View HMP8156A_2024856.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

1 ? september 2003 hmp8156 ntsc/pal encoder features ? (m) ntsc and (b, d, g, h, i, m, n, cn) pal operation ? itu-r bt.601 and square pixel operation ? digital input formats - 4:2:2 ycbcr - 8-bit or 16-bit - 4:4:4 rgb - 16-bit (5, 6, 5) or 24-bit (8, 8, 8) - linear or gamma-corrected - 8-bit parallel itu-r bt.656 - seven overlay colors ? analog output formats - y/c + two composite - rgb + composite (scart) ? flexible video timing control - timing master or slave - selectable polarity on each control signal - programmable blank output timing - field output ? closed caption encoding for ntsc and pal ? 2x upscaling of sif video ? four 2x oversampling, 10-bit dacs ?i 2 c interface ? verilog models available applications ? multimedia pcs ? video conferencing ? video editing ? related products - ntsc/pal encoders: hmp8170 - ntsc/pal decoders: hmp8117 description the hmp8156 ntsc and pal encoder is designed for use in systems requiring the generation of high-quality ntsc and pal video from digital image data. ycbcr or rgb digital video data drive the p0-p23 inputs. overlay inputs are processed an d the data is 2x upsampled. the y data is optionally lowpass filtered to 5mhz and drives the y analog output. cb and cr are each lowpass filtered to 1.3mhz, quadrature modulate d, and summed. the result drives the c analog output. the digital y and c data are also added together and drive the two composite analog outputs. the ycbcr data may also be converted to rgb data to drive the dacs, allowing support for the european scart con- nector. the dacs can drive doubly-terminated (37.5 ? ) lines, and run at a 2x oversampling rate to simplify the analog output filter requirements. table of contents page functional block diagram. . . . . . . . . . . . . . . . . . . . . . . . . 2 functional operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 pixel data input formats . . . . . . . . . . . . . . . . . . . . . . . . . 3 input processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 pixel input and control signal timing . . . . . . . . . . . . . . . 5 video timing control . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 video processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 analog outputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 host interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 pinout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 pin descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 applications information . . . . . . . . . . . . . . . . . . . . . . . . . . 30 evaluation kits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 ordering information part number temp. range ( o c) package pkg. no. hmp8156cn 0 to 70 64 pqfp q64.14x14 hmp8156eval1 daughter card evaluation platform (note) hmp8156eval2 frame grabber evaluation platform (note) note: described in the applications section fn4269.4 caution: these devices are sensitive to electrostatic discharge; follow proper ic handling procedures. 1-888-intersil or 321-724-7143 | intersil (and design) is a registered trademark of intersil americas inc. copyright ? intersil americas inc. 2003. all rights reserved. all other trademarks mentioned are the property of their respective owners. n o t r e co m m e nd e d f o r n e w d e s i g ns n o r e c o m m e n d e d re p l a ce m e nt c o n t a c t o u r t e c h n i c a l s u p p o r t ce n t e r a t 1 - 8 8 8 - i n t e r s i l o r w w w . i n t e r s i l . c o m / t s c
2 hmp8156 functional block diagram overlay processing sa scl sda reset host interface optional gamma correction p0 - p23 4:2:2 to 4:4:4 sample conversion color space conversion 4:4:4 y/cb/cr (optional) lp filter y cb/cr closed captioning processing dac dac dac dac y/g ntsc/ ntsc/ c/b pal 1 pal 2/ r hsync vsync blank clk video timing clk2 field control vref fs adjust internal 1.225v reference 2 x upsample chroma modulation lp filter 4:4:4 to 8:8:8 (2:2:2 sif) optional 2x upscaling (sif mode)
