. when the tx stat us fifo used space is greater than this value an tx status fifo level interr upt (tsfl) will be generated. r/w 00h 15-8 reserved ro - 7-0 rx status level. the value in this field sets the level, in number of dwords, at which the rx status fi fo level interrupt (rsfl) will be generated. when the rx status fifo used space is greater than this value an rx status fifo level interrupt (rsfl) will be generated. r/w 00h downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 86 smsc lan9221/lan9221i datasheet 5.3.7 rx_cfgreceive configuration register this register controls the la n9221/lan9221i receive engine. offset: 6ch size: 32 bits bits description type default 31:30 rx end alignment. this field specifies the alignment that must be maintained on the last data transfer of a buffer. the lan9221/lan9221i will add extra dwords of data up to the alignment specified in the table below. the host is responsible for removing these extra dwords. this mechanism can be used to maintain cache line alignment on host processors. please refer to ta b l e 5 . 2 for bit definitions note: the desired rx end alignment must be set before reading a packet. the rx end alignment can be changed between reading receive packets, but must not be changed if the packet is partially read. r/w 00b 29-28 reserved ro - 27-16 rx dma count (rx_dma_cnt). this 12-bit field indicates the amount of data, in dwords, to be transferred out of the rx data fifo before asserting the rxd_int. after being set, this field is decremented for each dword of data that is read from the rx data fifo. this field can be overwritten with a new value before it reaches zero. r/w 000h 15 force rx discard (rx_dump). this self-clearing bit clears the rx data and status fifos of all pending data. when a 1 is written, the rx data and status pointers are cleared to zero. note: please refer to section force receiver discard (receiver dump) on page 63 for a detailed description regarding the use of rx_dump. sc 0 14-13 reserved ro - 12-8 rx data offset (rxdoff). this field controls the offset value, in bytes, that is added to the beginning of an rx data packet. the start of the valid data will be shifted by the number of bytes specified in this field. an offset of 0-31 bytes is a valid number of offset bytes. note: the two lsbs of this field (d[9 :8]) must not be modified while the rx is running. the receiver must be halted, and all data purged before these two bits can be modified. the upper three bits (dword offset) may be modified while the receiver is running. modifications to the upper bits will take affect on the next dword read. r/w 00000 7-0 reserved ro - table 5.2 rx alignm ent bit definitions [31] [30] end alignment 0 0 4-byte alignment 0 1 16-byte alignment 1 0 32-byte alignment 11 r e s e r v e d downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 87 revision 2.9 (03-01-12) datasheet 5.3.8 tx_cfgtransmit configuration register this register controls the transmit functions on the lan9 221/lan9221i ethernet controller. offset: 70h size: 32 bits bits description type default 31-16 reserved. ro - 15 force tx status discard (txs_dump). this self-clearing bit clears the tx status fifo of all pending status dwords. when a 1 is written, the tx status pointers are cleared to zero. sc 0 14 force tx data discard (txd_dump). this self-clearing bit clears the tx data fifo of all pending data. when a 1 is written, the tx data pointers are cleared to zero. sc 0 13-3 reserved ro - 2 tx status allow overrun (txsao). when this bit is cleared, data transmission is suspended if the tx st atus fifo becomes full. setting this bit high allows the transmitter to continue operation with a full tx status fifo. note: this bit does not affect the operat ion of the tx status fifo full interrupt. r/w 0 1 transmitter en able (tx_on). when this bit is set (1), the transmitter is enabled. any data in the tx fifo will be sent. this bit is cleared automatically when stop_tx is se t and the transmitter is halted. r/w 0 0 stop transmitter (stop_tx). when this bit is set (1), the transmitter will finish the current frame, and will then stop transmitting. when the transmitter has stopped this bit will clear. all writes to this bit are ignored while this bit is high. sc 0 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 88 smsc lan9221/lan9221i datasheet 5.3.9 hw_cfghardware configuration register note: the transmitter and receiver must be stopped be fore writing to this register. refer to section 3.12.8, "stopping and starting the transmitter," on page 61 and section 3.13.4, "stopping and starting the receiver," on page 66 for details on stopping the transmitter and receiver. offset: 74h size: 32 bits bits description type default 31 reserved ro - 30 reserved ro - 29 fifo port endian or dering (fportend). this control bit determines the endianess of rx and tx data fifo host accesses when accessed through the rx/tx data fifo ports, including the alias addresses (any access from 00h to 3ch). when this bit is cleared, data fifo port accesses utilize little endian byte ordering. when this bit is set, data fifo port accesses utilize big endian byte ordering. please refer to section section 3.7.3, "mixed endian support," on page 34 for more information on this feature. r/w nasr 0 28 direct fifo access endian ordering (fselend). this control bit determines the endianess of rx and tx data fifo host accesses when accessed using the fifo_sel signal. when this bit is cleared, fifo_sel accesses utilize little endian byte ordering. when this bit is set, fifo_sel accesses utilize big endian byte ordering. please refer to section section 3.7.3, "mixed endian support," on page 34 for more information on this feature. r/w nasr 0 27-25 reserved ro - 24 amdix_en strap state . this read-only bit reflects the state of the amdix_en strap pin (pin 73). this pin can be overridden by phy registers 27.15 and 27.13 ro amdix strap pin 23-21 reserved ro 20 must be one (mbo). this bit must be set to 1 for normal device operation. r/w 0 16-19 tx fifo size (tx_fif_sz). sets the size of the tx fifos in 1kb values to a maximum of 14kb. the tx status fifo consumes 512 bytes of the space allocated by tx_fif_siz, and the tx data fifo consumes the remaining space specified by tx_fif _sz. the minimum size of the tx fifos is 2kb (tx data and status combi ned). the tx data fifo is used for both tx data and tx commands. the rx status and data fifos consume the remaining space, which is equal to 16kb C tx_fif_siz. see section 5.3.9.1, "allowable settings for configurable fifo memory allocation," on page 90 for more information. r/w 5h 15-2 reserved ro - 1 soft reset timeout (srst_to) . if a software reset is attempted when the phy is not in the operational state (rx_clk and tx_clk running), the reset will not complete and the soft reset operation will timeout and this bit will be set to a 1. the host processor must correct the pr oblem and issue another soft reset. ro 0 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 89 revision 2.9 (03-01-12) datasheet 0 soft reset (srst) . writing 1 generates a software initiated reset. this reset generates a full reset of the mac csrs. the scsrs (system command and status registers) are reset except for any nasr bits. soft reset also clears any tx or rx errors (txe/rxe). this bit is self-clearing. notes: ? do not attempt a soft reset unless th e phy is fully awake and operational. after a phy reset, or when returning from a reduced power state, the phy must given adequate time to return to the operational state before a soft reset can be issued. ? the lan9221/lan9221i must always be read at least once after power- up, reset, or upon return from a powe r-saving state or write operations will not function. sc 0 bits description type default downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 90 smsc lan9221/lan9221i datasheet 5.3.9.1 allowable settings for configurable fifo memory allocation tx and rx fifo space is configurable through th e csr - hw_cfg register defined above. the user must select the fifo allocation by setting the tx fifo size (tx_fif_sz) field in the hardware configuration (hw_cfg) register. th e tx_fif_sz field selects the total allocation for the tx data path, including the tx status fifo size. the tx status fifo size is fixed at 512 bytes (128 tx status dwords). the tx status fifo length is subtract ed from the total tx fifo size with the remainder being the tx data fifo size. note that tx dat a fifo space includes both commands and payload data. rx fifo size is the remainder of the unallocated fifo space (16384 bytes C tx fifo size). the rx status fifo size is always equal to 1/16 of the rx fifo size. the rx status fifo length is subtracted from the total rx fifo size with the remainder being the rx data fifo size. for example, if tx_fif_sz = 6 then: total tx fifo size = 6144 bytes (6kb) tx status fifo size = 512 bytes (fixed) tx data fifo size = 6144 C 512 = 5632 bytes rx fifo size = 16384 C 6144 = 10240 bytes (10kb) rx status fifo size = 10240 / 16 = 640 bytes (160 rx status dwords) rx data fifo size = 10240 C 640 = 9600 bytes table 5.3 shows every valid setting for t he tx_fif_sz field. note that se ttings not shown in this table are reserved and should not be used. note: the rx data fifo is considered full 4 dwords before the length that is specified in the hw_cfg register. table 5.3 valid tx/rx fifo allocations tx_fif_sz tx data fifo size (bytes) tx status fifo size (bytes) rx data fifo size (bytes) rx status fifo size (bytes) 2 1536 512 13440 896 3 2560 512 12480 832 4 3584 512 11520 768 5 4608 512 10560 704 6 5632 512 9600 640 7 6656 512 8640 576 8 7680 512 7680 512 9 8704 512 6720 448 10 9728 512 5760 384 11 10752 512 4800 320 12 11776 512 3840 256 13 12800 512 2880 192 14 13824 512 1920 128 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 91 revision 2.9 (03-01-12) datasheet in addition to the host-accessible fifos, the mac interface layer (mil) contains an additional 2k bytes of tx, and 128 bytes of rx fifo buffering. th ese sizes are fixed, and cannot be adjusted by the host. as space in the tx mil (mac interface layer) fifo frees, data is moved into it from the tx data fifo. depending on the size of the frames to be transmitt ed, the mil can hold up to two ethernet frames. this is in addition to any tx data that may be queued in the tx data fifo. conversely, as data is received by the lan9221/lan9 221i, it is moved from the mac to the rx mil fifo, and then into the rx data fifo. when the rx data fifo fills up, data will continue to collect in the rx mil fifo. if the rx mil fifo fills up and overruns, subseque nt rx frames will be lost until room is made in the rx data fifo. for each frame of data that is lost, the rx dropped frames counter (rx_drop) is incremented. rx and tx mil fifo levels are not visible to the host processor. rx and tx mil fifos operate independent of the tx data and rx data and status fifos. fifo levels set for the rx and tx data and status fifos do not take into consideration the mil fifos. 5.3.10 rx_dp_ctrlreceive datapath control register this register is used to discard unwanted receive frames. offset: 78h size: 32 bits bits description type default 31 rx data fifo fast forward (rx_ffwd): writing a 1 to this bit causes the rx data fifo to fast-forward to t he start of the next frame. this bit will remain high until the rx data fifo fast-forward operation has completed. no reads should be issued to the rx data fifo while this bit is high. note: please refer to sect ion receive data fifo fast forward on page 63 for detailed information regarding the use of rx_ffwd. r/w sc 0b 30-0 reserved ro - downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 92 smsc lan9221/lan9221i datasheet 5.3.11 rx_fifo_infreceive fifo information register this register contains the used space in the receive fifos of the lan9221/lan9221i ethernet controller. 