3 functional operation the hmp8156 is a fully integrated digital encoder. it accepts digital video input data and generates four analog video out- put signals. the input data format is selectable and includes ycbcr, rgb, and overlay dat a. the outputs are config- urable to be either two composite video signals and y/c (s- video) or one composite and component rgb video. the hmp8156 accepts pixel data in one of several formats and transforms it into 4:4:4 sampled luminance and chromi- nance (ycbcr) data. if enabled, the encoder also mixes overlay data with the input data. the encoder then interpo- lates the ycbcr data to twice the pixel rate and low pass fil- ters it to match the bandwidth of the video output format. if enabled, the encoder also adds closed captioning informa- tion to the y data. at the same time, the encoder modulates the chrominance data with a digitally synthesized subcarrier. finally, the encoder outputs the luminance, chrominance, and their sum as analog signals using 10-bit d/a converters. the hmp8156 provides operating modes to support all ver- sions of the ntsc and pal standards and accepts full and sif size input data with rectangular (itu-r bt.601) and square pixel ratios. it operates from a single clock at twice the pixel clock rate determined by the operating mode. the hmp8156?s video timing control is flexible. it may oper- ate as the master generating the system?s video timing con- trol signals or it may accept external timing controls. the polarity of the timing controls and the number of active pixels and lines are programmable. pixel data input formats the hmp8156 accepts pixel dat a via the p0-p23 input pins. the definition of each pixel input pin is determined by the input format selected in the input format register. the defini- tion for each mode is shown in table 2. ycbcr pixel data the hmp8156 accepts 4:2:2 sampled ycbcr input data. the luminance and color difference signals are each 8 bits, scaled 0 to 255. values outside their nominal ranges (16- 235 for y and 16-240 for cb and cr) are processed normally. the color difference signals are time multiplexed into one 8- bit bus beginning with a cb sample. the y and cbcr busses may be input in parallel (16-bit mode) or may be time multi- plexed and input as a single bus (8-bit mode). the single bus may also contain sav and eav video timing reference codes (itu-r bt.656 mode). rgb data the hmp8156 accepts 4:4:4 sampled rgb component video input data. the color signals may be (8,8,8) for 24-bit mode or (5,6,5) for 16-bit mode. in 24-bit mode, they are scaled 0 to 255, black to whit e. in 16-bit mode, the encoder left shifts the input so that it has the same scale as 24-bit input. the rgb data may be linear or gamma corrected; if enabled, the encoder will gamma correct the input data. overlay data the hmp8156 accepts 5 bits of pixel overlay input data and combines it with the input pixel data. the data specifies an overlay color and the fractions of the new and original colors to be summed. blue screen generation in blue screen mode, the hmp8156 ignores the pixel input data and generates a solid, blue screen. the overlay inputs may be used to place information over the blue screen. input processing color space conversion for linear rgb input formats, the encoder applies gamma- correction using a selectable ga mma value of 1/2.2 or 1/2.8. the gamma-corrected rgb data fr om either the correction function in linear mode or the input port otherwise is con- verted to 4:4:4 sampled ycbcr data. for the ycbcr input formats, the encoder converts the 4:2:2 sampled data to 4:4:4 sampled data. the conversion is done by 2x upsampling the cb and cr data. the upsampling func- tion uses linear interpolation. overlay processing the hmp8156 accepts overlay data via the ol0-ol2, m0, and m1 pins. overlay mixing is done using the 4:4:4 ycbcr pixel data from the color space converter. the ycbcr data following overlay processing is used as input data by the video processing functions. the ol0-ol2 inputs select the color to be mixed with the pixel data. overlay colors 1-7 are standard color bar colors. overlay color 0 is special and disables mixing on a pixel by pixel basis. the overlay color palette is shown in table 1. note that overlay capability is not available when the 24-bit rgb input format is used. the encoder provides 4 methods for mixing the overlay data with the pixel data: disabled, external mixing, internal mixing and no mixing. the method used is selected in the input for- mat control register. overlay mixing: disabled when overlay mixing is disabled, the ol0-ol2, m0, and m1 inputs are ignored and the pixel data is not changed. table 1. overlay color palette ol2-ol0 color 000 001 010 011 100 101 110 111 pixel data 75% blue 75% red 75% magenta 75% green 75% cyan 75% yellow 100% white hmp8156