5.3.12 tx_fifo_inftransmit fifo information register this register contains the free space in the tr ansmit data fifo and the used space in the transmit status fifo in the lan9221/lan9221i. offset: 7ch size: 32 bits bits description type default 31-24 reserved ro - 23-16 rx status fifo used space (rxsused). indicates the amount of space in dwords, used in the rx status fifo. ro 00h 15-0 rx data fifo used space (rxdused). ). reads the amount of space in bytes, used in the rx data fifo. for each receive frame, this field is incremented by the length of the receive data rounded up to the nearest dword (if the payload does not end on a dword boundary). ro 0000h offset: 80h size: 32 bits bits description type default 31-24 reserved ro - 23-16 tx status fifo used space (txsused). indicates the amount of space in dwords used in the tx status fifo. ro 00h 15-0 tx data fifo free space (tdfree). reads the amount of space in bytes, available in the tx data fifo. the appl ication should never write more data than is available, as indicated by this value. ro 1200h downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 93 revision 2.9 (03-01-12) datasheet 5.3.13 pmt_ctrl power mana gement control register this register controls the power management features. this register can be read while the lan9221/lan9221i is in a power saving mode. note: the lan9221/lan9221i must always be read at least once after power-up, reset, or upon return from a power-saving state or write operations will not function. offset: 84h size: 32 bits bits description type default 31:14 reserved ro - 13-12 power management mode (pm_mode) C these bits set the lan9221/lan9221i into the appropriat e power management mode. special care must be taken when modifying these bits. encoding : 00b C d0 (normal operation) 01b C d1 (wake-up frame and magic packet detection are enabled) 10b C d2 (can perform energy detect) 11b C reserved - do not set in this mode note: when the lan9221/lan9221i is in any of the reduced power modes, a write of any data to t he byte_test register will wake- up the device. do not perform writes to other addrresses while the ready bit in this register is cleared. sc 00b 11 reserved ro - 10 phy reset (phy_rst) C writing a 1 to this bit re sets the phy. the internal logic automatically holds the phy rese t for a minimum of 100us. when the phy is released from reset, th is bit is automatically cleared. all writes to this bit are ignored while this bit is high. sc 0b 9 wake-on-lan enable (wol_en) C when set, the pme signal (if enabled with pme_en) will be asserted in accordance with the pme_ind bit upon a wol event. when set, the pme_int will also be asserted upon a wol event, regardless of the setting of the pme_en bit. r/w 0b 8 energy-detect enable (ed_en) - when set, the pme signal (if enabled with pme_en) will be asserted in accordance with the pme_ind bit upon an energy-detect event. when set, the pme_int will also be asserted upon an energy detect event, regardless of the setting of the pme_en bit. r/w 0b 7 reserved ro - 6 pme buffer type (pme_type) C when cleared, enables pme to function as an open-drain buffer for use in a wi red-or configurati on. when set, the pme output is a push-pull driver. when configured as an open-drain output the pme_pol field is ignored, and the output is always active low. r/w nasr 0b downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 94 smsc lan9221/lan9221i datasheet 5-4 wake-up status (wups) C this field indicates the cause of a wake-up event detection as follows 00b -- no wake-up event detected 01b -- energy detected 10b -- wake-up frame or magic packet detected 11b -- indicates multiple events occurred wups bits are cleared by writing a 1 to the appropriate bit. the device must return to the d0 state (ready bit set) before these bits can be cleared. note: in order to clear this bit, it is required that all event sources be cleared as well. the event sources are described in figure 3.18 pme and pme_int signal generationon page 46 . r/wc 00 3 pme indication (pme_ind). the pme signal can be configured as a pulsed output or a static signal, which is asserted upon detection of a wake-up event. when set, the pme signal will pulse acti ve for 50ms upon detection of a wake-up event. when clear, the pme signal is driven continuously upon detection of a wake- up event. the pme signal can be deactivated by clearing the wups bits, or by clearing the appropriate enable (refer to section 3.10.2.3, "power management event indicators," on page 46 ). r/w 0b 2 pme polarity (pme_pol). this bit controls the pola rity of the pme signal. when set, the pme output is an active high signal. when reset, it is active low. when pme is configured as an o pen-drain output this field is ignored, and the output is always active low. r/w nasr 0b 1 pme enable (pme_en). when set, this bit enables the external pme signal. this bit does not affect the pme interrupt (pme_int). r/w 0b 0 device ready (ready). when set, this bit indicates that lan9221/lan9221i is ready to be accessed. this register can be read when lan9221/lan9221i is in any power management mode. upon waking from any power management mode, including power-up, the host processor can interrogate this field as an indication when lan9221/lan9221i has stabilized and is fully alive. reads and writes of any other address are invalid until this bit is set. note: with the exception of hw_cfg and pmt_ctrl, read access to any internal resources is forbidden while the ready bit is cleared. note: on power-up, this bit can be polled to indicate when a valid soft reset (srst) can be performed. ro - bits description type default downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 95 revision 2.9 (03-01-12) datasheet 5.3.14 gpio_cfggeneral purpos e io configuration register this register configures the gpio and led functions. offset: 88h size: 32 bits bits description type default 31 reserved ro - 30:28 led[3:1] enable (ledx_en). a 1 sets the associated pin as an led output. when cleared low, the pin functions as a gpio signal. led1/gpio0 C bit 28 led2/gpio1 C bit 29 led3/gpio2 C bit 30 r/w 000 27 reserved ro - 26:24 gpio interrupt polarity 0-2 (gpio_int_pol). when set high, a high logic level on the corresponding gpio pin will set the corresponding int_sts register bit. when cleared low, a lo w logic level on the corresponding gpio pin will set the corresponding int_sts register bit. gpio interrupts must also be enabled in gpiox_int_en in the int_en register. gpio0 C bit 24 gpio1 C bit 25 gpio2 C bit 26 note: gpio inputs must be active for gr eater than 40ns to be recognized as interrupt inputs. r/w 000 23 reserved ro - 22:20 eeprom enable (eepr_en). the value of this field determines the function of the external eedio and eeclk: please refer to ta b l e 5 . 4 for the eeprom enable bit function definitions. note: the host must not change the f unction of the eedio and eeclk pins when an eeprom read or writ e cycle is in progress. do not use reserved settings. r/w 000 19 reserved ro - 18:16 gpio buffer type 0-2 (gpiobufn). when set, the output buffer for the corresponding gpio signal is conf igured as a push/pull driver. when cleared, the corresponding gpio set co nfigured as an open-drain driver. gpio0 C bit 16 gpio1 C bit 17 gpio2 C bit 18 r/w 000 15:11 reserved ro - 10:8 gpio direction 0-2 (gpdirn). when set, enables the corresponding gpio as output. when cleared the gpio is enabled as an input. gpio0 C bit 8 gpio1 C bit 9 gpio2 C bit 10 r/w 0000 7:5 reserved ro - downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 96 smsc lan9221/lan9221i datasheet 5.3.15 gpt_cfg-general purpose timer configuration register this register configures the general purpose time r. the gp timer can be configured to generate host interrupts at intervals defined in this register. 4:3 gpo data 3-4 (gpodn). the value written is reflected on gpon. gpo3 C bit 3 gpo4 C bit 4 r/w 00 2:0 gpio data 0-2 (gpiodn) . when enabled as an output , the value written is reflected on gpion. when read, gpion reflects the current state of the corresponding gpio pin. gpio0 C bit 0 gpio1 C bit 1 gpio2 C bit 2 r/w 000 table 5.4 eeprom enab le bit definitions [22] [21] [20] eedio f unction eeclk function 0 0 0 eedio eeclk 0 0 1 gpo3 gpo4 0 1 0 reserved 0 1 1 gpo3 rx_dv 1 0 0 reserved 1 0 1 tx_en gpo4 1 1 0 tx_en rx_dv 11 1 t x _ c l k r x _ c l k offset: 8ch size: 32 bits bits description type default 31-30 reserved ro - 29 gp timer enable (timer_en). when a one is written to this bit the gp timer is put into the run state. when cleared, the gp timer is halted. on the 1 to 0 transition of this bit the gpt_load field will be preset to ffffh. r/w 0 28-16 reserved ro - 15-0 general purpose timer pre-load (gpt_load). this value is pre-loaded into the gp-timer. r/w ffffh bits description type default downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 97 revision 2.9 (03-01-12) datasheet 5.3.16 gpt_cnt-general purpose timer current count register this register reflects the cu rrent value of the gp timer. 5.3.17 word_swapword swap control this register controls how words from the host data bus are mapped to the csrs and data fifos inside the lan9221/lan9221i. the lan9221/lan9221i always sends data from the transmit dat a fifo to the network so that the lo w order word is sent first, and always receives data from the network to the receive data fifo so that the low order word is received first. offset: 90h size: 32 bits bits description type default 31-16 reserved ro - 15-0 general purpose timer cu rrent count (gpt_cnt). this 16-bit field reflects the current value of the gp timer. ro ffffh offset: 98h size: 32 bits bits description type default 31:0 word swap. if this field is set to 00000000h, or anything except 0xffffffffh, the lan9221/lan9221i maps words with address bit a[1]=1 to the high order words of the cs rs and data fifos, and words with address bit a[1]=0 to the low order wo rds of the csrs and data fifos. if this field is set to 0xffffffffh, the lan9221/lan9221i maps words with address bit a[1]=1 to the low order words of the csrs and data fifos, and words with address bit a[1]=0 to the hi gh order words of the csrs and data fifos. note: word swap is used in conjunction with the mixed endian functionality to determine the final byte ordering. refer to section 3.7.3, "mixed endian support" for more information. r/w nasr 00000000h downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 98 smsc lan9221/lan9221i datasheet 5.3.18 free_runfree-run 25mhz counter this register reflects the value of the free-running 25mhz counter. 5.3.19 rx_dropC receiver dropped frames counter this register indicates the number of receive frames that have been dropped. offset: 9ch size: 32 bits bits description type default 31:0 free running sclk counter (fr_cnt): note: this field reflects the value of a free-running 32-bit counter. at reset the counter starts at zero and is incremented for every 25mhz cycle. when the maximum count has been reached the counter will rollover. since the bus interface is 16-bits wide, and this is a 32-bit counter, the count value is latched on the first read. the free_run counter can take up to 160ns to clear after a reset event. note: this counter will run regardless of the power management states d0, d1 or d2. ro - offset: a0h size: 32 bits bits description type default 31-0 rx dropped frame counter (rx_dfc). this counter is incremented every time a receive frame is dropped. rx_d fc is cleared on any read of this register. an interrupt can be issued when this counter passes through its halfway point (7fffffffh to 80000000h). rc 00000000h downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 99 revision 2.9 (03-01-12) datasheet 5.3.20 mac_csr_cmd C mac csr synchronizer command register this register is used to control the read and write operatio ns with the mac csrs 5.3.21 mac_csr_data C mac csr synchronizer data register this register is used in conjunction with th e mac_csr_cmd register to perform read and write operations with the mac csrs offset: a4h size: 32 bits bits description type default 31 csr busy. when a 1 is written into this bit, the read or write operation is performed to the specified mac csr. this bit will remain set until the operation is complete. in the case of a read this means that the host can read valid data from the data register. the mac_csr_cmd and mac_csr_data registers should not be modified until this bit is cleared. sc 0 30 r/nw. when set, this bit indicates t hat the host is requesting a read operation. when clear, the host is performing a write. r/w 0 29-8 reserved. ro - 7-0 csr address. the 8-bit value in this field selects which mac csr will be accessed with the read or write operation. r/w 00h offset: a8h size: 32 bits bits description type default 31-0 mac csr data. value read from or written to the mac csrs. r/w 00000000h downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 100 smsc lan9221/lan9221i datasheet 5.3.22 afc_cfg C automatic flow control configuration register this register configures the mechanism that controls both the automatic, and software-initiated transmission of pause frames and back pressure. note: the lan9221/lan9221i will not transmit paus e frames or assert back pressure if the transmitter is disabled. offset: ach size: 32 bits bits description type default 31:24 reserved ro - 23:16 automatic flow control high level (afc_hi). specifies, in multiples of 64 bytes, the level at which flow control will trigger. when this limit is reached the chip will apply back pressure or will transmit a pause frame as programmed in bits [3:0] of this register. during full-duplex operation only a si ngle pause frame is transmitted when this level is reached. the pause ti me transmitted in this frame is programmed in the fcpt field of the flow register in the mac csr space. during half-duplex operation each incoming frame that matches the criteria in bits [3:0] of this register will be jammed for the period set in the back_dur field. r/w 00h 15:8 automatic flow control low level (afc_lo). specifies, in multiples of 64 bytes, the level at which a pause fr ame is transmitted with a pause time setting of zero. when the amount of data in the rx data fifo falls below this level the pause frame is transm itted. a pause time value of zero instructs the other transmitting device to immediately resume transmission. the zero time pause frame will only be transmitted if the rx data fifo had reached the afc_hi level and a pause frame was sent. a zero pause time frame is sent whenever automatic flow c ontrol in enabled in bits [3:0] of this register. note: when automatic flow control is enabled the afc_lo setting must always be less than the afc_hi setting. r/w 00h 7:4 backpressure duration (back_dur). when the lan9221/lan9221i automatically asserts back pressure, it will be asserted for this period of time. this field has no function and is not used in full-duplex mode. please refer to table 5.5 , describing backpressure duration bit mapping for more information. r/w 0h 3 flow control on multicast frame (fcmult). when this bit is set, the lan9221/lan9221i will assert back pressure when the afc level is reached and a multicast frame is received. this field has no function in full- duplex mode. r/w 0 2 flow control on broadcast frame (fcbrd). when this bit is set, the lan9221/lan9221i will assert back pressure when the afc level is reached and a broadcast frame is received. this field has no function in full- duplex mode. r/w 0 1 flow control on address decode (fcadd). when this bit is set, the lan9221/lan9221i will assert back pressure when the afc level is reached and a frame addressed to the lan9221/lan9221i is received. this field has no function in full-duplex mode. r/w 0 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 101 revision 2.9 (03-01-12) datasheet 0 flow control on any frame (fcany). when this bit is set, the lan9221/lan9221i will assert back pr essure, or transmit a pause frame when the afc level is reached and any frame is received. setting this bit enables full-duplex flow control when the lan9221/lan9221i is operating in full-duplex mode. when this mode is enabled during half-duplex operation, the flow controller does not decode the mac address and will send a pause frame upon receipt of a valid preamble (i.e., immediately at the beginning of the next frame after the rx data fifo level is reached). when this mode is enabled during full- duplex operation, the flow controller will immediately instruct the mac to send a pause frame when the rx data fifo level is reached. the mac will queue the pause frame transmission for the next available window. setting this bit overrides bits [3:1] of this register. r/w 0 table 5.5 backpressure duration bit mapping backpressure duration [19:16] 100mbs mode 10mbs mode 0h 5us 7.2us 1h 10us 12.2us 2h 15us 17.2us 3h 25us 27.2us 4h 50us 52.2us 5h 100us 102.2us 6h 150us 152.2us 7h 200us 202.2us 8h 250us 252.2us 9h 300us 302.2us ah 350us 352.2us bh 400us 402.2us ch 450us 452.2us dh 500us 502.2us eh 550us 552.2us fh 600us 602.2us bits description type default downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 102 smsc lan9221/lan9221i datasheet 5.3.23 e2p_cmd C eepr om command register this register is used to control the read and write operations with the serial eeprom. offset: b0h size: 32 bits bits description type default 31 epc busy: when a 1 is written into this bi t, the operation specified in the epc command field is performed at the specified eeprom address. this bit will remain set until the operation is complete. in the case of a read this means that the host can read valid data from the e2p data register. the e2p_cmd and e2p_data registers should not be modified until this bit is cleared. in the case where a write is attempted and an eeprom is not present, the epc busy remains busy until the epc time-out occurs. at that time the busy bit is cleared. note: epc busy will be high immediately following power-up or reset. after the eeprom controller has finished reading (or attempting to read) the mac address from the eeprom the epc busy bit is cleared. sc 0 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 103 revision 2.9 (03-01-12) datasheet 30-28 epc command. this field is used to issue commands to the eeprom controller. the epc will execute commands when the epc busy bit is set. a new command must not be issued un til the previous command completes. this field is encoded as follows: read (read location): this command will cause a read of the eeprom location pointed to by epc address. the result of the read is available in the e2p_data register. ewds (erase/write disable): after issued, the eeprom will ignore erase and write commands. to re-enable er ase/write operations issue the ewen command. ewen (erase/write enable): enables the eeprom for erase and write operations. the eeprom will allow eras e and write operations until the erase/write disable command is sent, or until power is cycled. note: the eeprom device will power-up in the erase/write-disabled state. any erase or write operati ons will fail until an erase/write enable command is issued. write (write location): if erase/write operations are enabled in the eeprom, this command will cause the contents of the e2p_data register to be written to the eeprom location selected by the epc address field. wral (write all): if erase/write operations are enabled in the eeprom, this command will cause the contents of the e2p_data register to be written to every eeprom memory location. erase (erase location): if erase/write operatio ns are enabled in the eeprom, this command will erase the location selected by the epc address field. eral (erase all): if erase/write operations are enabled in the eeprom, this command will initiate a bulk erase of the entire eeprom. reload (mac address reload): instructs the eeprom controller to reload the mac address from the eeprom. if a value of 0x a5 is not found in the first address of the eeprom, the eeprom is assumed to be un- programmed and mac address relo ad operation will fail. the mac address loaded bit indicates a successful load of the mac address. r/w 0 27-10 reserved. ro - bits description type default [30] [29] [28] operation 0 0 0 read 001 e w d s 0 1 0 ewen 011 w r i t e 100 w r a l 1 0 1 erase 1 1 0 eral 111 r e l o a d downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 104 smsc lan9221/lan9221i datasheet 5.3.24 e2p_data C eeprom data register this register is used in conjunction with the e2p_cmd register to perform read and write operations with the serial eeprom. 9 epc time-out. if an eeprom operation is per formed, and there is no response from the eeprom within 30ms, the eeprom controller will time- out and return to its idle state. this bit is set when a time-out occurs indicating that the last operation was unsuccessful. note: if the eedio signal pin is externally pulled-high, epc commands will not time out if the eeprom devic e is missing. in this case the epc busy bit will be cleared as soon as the command sequence is complete. it should also be noted that th e erase, eral, write and wral commands are the only epc commands that will time-out if an eeprom device is not present - and- the eedio signal is pulled low r/wc 0 8 mac address loaded. when set, this bit indicates that a valid eeprom was found, and that the mac ad dress programming has completed normally. this bit is set after a successful load of the mac address after power-up, or after a reload command has completed r/wc - 7-0 epc address. the 8-bit value in this field is used by the eeprom controller to address the specific me mory location in the serial eeprom. this is a byte aligned address. r/w 00h offset: b4h size: 32 bits bits description type default 31-8 reserved ro - 7:0 eeprom data. value read from or written to the eeprom. r/w 00h bits description type default downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 105 revision 2.9 (03-01-12) datasheet 5.4 mac control and status registers these registers are located in the mac module and are accessed indirectly through the mac-csr synchronizer port. table 5.6, "mac csr register map" , shown below, lists the mac registers that are accessible through the indexing method using the mac_csr_cmd and mac_csr_data registers (see sections mac_csr_cmd C mac csr synchr onizer command register and mac_csr_data C mac csr synchronizer data register). table 5.6 mac csr register map mac control and status registers index symbol register name default 1m a c _ c r mac control register 00040000h 2 addrh mac address high 0000ffffh 3 addrl mac address low ffffffffh 4 hashh multicast hash table high 00000000h 5 hashl multicast hash table low 00000000h 6m i i _ a c c mii access 00000000h 7 mii_data mii data 00000000h 8f l o w flow control 00000000h 9v l a n 1 vlan1 tag 00000000h av l a n 2 vlan2 tag 00000000h bw u f f wake-up frame filter 00000000h cw u c s r wake-up control and status 00000000h dc o e _ c r checksum offload engine control 00000000h downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 106 smsc lan9221/lan9221i datasheet 5.4.1 mac_crmac control register this register establishes the rx and tx operation modes and controls for address filtering and packet filtering. offset: 1 attribute: r/w default value: 00040000h size: 32 bits bits description 31 receive all mode (rxall). when set, all incoming packets will be received and passed on to the address filtering function for processing of the sele cted filtering mode on the received frame. address filtering then occurs and is reported in receive stat us. when reset, only frames that pass destination address filtering will be sent to the application. 30-24 reserved 23 disable receive own (rcvown). when set, the mac disables the reception of frames when the mii tx_en signal is asserted. the mac blocks the tr ansmitted frame on the receive path. when reset, the mac receives all packets the phy gives, including those transmitted by the mac.this bit should be reset when the full duplex mode bit is set. 22 reserved 21 loopback operation mode (loopbk). selects the loop back operation modes for the mac. this is only for full duplex mode 1b0: normal: no feedback 1b1: internal: through mii in internal loopback mode, the tx frame is received by the internal mii interface, and sent back to the mac without being sent to the phy. note: when enabling or disabling the loopback mode it can take up to 10 s for the mode change to occur. the transmitter and receiver must be stopped and disabled when modifying the loopbk bit. the transmit ter or receiver should not be enabled within10 s of modifying the loopbk bit. 20 full duplex mode (fdpx). when set, the mac operates in full-du plex mode, in which it can transmit and receive simultaneously. in full-duplex mode, the heartbeat check is disa bled and the heartbeat fail status should thus be ignored. 19 pass all multicast (mcpas). when set, indicates that all incoming frames with a multicast destination address (first bit in the destination address field is 1) are received. incoming frames with physical address (individual address/unicast) destinations are filtered and received only if the address matches the mac address. 18 promiscuous mode (prms). when set, indicates that any incoming frame is received regardless of its destination address. 17 inverse filtering (invfilt). when set, the address check function operates in inverse filtering mode. this is valid only during perfect filtering mode. 16 pass bad frames (passbad). when set, all incoming frames that passed address filtering are received, including runt frames and collided frames. 15 hash only filtering mode (ho). when set, the address check functi on operates in the imperfect address filtering mode both fo r physical and multicast addresses 14 reserved downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 107 revision 2.9 (03-01-12) datasheet 13 hash/perfect filtering mode (hpfilt). when reset (0), the lan9221/lan9221i will implement a perfect address filter on incoming frames accordi ng the address specified in the mac address register. when set (1), the address check function does im perfect address filtering of multicast incoming frames according to the hash table specified in the multicast hash table register. if the hash only filtering mode (ho) bit is set (1), then the physical (ia) are im perfect filtered too. if the hash only filtering mode (ho) bit is reset (0), then the ia addresses are perfect address filtered according to the mac address register 12 late collision control (lcoll). when set, enables retransmission of the collided frame even after the collision period (late collision). when reset, the mac disables frame transmission on a late collision. in any case, the late collision status is appropriately updated in the transmit packet status. 11 disable broadcast frames (bcast). when set, disables the rec eption of broadcast frames. when reset, forwards all broadcast frames to the application. note: when wake-up frame detection is en abled via the wuen bit of the wucsrwake-up control and status register , a broadcast wake-up frame will wake-up the device despite the state of this bit. 10 disable retry (disrty). when set, the mac attempts only one transmission. when a collision is seen on the bus, the mac ignores the current frame and goes to the next frame and a retry error is reported in the transmit status. when reset, the mac attempts 16 transmissions before signaling a retry error. 