4 overlay mixing: external when external overlay mixing is selected, mixing of overlay data and pixel data is controlled by the m1 and m0 inputs. m1 and m0 indicate the mixing level between the pixel inputs and the overlay inputs, on a pixel-by-pixel basis. m1 and m0 are ignored if ol2-ol0 = 000. otherwise, they select the percentage of each color to sum as shown in table 3. in external mixing mode, there is no minimum number of pix- els an overlay color or pixel color must be selected. the mix- ing level may also vary at any rate. overlay mixing: internal mixing of overlay and pixel data may also be controlled inter- nally, and the m1 and m0 input pins are ignored. a transition from pixel data to overlays, from overlays to pixel data, or between different overlay colors triggers the mixing function. an overlay color must be sele cted for a minimum of three pixels for proper overlay operation in this mode. internal overlay mixing should not be used with the bt.656 input for- mat. when going from pixel to overlay data, mixing starts one pixel before the selection of the overlay color (ol2-ol1!= 000). the first pixel output before t he overlay uses 12.5% overlay color plus 87.5% pixel color. the next output is aligned with the selection of the overlay color and uses 87.5% overlay color plus 12.5% pixel color. additional outputs use 100% overlay color. when going from overlay to pixel data, mixing starts one pixel before the selection of the pixel color (ol2-ol0 = 000). the last pixel output of the ov erlay uses 87.5% overlay color plus 12.5% pixel color. the ne xt output uses 12.5% overlay color plus 87.5% pixel color. additional outputs use 100% pixel color. when going from one overlay color to another, mixing starts one pixel before the selection of the new overlay color, and uses 12.5% new overlay color plus 87.5% old overlay color. the next output is aligned with t he selection of the new over- lay color and uses 87.5% new overlay color plus 12.5% old overlay color. additional outputs use 100% new overlay color. overlay mixing: no mixing with no overlay mixing sele cted, whenever the ol0-ol2 inputs are non-zero, the overlay color is displayed. the m0 and m1 inputs are ignored, and no internal mixing is done. essentially, this is a hard switch between overlay and pixel data. in this mode, there is no minimum number of pixels an overlay color or pixel color must be selected. 2x upscaling table 2. pixel data input formats pin name 16-bit 4:2:2 ycbcr 8-bit 4:2:2 ycbcr bt.656 blue screen 16-bit rgb (5, 6, 5) 24-bit rgb p0 p1 p2 p3 p4 p5 p6 p7 cb0, cr0 cb1, cr1 cb2, cr2 cb3, cr3 cb4, cr4 cb5, cr5 cb6, cr6 cb7, cr7 ignored b0 b1 b2 b3 b4 g0 g1 g2 b0 b1 b2 b3 b4 b5 b6 b7 p8 p9 p10 p11 p12 p13 p14 p15 y0 y1 y2 y3 y4 y5 y6 y7 y0, cb0, cr0 y1, cb1, cr1 y2, cb2, cr2 y3, cb3, cr3 y4, cb4, cr4 y5, cb5, cr5 y6, cb6, cr6 y7, cb7, cr7 ycbcr data, sav and eav sequences ignored g3 g4 g5 r0 r1 r2 r3 r4 g0 g1 g2 g3 g4 g5 g6 g7 p16 p17 p18 p19 p20 p21 p22 p23 ol0 ol1 ol2 m0 m1 - - - r0 r1 r2 r3 r4 r5 r6 r7 table 3. overlay mixing factors m1, m0 % overlay color % pixel color 00 01 10 11 0 12.5 87.5 100 100 87.5 12.5 0 hmp8156