9 reserved 8 automatic pad stripping (padstr). when set, the mac strips the pad field on all incoming frames, if the length field is less than 46 bytes. the fcs fi eld is also stripped, since it is computed at the transmitting station based on the data and pad fiel d characters, and is invalid for a received frame that has had the pad characters stripped. receive frames with a 46-byte or greater length field are passed to the application unmodified (fcs is not st ripped). when reset, the mac passes all incoming frames to the host unmodified. note: when padstr is enabled, the rx checksum offload engine must be disabled (bit 0 (rxcoe_en) of the coe_crchecksum offload en gine control register ) and vice versa. these functions cannot be enabled simultaneously. bits description downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 108 smsc lan9221/lan9221i datasheet 7-6 backoff limit (bolmt). the bolmt bits allow the user to set its back-off limit in a relaxed or aggressive mode. accordi ng to ieee 802.3, the mac has to wait for a random number [r] of slot- times** after it detects a collision, where: (eq.1)0 < r < 2 k the exponent k is dependent on how many times the cu rrent frame to be transmitted has been retried, as follows: (eq.2)k = min (n, 10) where n is the current number of retries. if a frame has been retried three times, then k = 3 an d r= 8 slot-times maximum. if it has been retried 12 times, then k = 10, and r = 1024 slot-times maximum. an lfsr (linear feedback shift register) 20-bit co unter emulates a 20bit random number generator, from which r is obtained. once a collision is detect ed, the number of the curr ent retry of the current frame is used to obtain k (eq.2). this value of k tr anslates into the number of bits to use from the lfsr counter. if the value of k is 3, the mac takes the value in the first three bits of the lfsr counter and uses it to count down to zero on every slot-tim e. this effectively causes the mac to wait eight slot-times. to give the user more flexibility, the bolmt value forces the number of bits to be used from the lfsr counter to a predetermined value as in the table below. thus, if the value of k = 10, the mac will look at the bol mt if it is 00, then use the lower ten bits of the lfsr counter for the wait countdown. if the bolmt is 10, then it will only use t he value in the first four bits for the wait countdown, etc. **slot-time = 512 bit times. (see ieee 802 .3 spec., secs. 4. 2.3.25 and 4.4.2.1) 5 deferral check (dfchk). when set, enables the deferral check in the mac. the mac will abort the transmission attempt if it has def erred for more than 24,288 bit times. deferral starts when the transmitter is ready to transmit, but is prevented fr om doing so because the crs is active. defer time is not cumulative. if the transmitter defers for 10,000 bit times, then transmit s, collides, backs off, and then has to defer again after completion of back-off, the deferral timer resets to 0 and restarts. when reset, the deferral check is disabled in the mac and the mac defers indefinitely. 4 reserved 3 transmitter enable (txen). when set, the macs transmitter is enabled and it will transmit frames from the buffer onto the cable. when reset, the macs transmitter is disabled and will not transmit any frames. 2 receiver enable (rxen). when set (1), the macs receiver is enabled and will receive frames from the internal phy. when reset, the macs receiver is disabled and will not receive any frames from the internal phy. 1-0 reserved bits description bolmt value # bits used from lfsr counter 2b00 10 2b01 8 2b10 4 2b11 1 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 109 revision 2.9 (03-01-12) datasheet 5.4.2 addrhmac a ddress high register the mac address high register contains the upper 16-bits of the physical address of the mac. the contents of this register ar e optionally loaded from the eeprom at power-on through the eeprom controller if a programmed eeprom is detected. the least si gnificant byte of this register (bits [7:0]) is loaded from address 0x05 of the eeprom. the seco nd byte (bits [15:8]) is loaded from address 0x06 of the eeprom. please refer to section 4.6 for more information on the eeprom. section 5.4.3 details the byte ordering of the addrl and addrh registers with respect to the reception of the ethernet physical address. offset: 2 attribute: r/w default value: 0000ffffh size: 32 bits bits description 31-16 reserved 15-0 physical address [47:32]. this field contains the upper 16-bit s (47:32) of the physical address of the lan9221/lan92 21i device. the content of this field is undefined unt il loaded from the eeprom at power-on. the host can update the contents of this field after the initialization process has completed. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 110 smsc lan9221/lan9221i datasheet 5.4.3 addrlmac address low register the mac address low register contains the lower 32 bits of the physical address of the mac. the contents of this register ar e optionally loaded from the eeprom at power-on through the eeprom controller if a programmed eeprom is detected. the least si gnificant byte of this register (bits [7:0]) is loaded from address 0x01 of the eeprom. the most significant byte of this register is loaded from address 0x04 of the eeprom. please refer to section 4.6 for more information on the eeprom. table 5.7 below illustrates the byte ordering of the addrl and addrh registers with respect to the reception of the ethernet physical address. also shown is the correlation between the eeprom addresses and addrl and addrh registers. as an example, if the desired ethernet physical address is 12-34-56-78-9a-bc, the addrl and addrh registers would be programmed as shown in figure 5.2 . the values required to automatically load this configuration from the eeprom are also shown. figure 5.2 example addrl, addrh and eeprom setup offset: 3 attribute: r/w default value: ffffffffh size: 32 bits bits description 31-0 physical address [31:0]. this field contains the lower 32 bits (31:0) of the physical address of the lan9221/lan9221i device. the c ontent of this field is undefined until loaded from the eeprom at power-on. the host can update the contents of this field after the initialization process has completed. table 5.7 addrl, addrh and eeprom byte ordering eeprom address addrn order of reception on ethernet 0x01 addrl[7:0] 1 st 0x02 addrl[15:8] 2 nd 0x03 addrl[23:16] 3 rd 0x04 addrl[31:24] 4 th 0x05 addrh[7:0] 5 th 0x06 addrh[15:8] 6 th 0x12 0 7 0x34 8 15 0x56 16 23 0x78 24 31 addrl 0x9a 0 7 0xbc 8 15 addrh xx 16 23 xx 24 31 0xa5 0x12 0x34 0x56 0x78 0x9a 0xbc 0x00 0x01 0x02 0x03 0x04 0x05 0x06 eeprom downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 111 revision 2.9 (03-01-12) datasheet note: by convention, the left most by te of the ethernet address (in this example 0x12) is the most significant byte and is transmitted/received first. 5.4.4 hashhmulticast h ash table high register the 64-bit multicast table is used for group address fi ltering. for hash filtering, the contents of the destination address in the incoming frame is used to index the contents of the hash table. the most significant bit determines the register to be used (h i/low), while the other five bits determine the bit within the register. a value of 00000 selects bit 0 of the multicast hash table lo register and a value of 11111 selects the bit 31 of the multicast hash table hi register. if the corresponding bit is 1, then the multicast frame is accepted. otherwise, it is rejected. if the pass all multicast (mcpas) bit is set (1), then all multicast frames are accepted regardless of the multicast hash values. the multicast hash table hi register contains the higher 32 bits of the hash table and the multicast hash table low register contains the lower 32 bits of the hash table. 5.4.5 hashlmulticast hash table low register this register defines the lower 32-bits of the multicast hash table. please refer to table 5.4.4, "hashhmulticast hash table high register" for further details. offset: 4 attribute: r/w default value: 00000000h size: 32 bits bits description 31-0 upper 32 bits of the 64-bit hash table offset: 5 attribute: r/w default value: 00000000h size: 32 bits bits description 31-0 lower 32 bits of the 64-bit hash table downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 112 smsc lan9221/lan9221i datasheet 5.4.6 mii_accmii access register this register is used to contro l the management cycles to the phy. 5.4.7 mii_datamii data register this register contains either the data to be written to the phy register specified in the mii access register, or the read data from the phy register whose index is specified in the mii access register. offset: 6 attribute: r/w default value: 00000000h size: 32 bits bits description 31-16 reserved 15-11 phy address: for every access to this register, this field must be set to 00001b. 10-6 mii register index (miirinda): these bits select the desired mii register in the phy. 5-2 reserved 1 mii write (miiwnr): setting this bit tells the phy that this will be a write operation using the mii data register. if this bit is not set, this will be a read operation, packing the data in the mii data register. 0 mii busy (miibzy): this bit must be polled to determine when the mii register access is complete. this bit must read a logical 0 before writing to this register and mii data register. the lan driver software must set (1) this bit in order for the lan9221/lan9221i to read or write any of the mii phy registers. during a mii register acce ss, this bit will be set, signifying a read or write access is in progress. the mii data register must be kept valid until the mac clears this bit during a phy write operation. the mii data register is invalid until the mac has cleared this bit during a phy read operation. offset: 7 attribute: r/w default value: 00000000h size: 32 bits bits description 31-16 reserved 15-0 mii data. this contains the 16-bit value read from the phy read operation or the 16-bit data value to be written to the phy before an mii write operation. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 113 revision 2.9 (03-01-12) datasheet 5.4.8 flowflow control register this register controls the generation and receptio n of the control (pause command) frames by the macs flow control block. the control frame fields are selected as specified in the 802.3x specification and the pause-time value from this register is used in the pause time field of the control frame. in full-duplex mode the fcbsy bi t is set until the control frame is tr ansferred onto the cable. in half- duplex mode fcbsy is set while back pressure is being asserted. the ho st has to make sure that the busy bit is cleared before writin g the register. the pass control frame bit (fcpass) does not affect the sending of the frames, including control frames, to the application interface. the flow control enable (fcen) bit enables the receive portion of the flow control block. this register is used in conjunction with the afc_ cfg register in the slave csrs to configure flow control. software flow control is initiated using the afc_cfg register. note: the lan9221/lan9221i will not transmit paus e frames or assert back pressure if the transmitter is disabled. offset: 8 attribute: r/w default value: 00000000h size: 32 bits bits description 31-16 pause time (fcpt). this field indicates the value to be used in the pause time field in the control frame. this field must be initialized before full-duplex automatic flow control is enabled. 15-3 reserved 2 pass control frames (fcpass). when set, the mac will pass the pause frame to the host. the application must accept or discard a received fram e based on the packet filter control bit. the mac receives, decodes and performs the pause functi on when a valid pause frame is received in full- duplex mode and when flow control is enabled (fce bit set). when reset, the mac resets the packet filter bit in the receive packet status. the mac always passes the data of all frames it receives (including flow control frames) to the application. frames that do not pass address filtering, as well as fr ames with errors, are passed to the application. the application must discard or retain the received frames data based on the received frames status field. filtering modes (promiscuous mode, for example) take precedence over the fcpass bit. 1 flow control enable (fcen). when set, enables the mac flow control function. the mac decodes all incoming frames for control frames; if it re ceives a valid control frame (pause command), it disables the transmitter for a specified time (decoded pause time x slot time). when reset, the mac flow control function is disabled; the mac does not decode frames for control frames. note: flow control is applicable when the mac is se t in full duplex mode. in half-duplex mode, this bit enables the backpressure function to cont rol the flow of received frames to the mac. 0 flow control busy (fcbsy). this bit is set high whenever a pause frame or back pressure is being transmitted. this bit should read logical 0 before writing to the flow control (flow) register. during a transfer of control frame, this bit continues to be set, signifying that a frame transmission is in progress. after the pause control frames transm ission is complete, th e mac resets to 0. notes: ? when writing this register th e fcbsy bit must always be zero. ? applications must always write a zero to this bit downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 114 smsc lan9221/lan9221i datasheet 5.4.9 vlan1vlan1 tag register this register contains the vlan tag field to iden tify vlan1 frames. for vlan frames the legal frame length is increased from 1518 bytes to 1522 bytes. 5.4.10 vlan2vlan2 tag register this register contains the vlan tag field to iden tify vlan2 frames. for vlan frames the legal frame length is increased from 1518 bytes to 1522 bytes. offset: 9 attribute: r/w default value: 00000000h size: 32 bits bits description 31-16 reserved 15-0 vlan1 tag identifier (vti1). this contains the vlan tag field to identify the vlan1 frames. this field is compared with the 13th and 14th bytes of the incoming frames for vlan1 frame detection. if used, this register must be set to 0x8100. offset: a attribute: r/w default value: 00000000h size: 32 bits bits description 31-16 reserved 15-0 vlan2 tag identifier (vti2). this contains the vlan tag field to identify the vlan2 frames. this field is compared with the 13th and 14th bytes of the incoming frames for vlan2 frame detection.if used, this register must be set to 0x8100. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 115 revision 2.9 (03-01-12) datasheet 5.4.11 wuffwake-up frame filter this register is used to conf igure the wake up frame filter. 5.4.12 wucsrwake-up control and status register this register contains data pertaining to the macs remote wake-up status and capabilities. offset: b attribute: wo default value: 00000000h size: 32 bits bits description 31-0 wake-up frame filter (wff). wake-up frame filter (wff). the wa ke-up frame filter is configured through this register using an indexing mechanism. af ter hardware reset, or soft reset, the mac loads the first value written to this location to the firs t dword in the wake-up fram e filter (filter 0 byte mask). the second value written to this location is loaded to the second dword in the wake-up frame filter (filter 1 byte mask) and so on. once all eight dwords have been written, the internal pointer will once again point to the first entry and the filter entries can be modifi ed in the same manner. note: this is a write-only register. offset: c attribute: r/w default value: 00000000h size: 32 bits bits description 31-10 reserved 9 global unicast enable (gue). when set, the mac wakes up from power-saving mode on receipt of a global unicast frame. a global unicast frame has the mac address [0] bit set to 0. 8-7 reserved 6 remote wake-up frame received (wufr). the mac, upon receiving a valid remote wake-up frame, sets this bit. 5 magic packet received (mpr). the mac, upon receiving a valid magic packet, sets this bit. 4-3 reserved 2 wake-up frame enabled (wuen). when set, remote wake-up mode is enabled and the mac is capable of detecting wake-up frames as programmed in the wa ke-up frame filter. 1 magic packet enable (mpen). when set, magic packet wake-up mode is enabled. 0 reserved downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 116 smsc lan9221/lan9221i datasheet 5.4.13 coe_crchecksum offload engine control register this register controls the transmit and receive checksum offload engines. offset: d attribute: r/w default value: 00000000h size: 32 bits bits description 31-17 reserved 16 tx checksum offload engine enable (txcoe_en). this bit enables/disables the transmit coe. this bit may only be changed if the tx data path is disabled. 0: the txcoe is bypassed 1: the txcoe is enabled 15-2 reserved 1 rx checksum offload engine mode (rxcoe_mode) this register indicates whether the rxcoe will check for vlan tags or a snap header prior to beginning its checksum calculation. in its default mode, the calculation will always begin 14 bytes into the frame. the rxcoe_mode may only be changed if the ess rx path is disabled. 0: begin checksum calculation after first 14 bytes of ethernet frame 1: begin checksum calculation at start of l3 packet by adjusting for vlan tags and/or snap header. 0 rx checksum offload engine enable (rxcoe_en). this bit enables/disables the receive coe. this bit may only be changed if the rx data path is disabled. 0: the rxcoe is bypassed 1: the rxcoe is enabled note: when the rxcoe is enabled, automatic pad strippi ng must be disabled (bit 8 (padstr) of the mac_crmac control register ) and vice versa. these functions cannot be enabled simultaneously. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 117 revision 2.9 (03-01-12) datasheet 5.5 phy registers the phy registers are not memory mapped. these registers are acce ssed indirectly through the mac via the mii_acc and mii_data registers. an index must be used to access individual phy registers. phy register indexes are shown in table 5.8, "lan9221/lan9221i phy control and status register" . note: the nasr ( not affected by software reset) designation is only applicable when bit 15 of the phy basic control re gister (reset) is set. table 5.8 lan9221/lan9221i phy control and status register phy control and status registers index (in decimal) register name 0 basic control register 1 basic status register 2 phy identifier 1 3 phy identifier 2 4 auto-negotiation advertisement register 5 auto-negotiation link partner ability register 6 auto-negotiation expansion register 17 mode control/status register 18 special modes register 27 special control/status indications 29 interrupt source register 30 interrupt mask register 31 phy special control/ status register downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 118 smsc lan9221/lan9221i datasheet 5.5.1 basic control register note 5.1 the default value of this bit is dete rmined by the auto-negotiation process. index (in decimal): 0 size: 16-bits bits description type default 15 reset. 1 = software reset. bit is self-clearing. for best results, when setting this bit do not set other bits in this register. rw/sc 0 14 loopback. 1 = loopback mode, 0 = normal operation rw 0 13 speed select. 1 = 100mbps, 0 = 10mbps. ignored if auto negotiation is enabled (0.12 = 1). rw see note 5.1 12 auto-negotiation enable. 1 = enable auto-negotiate process (overrides 0.13 and 0.8) 0 = disable auto-negotiate process. rw 1 11 power down. 1 = general power down-mode, 0 = normal operation. note: after this bit is cleared, the phy may auto-negotiate with it's partner station. this process may take a few seconds to complete. once auto-negotiation is complete, bit 5 of the phy's basic status register will be set. rw 0 10 reserved ro 0 9 restart auto-negotiate. 1 = restart auto-negotiate process 0 = normal operation. bit is self-clearing. rw/sc 0 8 duplex mode. 1 = full duplex, 0 = half duplex . ignored if auto negotiation is enabled (0.12 = 1). rw see note 5.1 7 collision test. 1 = enable col test, 0 = disable col test rw 0 6-0 reserved ro 0 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 119 revision 2.9 (03-01-12) datasheet 5.5.2 basic status register 5.5.3 phy identifier 1 index (in decimal): 1 size: 16-bits bits description type default 15 100base-t4. 1 = t4 able, 0 = no t4 ability ro 0 14 100base-tx full duplex. 1 = tx with full duplex, 0 = no tx full duplex ability. ro 1 13 100base-tx half duplex. 1 = tx with half duplex, 0 = no tx half duplex ability. ro 1 12 10base-t full duplex. 1 = 10mbps with full duplex 0 = no 10mbps with full duplex ability ro 1 11 10base-t half duplex. 1 = 10mbps with half duplex 0 = no 10mbps with half duplex ability ro 1 10-6 reserved ro 0 5 auto-negotiate complete. 1 = auto-negotiate process completed 0 = auto- negotiate process not completed ro 0 4 remote fault. 1 = remote fault condition detected 0 = no remote fault ro/lh 0 3 auto-negotiate ability. 1 = able to perform auto-negotiation function 0 = unable to perform auto-negotiation function ro 1 2 link status. 1 = link is up, 0 = link is down ro/ll 0 1 jabber detect. 1 = jabber condition det ected 0 = no jabber condition detected ro/lh 0 0 extended capabilities. 1 = supports extended capabilities registers 0 = does not support extended capabilities registers. ro 1 index (in decimal): 2 size: 16-bits bits description type default 15-0 phy id number. assigned to the 3rd through 18th bits of the organizationally unique identifier (oui), respectively. ro 0x0007h downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 120 smsc lan9221/lan9221i datasheet 5.5.4 phy identifier 2 5.5.5 auto-negotiation advertisement note 5.2 when both symmetric pause and asymmetric pause support are advertised (value of 11), the device will only be configured to, at most, one of the two settings upon auto- negotiation completion. index (in decimal): 3 size: 16-bits bits description type default 15-10 phy id number assigned to the 19th through 24th bits of the oui. ro 0xc0c3h 9 - 4 model number. six-bit manufacturers model number. ro 3 - 0 revision number. four-bit manufacturers revision number. ro index (in decimal): 4 size: 16-bits bits description type default 15:14 reserved ro 00 13 remote fault. 1 = remote fault detected, 0 = no remote fault r/w 0 12 reserved r/w 0 11:10 pause operation. (see note 5.2 ) 00 no pause 01 symmetric pause 10 asymmetric pause 11 advertise support for both symmetric pause and asymmetric pause r/w 00 9 reserved ro 0 8 100base-tx full duplex. 1 = tx with full duplex, 0 = no tx full duplex ability r/w 1 7 100base-tx. 1 = tx able, 0 = no tx ability r/w 1 6 10base-t full duplex. 1 = 10mbps with full duplex 0 = no 10mbps with full duplex ability r/w 1 5 10base-t. 1 = 10mbps able, 0 = no 10mbps ability r/w 1 4:0 selector field. [00001] = ieee 802.3 r/w 00001 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 121 revision 2.9 (03-01-12) datasheet 5.5.6 auto-negotiation link partner ability index (in decimal): 5 size: 16-bits bits description type default 15 next page. 1 = next page capable, 0 = no next page ability. this device does not support next page ability. ro 0 14 acknowledge. 1 = link code word received from partner 0 = link code word not yet received note: this bit will always read 0 ro 0 13 remote fault. 1 = remote fault detected, 0 = no remote fault ro 0 12 reserved ro 0 11-10 pause operation . 00 no pause supported by partner station 01 symmetric pause supported by partner station 10 asymmetric pause supported by partner station 11 both symmetric pause and asymme tric pause supported by partner station ro 00 9 100base-t4. 1 = t4 able, 0 = no t4 ability ro 0 8 100base-tx full duplex. 1 = tx with full duplex, 0 = no tx full duplex ability ro 0 7 100base-tx. 1 = tx able, 0 = no tx ability ro 0 6 10base-t full duplex. 1 = 10mbps with full duplex 0 = no 10mbps with full duplex ability ro 0 5 10base-t. 1 = 10mbps able, 0 = no 10mbps ability ro 0 4:0 selector field. [00001] = ieee 802.3 ro 00001 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 122 smsc lan9221/lan9221i datasheet 5.5.7 auto-negotiation expansion 5.5.8 mode control/status note 5.3 the default value of this bit will vary depe ndant on the current link state of the line. index (in decimal): 6 size: 16-bits bits description type default 15:5 reserved ro 0 4 parallel detection fault. 1 = fault detected by parallel detection logic 0 = no fault detected by parallel detection logic ro/lh 0 3 link partner next page able. 1 = link partner has next page ability 0 = link partner does not have next page ability ro 0 2 next page able. 1 = local device has next page ability 0 = local device does not have next page ability ro 0 1 page received. 1 = new page received 0 = new page not yet received ro/lh 0 0 link partner auto-negotiation able. 1 = link partner has auto-negotiation ability 0 = link partner does not have auto-negotiation ability ro 0 index (in decimal): 17 size: 16-bits bits description type default 15-14 reserved. write as 0; ignore on read. rw 0 13 edpwrdown. enable the energy detect power-down mode: 0=energy detect power-down is disabled 1=energy detect power-down is enabled rw 0 12-2 reserved. write as 0, ignore on read rw 0 1 energyon. indicates whether energy is detec ted. this bit goes to a 0 if no valid energy is detected within 256ms. reset to 1 by hardware reset, unaffected by sw reset. ro see note 5.3 0 reserved. write as 0. ignore on read. rw 0 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 123 revision 2.9 (03-01-12) datasheet 5.5.9 special modes note 5.4 when mode=111, the register 0 bits 13 and 8 are variable dependant on the auto- negotiated speed and duplex. index (in decimal): 18 size: 16-bits address description type default 15-8 reserved rw, nasr 7:5 mode: phy mode of operation. refer to table 5.9 for more details. rw, nasr 111 4:0 phyad: phy address: the phy address is used for the smi address. rw, nasr 00001b table 5.9 mode control mode mode definitions default register bit values register 0 register 4 [13,12,10,8] [8,7,6,5] 000 10base-t half duplex. auto-negotiation disabled. 0000 n/a 001 10base-t full duplex. auto-negotiation disabled. 0001 n/a 010 100base-tx half duplex. auto-negotiation disabled. crs is active during transmit & receive. 1000 n/a 011 100base-tx full duplex. auto-negotiation disabled. crs is active during receive. 1001 n/a 100 100ase-tx half duplex is advertised. auto- negotiation enabled. crs is active during transmit & receive. 1100 0100 101 repeater mode. auto -negotiation enabled. 100base-tx half duplex is advertised. crs is active during receive. 1100 0100 110 reserved - do not set the lan9221/lan9221i in this mode. n/a n/a 111 all capable. auto-negotiation enabled. x10x note 5.4 1111 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 124 smsc lan9221/lan9221i datasheet 5.5.10 special control/status indications index (in decimal): 27 size: 16-bits address description mode default 15 override amdix strap 0 - amdix_en (pin 52) enables or disables hp auto mdix 1 - override pin 52. phy register 27.14 and 27.13 determine mdix function rw 0 14 auto-mdix enable: only effective when 27. 15=1, otherwise ignored. 0 = disable auto-mdix. 27.13 determines normal or reversed connection. 1 = enable auto-mdix. 27.13 must be set to 0. rw 0 13 auto-mdix state. only effective when 27. 15=1, otherwise ignored. when 27.14 = 0 (manually set mdix state): 0 = no crossover (tpo = output, tpi = input) 1 = crossover (tpo = input, tpi = output) when 27.14 = 1 (automatic mdix ) this bit must be set to 0. do not use the combination 27.15=1, 27.14=1, 27.13=1. rw 0 12:11 reserved: write as 0. ignore on read. rw 0 10 vcooff_lp: forces the receive pll 10m to lock on the reference clock at all times: 0 - receive pll 10m can lock on reference or line as needed (normal operation) 1 - receive pll 10m is locked on the reference clock. in this mode 10m data packets cannot be received. rw, nasr 0 9-5 reserved: write as 0. ignore on read. rw 0 4 xpol: polarity state of the 10base-t: 0 - normal polarity 1 - reversed polarity ro 0 3:0 reserved: read only - writing to these bits have no effect. ro xxxxb downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 125 revision 2.9 (03-01-12) datasheet 5.5.11 interrupt source flag note 5.5 the default value of this bit will vary depe ndant on the current link state of the line. 5.5.12 interrupt mask index (in decimal): 29 size: 16-bits bits description type default 15-8 reserved. ignore on read. ro/lh 0 7 int7. 1= energyon generated, 0= not source of interrupt ro/lh 0 6 int6. 1= auto-negotiation complete, 0= not source of interrupt ro/lh 0 5 int5. 1= remote fault detected, 0= not source of interrupt ro/lh 0 4 int4. 1= link down (link status negated), 0= not source of interrupt ro/lh see note 5.5 3 int3. 1= auto-negotiation lp acknowled ge, 0= not source of interrupt ro/lh 0 2 int2. 1= parallel detection fault, 0= not source of interrupt ro/lh 0 1 int1. 1= auto-negotiation page received, 0= not source of interrupt ro/lh 0 0 reserved. ro/lh 0 index (in decimal): 30 size: 16-bits bits description type default 15-8 reserved. write as 0; ignore on read. ro 0 7-0 mask bits. 1 = interrupt source is enabled 0 = interrupt source is masked rw 0 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 126 smsc lan9221/lan9221i datasheet 5.5.13 phy special control/status note 5.6 the default value of this bit is dete rmined by the auto-negotiation process. index (in decimal): 31 size: 16-bits bits description type default 15 - 13 reserved ro 000b 12 autodone. auto-negotiation done indication: 0 = auto-negotiation is not done or disabled (or not active) 1 = auto-negotiation is done ro 0b 11-5 reserved. write as 0000010b, ignore on read. rw 0000010b 4-2 speed indication. hcdspeed value: [001]=10mbps half-duplex [101]=10mbps full-duplex [010]=100base-tx half-duplex [110]=100base-tx full-duplex ro see note 5.6 1-0 reserved. write as 0; ignore on read ro 00b downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 127 revision 2.9 (03-01-12) datasheet chapter 6 timing diagrams 6.1 equivalent test load output timing specifications assume an equivalent test load as illustrated in figure 6.1 below. 6.2 host interface timing the lan9221/lan9221i supports the following host cycles: read cycles: ? pio reads (ncs or nrd controlled) ? pio burst reads (ncs or nrd controlled) ? rx data fifo direct pio re ads (ncs or nrd controlled) ? rx data fifo direct pio burst reads (ncs or nrd controlled) write cycles: ? pio writes (ncs and nwr controlled) ? tx data fifo direct pio wr ites (ncs or nwr controlled) all timing measurements were verified under the following conditions: note: the specified test load used for output timing is dependant on vddvario. temperature: ................................................................................................................... ............. note 6.1 load capacitance (vddvario = 3.3v +/- 300mv):........................................................................ 25pf load capacitance (vddvario = 2.5v +/- 10%): ............................................................................ 10pf load capacitance (vddvario = 1.8v +/- 10%): ............................................................................ 10pf note 6.1 0 o c to +70 o c for commercial version, -40 o c to +85 o c for industrial version. figure 6.1 equivalent test load output test load varies dependent on vddvario: vddvario = 3.3v 300mv: 25pf vddvario = 2.5v 10%: 10pf vddvario=1.8v 10%: 10pf downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 128 smsc lan9221/lan9221i datasheet 6.2.1 special restrictions on back-to-back write/read cycles it is important to note that there are specific re strictions on the timing of back-to-bac k write-read operations. these restrictions c oncern reading the control regist ers after any write cycle to the lan9221/lan9221i device. in many cases there is a required minimum delay between writing to the lan9221/lan9221i, and the subsequent side effect (change in the control register value). for example, when writing to the tx data fifo, it ta kes up to 135ns for the level indication to change in the tx_fifo_inf register. in order to prevent the host from reading stale data after a write operation, minimum wait periods must be enforced. these periods are specified in table 6.1, "read after write timing rules" . the host processor is required to wait t he specified period of time after any write to the lan9221/lan9221i before reading the resource specified in the table. these wait periods are for read operations that immediately follow any write cycle. note that the required wait period is dependant upon the register being read after the write. performing "dummy" reads of the byte_test register is a c onvenient way to guarantee that the minimum write-to-read timi ng restriction is met. table 6.1 also shows the number of dummy reads that are required before reading the register indicat ed. the number of byte_test reads in this table is based on the minimum timing for tcycle (45ns). fo r microprocessors with slower busses the number of reads may be reduced as long as the total time is equal to, or greater than the time specified in the table. note that dummy reads of the byte_test register are not required as long as the minimum time period is met. table 6.1 read after write timing rules register name minimum wait time for read following any write cycle (in ns) number of byte_test reads (assuming t cycle of 45ns) id_rev 0 0 irq_cfg 135 3 int_sts 90 2 int_en 45 1 byte_test 0 0 fifo_int 45 1 rx_cfg 45 1 tx_cfg 45 1 hw_cfg 45 1 rx_dp_ctrl 45 1 rx_fifo_inf 0 0 tx_fifo_inf 135 3 pmt_ctrl 315 7 gpio_cfg 45 1 gpt_cfg 45 1 gpt_cnt 135 3 word_swap 45 1 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 129 revision 2.9 (03-01-12) datasheet 6.2.2 special restrictions on back-to-back read cycles there are also restrictions on specific back-to -back read operations. thes e restrictions concern reading specific registers after readi ng resources that have side effects. in many cases there is a delay between reading the lan9221/lan9221i, and the subs equent indication of the expected change in the control register values. in order to prevent the host from reading stale data on back-to-back reads, minimum wait periods have been established. these periods are specified in table 6.2, "read after read timing rules" . the host processor is required to wait the specified period of time between read operations of specific combinations of resources. the wait period is dep endant upon the combination of registers being read. performing "dummy" reads of the byte_test register is a c onvenient way to guarantee that the minimum wait time restriction is met. table 6.2 also shows the number of dummy reads that are required for back-to-back read operations. the number of byte_test reads in this table is based on the minimum timing for t cycle (45ns). for microprocessors with slower busses the number of reads may be reduced as long as the total time is equal to , or greater than the time specified in the table. dummy reads of the byte_test register are not requ ired as long as the minimum time period is met. free_run 180 4 rx_drop 0 0 mac_csr_cmd 45 1 mac_csr_data 45 1 afc_cfg 45 1 e2p_cmd 45 1 e2p_data 45 1 table 6.2 read after read timing rules after reading... wait for this many ns or perform this many reads of byte_test (assuming t cycle of 45ns) before reading... rx data fifo 135 3 rx_fifo_inf rx status fifo 135 3 rx_fifo_inf tx status fifo 135 3 tx_fifo_inf rx_drop 180 4 rx_drop table 6.1 read after writ e timing rules (continued) register name minimum wait time for read following any write cycle (in ns) number of byte_test reads (assuming t cycle of 45ns) downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 130 smsc lan9221/lan9221i datasheet 6.3 pio reads pio reads can be used to access csrs or rx data and rx/tx status fifos. in this mode, counters in the csrs are latched at the beginning of the read cycl e. read data is valid as indicated in the timing diagram. pio reads can be performed using chip select (ncs) or read enable (nrd). either or both of these control signals mu st go high between cycles for the period specified. note: some registers have restrictions on the ti ming of back-to-back, write-read and read-read cycles. note: the data bus width is 16 bits note 6.2 when vddvario is 3.3v or 2.5v, the maximum t doff time is 7ns. when vddvario is 1.8v, the maximum t doff time is 9ns. note: a pio read cycle begins when both ncs and nrd are asserted . the cycle ends when either or both ncs and nrd are deasserted. they may be asserted and deasserted in any order. figure 6.2 pio read cycle timing table 6.3 pio read timing symbol description min typ max units t cycle read cycle time 45 ns t csl ncs, nrd assertion time 32 ns t csh ncs, nrd deassertion time 13 ns t csdv ncs, nrd valid to data valid 30 ns t asu address setup to ncs, nrd valid 0 ns t ah address hold time 0 ns t don data buffer turn on time 0 ns t doff data buffer turn off time note 6.2 ns t doh data output hold time 0 ns data bus ncs, nrd a[ 7:1 ] downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 131 revision 2.9 (03-01-12) datasheet 6.4 pio burst reads in this mode, performance is improved by allo wing up to 16 wo rd read cycles back-to-back. pio burst reads can be performed using chip select (ncs ) or read enable (nrd). either or both of these control signals must go high between bursts for the period specified. note: the data bus width is 16 bits note 6.3 when vddvario is 3.3v or 2.5v, the maximum t doff time is 7ns. when vddvario is 1.8v, the maximum t doff time is 9ns. note: a pio burst read cycle begins when both ncs and nrd are asserted. the cycle ends when either or both ncs and nrd are deasserted. they may be asserted and deasserted in any order. figure 6.3 pio burst read cycle timing table 6.4 pio burst read timing symbol description min typ max units t csh ncs, nrd deassertion time 13 ns t csdv ncs, nrd valid to data valid 30 ns t acyc address cycle time 45 t asu address setup to ncs, nrd valid 0 ns t adv address stable to data valid 40 t ah address hold time 0 ns t don data buffer turn on time 0 ns t doff data buffer turn off time note 6.3 ns t doh data output hold time 0 ns data bus ncs, nrd a[ 7:5 ] a[ 4:1 ] downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 132 smsc lan9221/lan9221i datasheet 6.5 rx data fifo direct pio reads in this mode the upper address inputs are not dec oded, and any read of the lan9221/lan9221i will read the rx data fifo. this mode is enabled when fifo_sel is driven high during a read access. this is normally accomplished by connecting the fifo_sel signal to high-order address line. this mode is useful when the host processor must increment its address when accessing the lan9221/lan9221i. timing is identical to a pio re ad, and the fifo_sel signal has the same timing characteristics as the address lines. note that address lines a[2:1] are stil l used, and address bits a[7:3] are ignored. note: the data bus width is 16 bits. note 6.4 when vddvario is 3.3v or 2.5v, the maximum t doff time is 7ns. when vddvario is 1.8v, the maximum t doff time is 9ns. note: an rx data fifo direct pio read cycle begins when both ncs and nrd are asserted. the cycle ends when either or both ncs and nrd are de-asserted. they may be asserted and de- asserted in any order. figure 6.4 rx data fifo direct pio read cycle timing table 6.5 rx data fifo direct pio read timing symbol description min typ max units t cycle read cycle time 45 ns t csl ncs, nrd assertion time 32 ns t csh ncs, nrd deassertion time 13 ns t csdv ncs, nrd valid to data valid 30 ns t asu address, fifo_sel setup to ncs, nrd valid 0 ns t ah address, fifo_sel hold time 0 ns t don data buffer turn on time 0 ns t doff data buffer turn off time note 6.4 ns t doh data output hold time 0 ns data bus ncs, nrd fifo _ sel a [ 2:1] downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 133 revision 2.9 (03-01-12) datasheet 6.6 rx data fifo di rect pio burst reads in this mode the upper address inputs are not decoded, and any burst read of the lan9221/lan9221i will read the rx data fifo. this mode is enab led when fifo_sel is driven high during a read access. this is normally accomplished by connecti ng the fifo_sel signal to a high-order address line. this mode is useful when the host processor must increment its address when accessing the lan9221/lan9221i. timing is identical to a pio bu rst read, and the fifo_sel signal has the same timing characteristics as the address lines. in this mode, performance is im proved by allowing an unlimited nu mber of back-to-back read cycles. rx data fifo direct pio burst reads can be per formed using chip select (ncs) or read enable (nrd). when either or both of these control signals go high, they must remain high for the period specified. note that address lines a[2:1] are stil l used, and address bits a[7:3] are ignored. note: the data bus width is 16 bits. note 6.5 when vddvario is 3.3v or 2.5v, the