5 following overlay processing, 2x upscaling may optionally be applied to the pixel data. in this mode, the hmp8156 accepts sif resolution video at 50 or 59.94 frames per sec- ond and generates standard interlaced video at 262.5 lines per field (240 active) at 59.94 fields per second for (m, nsm) ntsc and (m) pal, and 3 12.5 lines per field (288 active) at 50 fields per second for (b, d, g, h, i, n, cn) pal. this mode of operation allows sif video to be upscaled to full resolution and recorded on a vcr or displayed on a tv. the input pixel data rate is reduced by half when 2x upscal- ing is enabled. the color space conversion generates, and the overlay mixer uses, 2:2:2 ycbcr data instead of 4:4:4 data. for rectangular pixel ntsc and pal video, the input rate is 6.75mhz during the active portion of each line instead of 13.5mhz. example sif input resolutions and resulting output resolutions are shown in table 4. the hmp8156 performs horizontal 2x upscaling by linear interpolation. the vertical scaling is done by line duplication. for typical line duplication, the same frame of sif pixel input data is used for both the odd and even fields. note that a frame of sif size input has about the same number of lines as a field of full size input. after 2x upscaling, the input is 4:4:4 ycbcr data ready for video processing. pixel input and c ontrol signal timing the pixel input timing and the video control signal input/out- put timing of the hmp8156 depend on the part?s operating mode. the periods when the encoder samples its inputs and generates its outputs are summarized in table 5. figures 1-9 show the timing of clk, clk2, blank , and the pixel and overlay input data with respect to each other. blank may be an input or an output; the figures show both. when it is an input, blank must arrive coin cident with the pixel and overlay input data; all are sampled at the same time. when blank is an output, its timing with respect to the pixel and overlay inputs depends on the blank timing select bit in the timing_i/o_1 register. if t he bit is cleared, the hmp8156 deasserts blank one clk cycle before it samples the pixel and overlay inputs. as shown in the timing figures, the encoder samples the inputs 1-7 clk2 periods after negating blank , depending on the operating mode. if the bit is set, the encoder deasserts blank during the same clk cycle in which it samp les the input data. in effect, the input data must arrive one clk cycle earlier than when the bit is cleared. this mode is not shown in the figures. 8-bit ycbcr format wi thout 2x upscaling when 8-bit ycbcr format is selected and 2x upscaling is not enabled, the data is latched on each rising edge of clk2. the pixel data must be [cb y cr y? cb y cr y?. . . ], with the first active data each scan line being cb data. overlay data table 4. typical resolutions for 2x upscaling input active resolution output active resolution 352 x 240 352 x 288 320 x 240 384 x 288 704 x 480 704 x 576 640 x 480 768 x 576 table 5. pixel input and control signal i/o timing input format 2x upscaling input port sampling video timing control (note) clk frequency pixel data overlay data input sample output on input output 8-bit ycbcr off every rising edge of clk2 same edge that latches y every rising edge of clk2 any rising edge of clk2 ignored one-half clk2 on rising edge of clk2 when clk is low. same edge that latches y data rising edge of clk2 when clk is low. rising edge of clk2 when clk is high. one-half clk2 16-bit ycbcr, 16-bit rgb, or 24-bit rgb off rising edge of clk2 when clk is low rising edge of clk2 when clk is high. one-half clk2 on 2nd rising edge of clk2 when clk is low either rising clk2 edge when clk is high one-fourth clk2 bt.656 off every rising edge of clk2 same edge that latches y not allowed any rising edge of clk2 ignored one-half clk2 on not available note: video timing control signals include hsync , vsync , blank and field. the sync and blanki ng i/o directions are independent; field is always an output. hmp8156
6 is latched when the y input data is latched. the pixel and overlay input timing is shown in figure 1. as inputs, blank , hsync , and vsync are latched on each rising edge of clk2. as outputs, blank , hsync , and vsync are output following the rising edge of clk2. if the clk pin is configured as an inpu t, it is ignor ed. if configured as an output, it is one-half the clk2 frequency 8-bit ycbcr format wi th 2x upscaling when 8-bit ycbcr format is selected, the data is latched on the rising edge of clk2 while clk is low. the pixel data must be [cb y cr y? cb y cr y?. . . ], with the first active data each scan line being cb data. overlay data is latched on the rising edge of clk2 that latches y pixel input data. the pixel and overlay input timing is shown in figure 2. as inputs, blank , hsync , and vsync are latched on the rising edge of clk2 while clk is low. as outputs, hsync , vsync , and blank are output following the rising edge of clk2 while clk is high. in this mode of operation, clk is one-half the clk2 frequency. 16-bit ycbcr, 16-bit rgb, 24-bit rgb formats without 2x upscaling when 16-bit ycbcr, 16-bit rg b data, or 24-bit rgb format is selected without 2x upscaling, the pixel data is latched on the rising edge of clk2 while clk is low. overlay data is also latched on the rising edge of clk2 while clk is low. the pixel and overlay input timing is shown in figures 3 - 5. as inputs, blank , hsync , and vsync are latched on the rising edge of clk2 while clk is low. as outputs, hsync , vsync , and blank are output following the rising edge of clk2 while clk is high. in these modes of operation, clk is one-half the clk2 frequency. 16-bit ycbcr, 16-bit rgb, 24-bit rgb formats with 2x upscaling when 16-bit ycbcr, 16-bit rgb data, or 24-bit rgb format is selected and 2x upscaling is enabled, data is latched on the rising edge of clk2 while clk is low. overlay data is latched on the rising edge of clk2 while clk is low. the pixel and overlay input timing is shown in figures 6-8 figure 1. pixel and overlay input timing - 8-bit ycbcr without 2x upscaling figure 2. pixel and overlay input timing - 8-bit ycbcr with 2x upscaling cb 0y 0cr 0y 1cb 2y 2 clk2 p8-p15 blank (input) pixel 0 pixel 1 pixel 2 ol0-ol2, m1, m0 y n pixel n blank (output) clk clk2 cb 0y 0cr 0y 1cb 2y 2 p8-p15 blank (input) pixel 0 pixel 1 pixel 2 ol0-ol2, m1, m0 y n pixel n blank (output) hmp8156
7 as inputs, blank , hsync , and vsync are latched on the rising edge of clk2 while clk is low. as outputs, hsync , vsync , and blank are output following the rising edge of clk2 while clk is high. clk is one-fourth the clk2 fre- quency. figure 3. pixel and overlay input timing 6-bit ycbcr without 2x upscaling figure 4. pixel and overlay input timing - 16-bit rgb without 2x upscaling y 0y 1y 2y 3y 4y 5 clk p8-p15 blank (input) cb 0cr 0cb 2cr 2cb 4cr 4 p0-p7 pixel 0 pixel 1 pixel 2 pixel 3 pixel 4 pixel 5 ol0-ol2, m1, m0 clk2 y n cr n-1 pixel n blank (output) rgb 0rgb 1rgb 2rgb 3rgb 4rgb 5 clk p0-p15 blank (input) pixel 0 pixel 1 pixel 2 pixel 3 pixel 4 pixel 5 ol0-ol2, m1, m0 clk2 rgb n pixel n blank (output) hmp8156
8 figure 5. pixel and overlay input timing - 24-bit rgb without 2x upscaling figure 6. pixel and overlay input timing - 16-bit ycbcr with 2x upsampling rgb 0rgb 1rgb 2rgb 3rgb 4rgb 5 clk p0-p24 blank (input) clk2 rgb n blank (output) blank (input) p8-p15 clk2 y 0 y 1 y n clk blank (output) ol0-ol2, m1, m0 pixel 0 pixel 1 pixel n p0-p7 cb 0 cr 0 cr n-1 hmp8156
9 8-bit parallel itu-r bt.656 format when itu-r bt.656 format is selected, data is latched on each rising edge of clk2. overlay data is latched when the y input data is latched. however, the overlay data must arrive three pixels after its corresponding y data. the pixel and overlay input timing is shown in figure 9. as inputs, the blank , hsync , and vsync pins are ignored since all timing is derived from the eav and sav sequences within the data stream. as outputs, blank , hsync and vsync are output following the rising edge of clk2. if the clk pin is configured as an input, it is ignored. if configured as an output, it is one-half the clk2 frequency. square pixel operation, overlay processing with internal mix- ing, and sif mode 2x upsampling are not supported for the bt.656 input format . also, the hsync , vsync , and blank signals must be configured as outputs. figure 7. pixel and overlay input timi ng - 16-bit rgb with 2x upsampling figure 8. pixel and overlay input timi ng - 24-bit rgb with 2x upsampling figure 9. pixel and overla y input timing - bt.656 blank (input) p0-p15 clk2 rgb 0 rgb 1 rgb n clk blank (output) ol0-ol2, m1, m0 pixel 0 pixel 1 pixel n blank (input) p0-p24 clk2 rgb 0 rgb 1 rgb n clk blank (output) cb 2 y 2 cr 2 y 3 cb 4 y 4 clk2 p8-p15 pixel 0 pixel 1 ol0-ol2, m1, m0 "ff" "00" blank ( output) "00" "xy" "10" pixel n-2 pixel n-1 "10" "80" pixel n hmp8156