maximum t doff time is 7ns. when vddvario is 1.8v, the maximum t doff time is 9ns. note: an rx data fifo direct pio burst read cycle begins when both ncs and nrd are asserted. the cycle ends when either or bo th ncs and nrd are deasserted. they may be asserted and deasserted in any order. figure 6.5 rx data fifo direct pio burst read cycle timing table 6.6 rx data fifo dir ect pio burst read cycle timing symbol description min typ max units t csh ncs, nrd deassertion time 13 ns t csdv ncs, nrd valid to data valid 30 ns t acyc address cycle time 45 t asu address, fifo_sel setup to ncs, nrd valid 0 ns t adv address stable to data valid 40 t ah address, fifo_sel hold time 0 ns t don data buffer turn on time 0 ns t doff data buffer turn off time note 6.5 ns t doh data output hold time 0 ns data bus ncs, nrd fifo _ sel a [ 2:1] downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 134 smsc lan9221/lan9221i datasheet 6.7 pio writes pio writes are used for all lan 9221/lan9221i write cycles. pio wr ites can be performed using chip select (ncs) or write enable (nwr). either or both of these control signals must go high between cycles for the period specified. note: the data bus width is 16 bits. note: a pio write cycle begins when both ncs and nwr are asserted. the cycle ends when either or both ncs and nwr are deasserted. they may be asserted and deasserted in any order. figure 6.6 pio write cycle timing table 6.7 pio write cycle timing symbol description min typ max units t cycle write cycle time 45 ns t csl ncs, nwr assertion time 32 ns t csh ncs, nwr deassertion time 13 ns t asu address setup to ncs, nwr assertion 0 ns t ah address hold time 0 ns t dsu data setup to ncs, nwr deassertion 7 ns t dh data hold time 0 ns data bus ncs, nwr a [ 7:1 ] downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 135 revision 2.9 (03-01-12) datasheet 6.8 tx data fifo direct pio writes in this mode the upper address inputs are not decoded, and any write to the lan9221/lan9221i will write the tx data fifo. this mode is enabled wh en fifo_sel is driven hi gh during a write access. this is normally accomplished by connecting the fifo_sel signal to a high-order address line. this mode is useful when the host processor must increment its address when accessing the lan9221/lan9221i. timing is identical to a pio writ e, and the fifo_sel signal has the same timing characteristics as the address lines. note: the data bus width is 16 bits. note: a tx data fifo direct pio write cycle begins when both ncs and nw r are asserted. the cycle ends when either or bo th ncs and nwr are deasserted . they may be asserted and deasserted in any order. figure 6.7 tx data fifo direct pio write timing table 6.8 tx data fifo direct pio write timing symbol description min typ max units t cycle write cycle time 45 ns t csl ncs, nwr assertion time 32 ns t csh ncs, nwr deassertion time 13 ns t asu address, fifo_sel setup to ncs, nwr assertion 0 ns t ah address, fifo_sel hold time 0 ns t dsu data setup to ncs, nwr deassertion 7 ns t dh data hold time 0 ns data bus ncs, nwr fifo _ sel a [ 2:1 ] downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 136 smsc lan9221/lan9221i datasheet 6.9 power sequence timing figure 6.8 illustrates the devices power sequencing requirements. device power supplies can turn on in any order provided they all reach operational levels within the specified time period t pon . device power supplies can turn off in any order provided th ey all reach 0 volts within the specified time period t poff . figure 6.8 power sequence timing note: power must be applied and removed to all 3.3v power supply pins simultaneously (including the ethernet magnetics). note: power must be applied and removed to all vddvario power supply pins simultaneously. table 6.9 power sequence timing parameter description min typ max units notes t pon power supply turn on time 50 ms t poff power supply turn off time 500 ms vddvario 3.3v (all) t pon t poff downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 137 revision 2.9 (03-01-12) datasheet 6.10 reset timing figure 6.9 reset timing table 6.10 reset timing parameter description min typ max units notes t6.1 reset pulse width 30 ms t6.2 configuration input setup to nreset rising 200 ns t6.3 configuration input hold after nreset rising 10 ns t6.4 output drive after nreset rising 16 ns t6.1 t6.2 t6.3 nreset configuration signals output drive t6.4 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 138 smsc lan9221/lan9221i datasheet 6.11 eeprom timing the following specifies the eeprom timing requirements for t he lan9221/lan9221i: figure 6.10 eeprom timing table 6.11 eeprom timing values symbol description min typ max units t ckcyc eeclk cycle time 1110 1130 ns t ckh eeclk high time 550 570 ns t ckl eeclk low time 550 570 ns t cshckh eecs high before rising edge of eeclk 1070 ns t cklcsl eeclk falling edge to eecs low 30 ns t dvckh eedio valid before rising edge of eeclk (output) 550 ns t ckhdis eedio disable after rising edge eeclk (output) 550 ns t dsckh eedio setup to rising edge of eeclk (input) 90 ns t dhckh eedio hold after rising edge of eeclk (input) 0n s t ckldis eeclk low to data disable (output) 580 ns t cshdv eedio valid after eecs high (verify) 600 ns t dhcsl eedio hold after eecs low (verify) 0 ns t csl eecs low 1070 ns downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 139 revision 2.9 (03-01-12) datasheet chapter 7 operational characteristics 7.1 absolute maximum ratings* supply voltage (vddvario, vdd33reg, vdd33a) ( note 7.1 ) . . . . . . . . . . . . . . . . . . . . . 0v to +3.6v positive voltage on signal pins, with respect to ground ( note 7.2 ) . . . . . . . . . . . . . . . . . . . . . . . . . . +6v negative voltage on signal pins, with respect to ground ( note 7.3 ) . . . . . . . . . . . . . . . . . . . . . . . . -0.5v positive voltage on xtal1/clkin, with respect to ground . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +4.6v positive voltage on xtal2, with respect to ground . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +2.5v ambient operating temperature in still air (t a ). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . note 7.4 storage temperature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .-65 o c to +150 o c lead temperature range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . refer to jedec spec. j-std-020 hbm esd performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..+/- 5kv note 7.1 when powering this device from laboratory or system power supplies, it is important that the absolute maximum ratings not be exceeded or device failure can result. some power supplies exhibit voltage spikes on their outputs when ac power is switched on or off. in addition, voltage transients on the ac power line may appear on the dc output. if this possibility exists, it is suggeste d that a clamp circuit be used. note 7.2 this rating does not apply to the following pins: xtal1/clkin, xtal2, exres. note 7.3 this rating does not apply to the following pins: exres. note 7.4 0 o c to +70 o c for commercial version, -40 o c to +85 o c for industrial version. *stresses exceeding those listed in this section c ould cause permanent damage to the device. this is a stress rating only. exposure to absolute maximum rating conditions for extended periods may affect device reliability. functional operation of the device at any condition exceeding those indicated in section 7.2, "operating conditions**" , section 7.6, "dc electr ical specifications" , or any other applicable section of this sp ecification is not implied. 7.2 operating conditions** supply voltage (vddvario) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +1.62v to +3 .6v supply voltage (vdd33reg, vdd33a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +3.3v+/-300mv ambient operating temperature in still air (t a ). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . note 7.4 note: do not drive input signals withou t power supplied to the device. **proper operation of the lan9221/lan9221i is guar anteed only within the ra nges specified in this section. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 140 smsc lan9221/lan9221i datasheet 7.3 power consumption (device only) this section provides typical power consumption values for the lan9221/lan9221i in various modes of operation. these measurements were taken under the following conditions: temperature: ................................................................................................................... ................ +25 c device vdd (vddvario, vdd33reg, vdd33a): ........... ............................................................ +3.30v note: power dissipation is determined by operating fr equency, temperature, and supply voltage, as well as external source/sink requirements. note 7.5 d0 = normal operation, d1 = wol (wake on lan mode), d2= low power energy detect. table 7.1 power consumption device only mode total power - typical (mw) 10base-t operation d0, 10base-t /w traffic 249 d0, idle 243 d1, idle 146 d2, energy detect power down 64 d2, general power down 11 100base-tx operation d0, 100base-tx /w traffic 380 d0, idle 370 d1, idle 272 d2, energy detect power down (cable disconnected) 64 d2, general power down 11 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 141 revision 2.9 (03-01-12) datasheet 7.4 power consumption (device and system components) this section provides typical power consumption values for a comp lete ethernet interface based on the lan9221/lan9221i, including the power diss ipated by the magnetics and other passive components. note: the power measurements list below were taken under the following conditions: temperature: ................................................................................................................... ................ +25 c device vdd (vddvario, vdd33reg, vdd33a): ........... ............................................................ +3.30v note: power dissipation is determined by operating fr equency, temperature, and supply voltage, as well as external source/sink requirements. table 7.2 power consumption device and system components mode total power - typical (mw) 10base-t operation d0, 10base-t /w traffic 596 d0, idle 591 d1, idle 492 d2, energy detect power down 64 d2, general power down 11 100base-tx operation d0, 100base-tx /w traffic 522 d0, idle 510 d1, idle 416 d2, energy detect power down 64 d2, general power down 11 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 142 smsc lan9221/lan9221i datasheet 7.5 worst case current consumption this section details the worst case current cons umption for each device pow er supply. these values are provided to assist system designers with proper power supply design. these values cannot be used to determine typical power consumption of the device. the maximum supply current for each power supply is provided for the maximum vddvario voltage in all supported ranges as detailed in table 7.3 , ta b l e 7 . 4 , and ta b l e 7 . 5 . note: the current measurements listed below were taken under the following conditions: temperature: ................................................................................................................... ............. note 7.6 device vdd (vddvario, vdd33reg , vdd33a): .................................................................... note 7.7 note: current consumption is determined by operat ing frequency, temperature, and supply voltage, as well as external source/sink requirements. note 7.6 0 o c to +70 o c for commercial version, -40 o c to +85 o c for industrial version. note 7.7 the device vdd voltage levels are indicat ed in the supply current tables below. note: above values do not include the supply current for the magnetics. based on the recommended implementation, the maximum supply curr ent needed for the magnetics is 108ma. table 7.3 maximum supply current characteristics vddvario=3.3v + 300mv parameter supply name supply voltage max current units variable voltage supply current vddvario 3.6v 12 ma +3.3v regulator supply current vdd33reg 3.6v 85 ma +3.3v analog supply current vdd33a 3.6v 48 ma table 7.4 maximum supply current characteristics v ddvario=2.5v + 10% parameter supply name supply voltage max current units variable voltage supply current vddvario 2.75v 9 ma +3.3v regulator supply current vdd33reg 3.6v 85 ma +3.3v analog supply current vdd33a 3.6v 48 ma table 7.5 maximum supply current characteristics v ddvario=1.8v + 10% parameter supply name supply voltage max current units variable voltage supply current vddvario 1.98v 7 ma +3.3v regulator supply current vdd33reg 3.6v 85 ma +3.3v analog supply current vdd33a 3.6v 48 ma downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 143 revision 2.9 (03-01-12) datasheet 7.6 dc electri cal specifications this section details the dc electrical specific ations of the lan9221/lan9221i i/o buffers. the electrical specifications in this section are valid over the indicated voltage range and the temperature range specified in section 7.2, "operating conditions**" . note: when operating at reduced vddvario voltage levels (less than 3.0v), do not rely on internal pull-up and pull-down resistors to determine signal state. refer to the individual signal descriptions in chapter 2, pin description and configuration for more information. table 7.6 i/o buffer characteristics vddvario=3.3v +/- 300mv parameter symbol min typ max units notes