10 video timing control the pixel and overlay data must be presented to the hmp8156 at 50 or 59.94 fields per second (interlaced). the video timing is controlled by the blank , hsync , vsync , field, and clk2 pins. hsync , vsync , and field timing the leading edge of hsync indicates the beginning of a horizontal sync interval. if hsync is an output, it is asserted for about 4.7 s. if hsync is an input, it must be active for at least two clk2 periods. the widt h of the horizontal compos- ite sync tip is determined fr om the video standard and does not depend on the width of hsync . the leading edge of vsync indicates the beginning of a vertical sync interval. if vsync is an output, it is asserted for 3 scan lines in (m, nsm) ntsc and (m, n) pal modes or 2.5 scan lines in (b, d, g, h, i, cn) pal modes. if vsync is an input, it must be asserted for at least two clk2 periods. when hsync and vsync are configured as outputs, their leading edges will occur simultaneously at the start of an odd field. at the start of an even field, the leading edge of vsync occurs in the middle of the line. when hsync and vsync are configured as inputs, if the leading edge of hsync occurs within 127 clk2 cycles of the leading edge of vsync , the encoder assumes it is at the start of an odd field. otherwise, it assumes it is processing an even field. the field signal is always an output and changes state near each leading edge of vsync . the delay between the syncs and field depends on the encoder?s operating mode as summarized in table 6. in modes in which the encoder uses clk to gate its inputs and outputs, the field signal may be delayed 0-12 additional clk2 periods. figure 10 illustrates the hsync , vsync , and field gen- eral timing for (m, nsm) ntsc and (m, n) pal. figure 11 illustrates the general timing for (b, d, g, h, i, cn) pal. in the figures, all the signals are shown active low (their reset state), and field is low during odd fields. table 6. field output timing operating mode clk2 delay comments sync i/o direction blank i/o direction input input 148 field lags vsync switch- ing from odd to even. field lags the earlier of vsync and hsync when syncs are aligned when switching from even to odd. input output 138 field lags vsync . output don?t care 32 field leads vsync . figure 10a. beginning an odd field figure 10b. beginning an even field figure 10. hsync , vsync , and field timing for (m, nsm) ntsc and (m, n) pal figure 11a. beginning an odd field figure 11b. beginning an even field figure 11. hsync , vsync , and field timing for (b, d, g, h, i, cn) pal hsync vsync field hsync vsync field hsync vsync field hsync vsync field hmp8156
11 blank timing the encoder uses the hsync , vsync , field signals to generate a standard composite video waveform with no active video. the signal includes only sync tips, color burst, and optionally, a 7.5 ire blanking setup. based on the blank signal, the encoder adds the pixel and overlay input data to the video waveform. the encoder ignores the pixel and overlay input data when blank is asserted. instead of the input data, the encoder generates the blanking level. the encoder also ignores the pixel and overlay inputs when generating closed captioning data on a specific line, even if blank is negated. there must be an even number of active and total pixels per line. in the 8-bit ycbcr modes, the number of active and total pixels per line must be a multiple of four. note that if blank is an output, half-line blanking on the output video cannot be done. the hmp8156 never adds a 7.5 ire blanking setup during the active line time on scan lines 1-21 and 263-284 for (m, nsm) ntsc, scan lines 523-18 and 260-281 for (m) pal, and scan lines 623-22 and 311-335 for (n) pal, allowing the generation of video test signals, timecode, and other infor- mation by controlling the pixel inputs appropriately. the relative timing of blank , hsync , and the output video depends on the blanking and sync i/o directions. the typical timing relation is shown in figure 12. the delays which vary with operating mode are indicate d. the width of the compos- ite sync tip and the location and duration of the color burst are fixed based on the video format. when blank is an output, the encoder asserts it during the inactive portions of active scan lines and for all of each inac- tive scan line. the inactive scan lines blanked each field are determined by the start_v_blank and end_v_blank registers. the inactive portion of active scan lines is determined by the start_h_blank and end_h_blank registers. the zero count for horizontal blanking is 32 clk2 cycles before the 50% point of the co mposite sync. from this zero point, the hmp8156 counts every other clk2 cycle. when the count reaches the value in the start_h_blank register, the encoder negates blank . when the count reaches the value in the end_h_blank register, blank is asserted. there may be an additional 0-7 clk2 delays in modes which use clk. the data pipeline delay through the hmp8156 is 26 clk2 cycles. in operating modes whic h use clk to ga te the inputs into the encoder, the delay may be an additional 0-7 clk2 cycles. the delay from blank to the start or end of active video is an additional one-half clk cycle when the blank tim- ing select bit is cleared. the active video may also appear to end early or start late since the hmp8156 controls the blank- ing edge rates. the delay from the active edge of hsync to the 50% point of the composite sync is 4-39 cl k2 cycles depending on the hmp8156 operating mode. the delay is shortest when the encoder is the timing master; it is longest when in slave mode. clk2 input the clk2 input clocks all of the hmp8156, including its video timing counters. for proper operation, all of the hmp8156 inputs must be synchronous with clk2. the fre- quency of clk2 depends on the device?s operating mode and the total number of pixels per line. the standard clock frequencies are shown in table 7. note that the color subcarrier is derived from the clk2 input. any jitter on clk2 will be transferred to the color subcarrier, resulting in color changes. just 400ps of jitter on clk2 causes up to a 1 o color subcarrier phase shift. thus, clk2 should be derived from a stable clock source, such as a crystal. the use of a pll to generate clk2 is not recom- mended. figure 12. hsync , blank , and output video timing blank composite video out hsync sync delay start h blank data pipe delay hmp8156
12 video processing upsampling video processing begins with the 4:4:4 sampled ycbcr data from the input processor. after overlay mixing and optional 2x upscaling, the hmp8156 upsamples the 4:4:4 data to generate 8:8:8 data. the encoder uses linear interpolation for the upsampling. filtering if enabled, the hmp8156 lowpass filters the y data to 5.0mhz. lowpass filtering y re moves any alia sing artifacts due to the upsampling process, and simplifies the analog output filters. the y 5.0mhz lowpass filter response is shown in figure 13. at this point, the hmp8156 also scales the y data to generate the proper output levels for the vari- ous video standards the hmp8156 lowpass filters th e cb and cr data to 1.3mhz prior to modulation. the lowpass filtering removes any alias- ing artifacts due to the upsampling process (simplifying the analog output filters) and al so properly bandwidth-limits cb and cr prior to modulation. the chrominance filtering is not optional like luminance filtering. the cb and cr 1.3mhz low- pass filter response is shown in figure 14. . table 7. typical video timing parameters video standard pixels per line hblank register values vblank register values clk2 (mhz) total active start end start end full input resolution, rectangular pixels (m, nsm) ntsc (b, d, g, h, i) pal (m) pal (n) pal (cn) pal 858 864 858 864 864 720 720 720 720 720 842 (0x34a) 853 (0x355) 842 (0x34a) 853 (0x355) 853 (0x355) 122 (0x7a) 133 (0x85) 122 (0x7a) 133 (0x85) 133 (0x85) 259 (0x103) 310 (0x136) 259 (0x103) 309 (0x135) 310 (0x136) 19 (0x13) 22 (0x16) 19 (0x13) 21 (0x15) 22 (0x16) 27.0 27.0 27.0 27.0 27.0 full input resolution, square pixels (m, nsm) ntsc (b, d, g, h, i) pal (m) pal (n) pal (cn) pa l 780 944 780 944 944 640 768 640 784 768 758 (0x2f6) 923 (0x39b) 758 (0x2f6) 923 (0x39b) 923 (0x39b) 118 (0x76) 155 (0x9b) 118 (0x76) 155 (0x9b) 155 (0x9b) 259 (0x103) 310 (0x136) 259 (0x103) 309 (0x135) 310 (0x136) 19 (0x13) 22 (0x16) 19 (0x13) 21 (0x15) 22 (0x16) 24.54 29.5 24.54 29.5 29.5 sif input resolution, rectangular pixels (m, nsm) ntsc (b, d, g, h, i) pal (m) pal (n) pal (cn) pal 429 432 429 432 432 352 352 352 352 352 834 (0x342) 845 (0x34d) 842 (0x34a) 853 (0x355) 853 (0x355) 130 (0x82) 141 (0x8d) 122 (0x7a) 133 (0x85) 133 (0x85) 259 (0x103) 310 (0x136) 259 (0x103) 309 (0x135) 310 (0x136) 19 (0x13) 22 (0x16) 19 (0x13) 21 (0x15) 22 (0x16) 27.0 27.0 27.0 27.0 27.0 sif input resolution, square pixels (m, nsm) ntsc (b, d, g, h, i) pal (m) pal (n) pal (cn) pal 390 472 390 472 472 320 384 320 392 384 758 (0x2f6) 923 (0x39b) 758 (0x2f6) 923 (0x39b) 923 (0x39b) 118 (0x76) 155 (0x9b) 118 (0x76) 155 (0x9b) 155 (0x9b) 259 (0x103) 310 (0x136) 259 (0x103) 309 (0x135) 310 (0x136) 19 (0x13) 22 (0x16) 19 (0x13) 21 (0x15) 22 (0x16) 24.54 29.5 24.54 29.5 29.5 figure 13a. full spectrum figure 13. y lowpass filter response ntsc square pixel clk2 = 24.54mhz ntsc or pal clk2 = 27.00mhz pal square pixel clk2 = 29.50mhz rectangular pixel 0 -10 -20 -30 -40 -50 -60 0 2 4 6 8 10 12 14 frequency (mhz) attenuation (db) hmp8156
13 . chrominance modulation the hmp8156 uses a numerically controlled oscillator (nco) clocked by clk2 and a sine look up rom to generate the color subcarrier. the subcarrier from the rom is pre-scaled to generate the proper levels for the various video standards. prescaling outside the cbcr data path minimizes color pro- cessing artifacts. the hmp815 6 modulates the filtered 8:8:8 chrominance data with the synthesized subcarrier. subcarrier phase the sch phase is 0 o after reset but then changes monotoni- cally over time due to residue in the nco. in an ideal sys- tem, zero sch phase would be maintained forever. in reality, this is impossible to achieve due to pixel clock fre- quency tolerances. if enabled, the hmp8156 resets the nco periodically to avoid an accumulation of sch phase error. the reset occurs at the beginning of each field to burst phase sequence. the sequence repeats every 4 fields for ntsc or 8 fields for pal. resetting the sch phase every four fields (ntsc) or eight fields (pal) avoids the accumulation of sch phase error at the expense of requiring any ntsc/pal decoder after the encoder be able to handle very minor ?jumps? (up to 2 o ) in the sch phase at the beginning of each four-field or eight- field sequence. most ntsc/pal decoders are able to han- dle this due to video editing requirements. composite video limiting the hmp8156 adds the luminance and modulated chromi- nance together with the syn c, color burst, and optional blanking pedestal to form the composite video data. if enabled in the video processing register, the encoder limits the active video so that it is always greater than one-eighth of full scale. this corresponds to approximately one-half the sync height. this allows the generation of ?safe? video in the event non-standard ycbcr values are input to the device. closed captioning if enabled in the auxiliary data control register, the hmp8156 generates closed captioning data on specified scan lines. the captioning data stream includes clock run-in and start bits followed by the captioning data. during closed caption- ing encoding, the pixel and overlay inputs are ignored on the scan lines containing captioning information. the hmp8156 has two 16-bit registers containing the cap- tioning information. each 16-bit register is organized as two cascaded 8-bit registers. one 16 -bit register (caption 21) is read out serially during line 18 , 21 or 22; the other 16-bit reg- ister (caption 284) is read out serially during line 281, 284 or 335. the data registers are shifted out lsb first. the bytes may be written in any order but both must be writ- ten within one frame time for proper operation. if the regis- ters are not updated, the enc oder resends the previously loaded values. figure 14a. full spectrum figure 14b. pass band figure 14. cb and cr lowpass filter response figure 13b. pass band figure 13. y lowpass filter response 0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0 01 234 5 67 attentuation (db) frequency (mhz) ntsc square pixel clk2 = 24.54mhz ntsc or pal clk2 = 27.00mhz pal square pixel clk2 = 29.50mhz rectangular pixel 0