vis type input buffer low input level v ili -0.3 v high input level v ihi 5.5 v negative-going threshold v ilt 1.01 1.18 1.35 v schmitt trigger positive-going threshold v iht 1.40 1.61 1.81 v schmitt trigger schmitt trigger hysteresis (v iht - v ilt ) v hys 350 423 492 mv input leakage (v in = vss or vddvario) i ih -10 +10 ua note 7.8 input leakage (v in = 5.5v) i ih 120 ua note 7.8 , note 7.9 input capacitance c in 2.5 pf vo12 type buffer low output level v ol 0.4 v i ol = 12ma high output level v oh vddvario - 0.4 vi oh = -12ma vod12 type buffer low output level v ol 0.4 v i ol = 12ma vod8 type buffer low output level v ol 0.4 v i ol = 8ma vo8 type buffer low output level v ol 0.4 v i ol = 8ma high output level v oh vddvario - 0.4 vi oh = -8ma iclk input buffer low input level v ilck -0.3 0.5 v high input level v ihck 1.4 3.6 v downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 144 smsc lan9221/lan9221i datasheet table 7.7 i/o buffer characteristics vddvario=2.5v +/- 10% parameter symbol min typ max units notes vis type input buffer low input level v ili -0.3 v high input level v ihi 5.5 v negative-going threshold v ilt 0.78 0.94 1.1 v schmitt trigger positive-going threshold v iht 1.13 1.32 1.51 v schmitt trigger schmitt trigger hysteresis (v iht - v ilt ) v hys 304 384 472 mv input leakage (v in = vss or vddvario) i ih -10 +10 ua note 7.8 input leakage (v in = 5.5v) i ih 120 ua note 7.8 , note 7.9 input capacitance c in 2.5 pf vo12 type buffer low output level v ol 0.4 v i ol = 9ma high output level v oh vddvario - 0.4 vi oh = -9ma vod12 type buffer low output level v ol 0.4 v i ol = 9ma vod8 type buffer low output level v ol 0.4 v i ol = 6ma vo8 type buffer low output level v ol 0.4 v i ol = 6ma high output level v oh vddvario - 0.4 vi oh = -6ma iclk input buffer low input level v ilck -0.3 0.5 v high input level v ihck 1.4 3.6 v downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 145 revision 2.9 (03-01-12) datasheet note 7.8 this specification applies to all inputs and tri-stated bi-directional pins. internal pull-down and pull-up resistors add +/- 50ua per-pin (typ ical). when operating at reduced vddvario voltage levels (less than 3.0v), do not rely on internal pull-up and pull-down resistors to determine signal state. refer to the individual signal descriptions in chapter 2, pin description and configuration for more information. note 7.9 this is the total v in input leakage for the entire device. this value should be divided by the number of pins driven to v in max to calculate per-pin leakage. for example, if 10 pins are driven to the maximum operational limit for v in (5.5v or 5.25v, dependant on vddvario), the per-pin input leakage is the maximum input leakage current divided by 10. table 7.8 i/o buffer characteristics vddvario=1.8v +/- 10% parameter symbol min typ max units notes vis type input buffer low input level v ili -0.3 v high input level v ihi 5.25 v negative-going threshold v ilt 0.54 0.69 0.83 v schmitt trigger positive-going threshold v iht 0.878 1.06 1.24 v schmitt trigger schmitt trigger hysteresis (v iht - v ilt ) v hys 277 368 525 mv input leakage (v in = vss or vddvario) i ih -10 +10 ua note 7.8 input leakage (v in = 5.25v) i ih 120 ua note 7.8 , note 7.9 input capacitance c in 2.5 pf vo12 type buffer low output level v ol 0.4 v i ol = 4ma high output level v oh vddvario - 0.4 vi oh = -4ma vod12 type buffer low output level v ol 0.4 v i ol = 4ma vod8 type buffer low output level v ol 0.4 v i ol = 3ma vo8 type buffer low output level v ol 0.4 v i ol = 3ma high output level v oh vddvario - 0.4 vi oh = -3ma iclk input buffer low input level v ilck -0.3 0.5 v high input level v ihck 1.4 3.6 v downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 146 smsc lan9221/lan9221i datasheet note 7.10 measured at the line side of the transformer, line replaced by 100 (+/- 1%) resistor. note 7.11 offset from16 ns pulse width at 50% of pulse peak note 7.12 measured differentially. note 7.13 min/max voltages guaranteed as measured with 100 resistive load. table 7.9 100base-tx transceiver characteristics parameter symbol min typ max units notes peak differential output voltage high v pph 950 - 1050 mvpk note 7.10 peak differential output voltage low v ppl -950 - -1050 mvpk note 7.10 signal amplitude symmetry v ss 98 - 102 % note 7.10 signal rise & fall time t rf 3.0 - 5.0 ns note 7.10 rise & fall time symmetry t rfs --0 . 5n s note 7.10 duty cycle distortion d cd 35 50 65 % note 7.11 overshoot & undershoot v os --5% jitter 1.4 ns note 7.12 table 7.10 10base-t transceiver characteristics parameter symbol min typ max units notes transmitter peak differential output voltage v out 2.2 2.5 2.8 v note 7.13 receiver differential squelch threshold v ds 300 420 585 mv downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 147 revision 2.9 (03-01-12) datasheet 7.7 clock circuit the lan9221/lan9221i can accept either a 25mhz cryst al (preferred) or a 25 mhz single-ended clock oscillator ( 50 ppm) input. the lan9221/lan9221i shares the 25mhz clock oscillator input (clkin) with the crystal input xtal1/clkin. if the single-en ded clock oscillator method is implemented, xtal2 should be left unconnected and clkin should be driven with a nominal 0-3.3v clock signal. the input clock duty cycle is 40% minimum, 50% typical and 60% maximum. it is recommended that a crystal utilizing matching parallel load capacitors be used for the lan9221/lan9221i crystal input/out put signals (xtal1, xtal2). see table 7.11, "lan9221/lan9221i crystal specifications" for crystal specifications. refer to application note an10.7 - parallel crystal circuit input voltage control for additional information. note 7.14 the maximum allowable values for frequency tolerance and frequency stability are application dependant. since any particular application must meet the ieee +/-50 ppm total ppm budget, the combination of these two values must be approximately +/-45 ppm (allowing for aging). note 7.15 frequency deviation over time is also referred to as aging. note 7.16 the total deviation for the transmitter clock frequency is specified by ieee 802.3u as +/- 50 ppm. note 7.17 0 o c for commercial version, -40 o c for industrial version. note 7.18 +70 o c for commercial version, +85 o c for industrial version. note 7.19 this number includes the pad, the bond wire and the lead frame. pcb capacitance is not included in this value. the xtal1/clkin and xtal2 pin and pcb capacitance values are required to accurately calculate the value of the two external load capacitors. these two external load capacitors determine th e accuracy of the 25.000 mhz frequency. table 7.11 lan9221/lan9221i crystal specifications parameter symbol min nom max units notes crystal cut at, typ crystal oscillation mode fundamental mode crystal calibration mode parallel resonant mode frequency f fund - 25.000 - mhz frequency tolerance @ 25 o cf tol - - +/-50 ppm note 7.14 frequency stability over temp f temp - - +/-50 ppm note 7.14 frequency deviation over time f age - +/-3 to 5 - ppm note 7.15 total allowable ppm budget - - +/-50 ppm note 7.16 shunt capacitance c o -7 t y p-p f load capacitance c l - 20 typ - pf drive level p w 300 - - uw equivalent series resistance r 1 --5 0 o h m operating temperature range note 7.17 - note 7.18 o c lan9221/lan9221i xtal1/clkin pin capacitance -3 t y p-p f note 7.19 lan9221/lan9221i xtal2 pin capacitance -3 t y p-p f note 7.19 downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 148 smsc lan9221/lan9221i datasheet chapter 8 package outline 8.1 56-qfn package notes: 1. all dimensions are in millimeters. 2. position tolerance of each terminal and exposed pad is 0.05 mm at maximum material condition. dimension b applies to plated terminals and is measured between 0.15 and 0.30 mm from the terminal tip. 3. the pin 1 identifier may vary, but is always located within the zone indicated. figure 8.1 56 pin qf n package definition table 8.1 56 pin qfn package parameters min nominal max remarks a 0.70 ~ 1.00 overall package height a1 0.00 0.02 0.05 standoff a2 ~ ~ 0.90 mold cap thickness d/e 7.85 8.00 8.15 x/y body size d1/e1 7.55 ~ 7.95 x/y mold cap size d2/e2 5.75 5.90 6.05 x/y exposed pad size l 0.30 ~ 0.50 terminal length b 0.18 0.25 0.30 terminal width e 0.50 basic terminal pitch downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 149 revision 2.9 (03-01-12) datasheet figure 8.2 56 pin qfn recommended pcb land pattern downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet revision 2.9 (03-01-12) 150 smsc lan9221/lan9221i datasheet chapter 9 datasheet revision history table 9.1 customer revision history revision level & date secti on/figure/entry correction rev. 2.9 (03-01-12) section 5.4.8, "flowflow control register," on page 113 updated pass control frames (bit 2) description to when set, the mac will pass the pause frame to the host... rev. 2.8 (07-14-11) table 2.4, system and power signals, on page 18 nled1 description modified to indicate that the signal is driven low when operating speed is 100mbs and is driven high during autonegotiation, when the cable is disconnected, and during 10 mbs operation. section 5.2.1, "rx fifo ports," on page 78 rx data fifo port is aliased to 16 word locations, not 16 dword locations. section 5.2.2, "tx fifo ports," on page 78 tx data fifo port is aliased to 16 word locations, not 16 dword locations. section 3.6.2.1, "tx checksum calculation," on page 33 added note stating tx checksum calculation should not be used for udp packets under ipv6. rev. 2.7 (03-15-10) chapter 2, "pin description and configuration," on page 15 added pin 1 designator to pin diagram section 7.2, "operating conditions**," on page 139 added note: do not drive input signals without power supplied to the device. section 6.9, "power sequence timing," on page 136 added power sequence timing section rev. 2.5 (11-13-08) all fixed various typos rev. 2.4 (10-24-08) all fixed various typos table 7.1 on page 140 and table 7.2 on page 141 added power consumption values. section 7.5, "worst case current consumption," on page 142 updated with current consumption values for various vddvario values. section 7.6, "dc electrical specifications," on page 143 added input capacitance and input leakage values. section 3.8, "general purpose timer (gp timer)," on page 37 changed incorrect gpt_cnt reference to gpt_load: on a reset, or when the timer_en bit changes from set 1 to cleared 0, the gpt_load field is initialized to ffffh. section 5.3.23, "e2p_cmd C eeprom command register," on page 102 corrected mac address loaded (bit 8) type from ro to r/wc table 7.11 on page 147 updated crystal specifications: drive level: 300uw esr: 50 ohms. downloaded from: http:///
high-performance 16-bit non-pci 10/100 ethernet controller with variable voltage i/o datasheet smsc lan9221/lan9221i 151 revision 2.9 (03-01-12) datasheet rev. 2.3 (08-18-08) note 7.9 on page 145 note following i/o buffer characteristics table modified: changed from: ".....the per-pin input leakage is 10 divided by the maximum input leakage current." to: ".....the per-pin input leakage is the maximum input leakage current divided by 10." rev. 2.2 (06-19-08) figure 1.2, "internal block diagram" diagram redone. the word core was added to the regulator block title. table 2.4, system and power signals, on page 18 changed vdd_core/vdd18core bulk capacitor value from 10uf to 4.7uf. rev. 2.2 (06-10-08) auto-negotiation advertisement on page 120 bits 9 and 15 relabeled as reserved, read-only (ro), with a default of 0. auto-negotiation advertisement on page 120 fixed definition of bits 11:10 when equal to 11 by adding advertise support for.. to beginning of definition. also added note stating when both symmetric pause and asymmetric pause support are advertised, the device will only be configured to, at most, one of the two settings upon auto-negotiation completion. section 3.5, "wake-up frame detection," on page 26 and section 5.4.1, "mac_crmac control register," on page 106 added note: when wake-up frame detection is enabled via the wuen bit of the wucsrwake- up control and status register , a broadcast wake- up frame will wake-up the device despite the state of the disable broadcast frame (bcast) bit in the mac_crmac control register . section 5.4.12, "wucsrwake-up control and status register," on page 115 fixed typo in bit 9: ... mac address [1:0] bit set to 0. was changed to ...mac ad dress [0] bit set to 0. section 3.6.1.1, "rx checksum calculation," on page 32 checksum = [b0, b1] + c0 + [b2, b3] + c1 + + [0, bn] + cn-1 changed to checksum = [b1, b0] + c0 + [b3, b2] + c1 + + [0, bn] + cn-1 section 2.2, external pull- up/pull-down resistors added section. added references to this section throughout the pin description tables as applicable. rev. 2.1 (05-13-08) section 1.1, "block diagram" removed the system memo ry block and arrow above the microprocessor/ microcontroller rev. 1.92 (10-22-07) chapter 2 pin description and configurationon page 15 pin assignment information re-organized into separate table. eeclk pin description in chapter 2 pin description and configurationon page 15 note added to eeclk pin description to indicate proper usage. table 9.1 customer revision history (continued) revision level & date secti on/figure/entry correction downloaded from: http:///

▲Up To Search▲

Price & Availability of LAN9221-ABZJ

	To Download LAN9221-ABZJ Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .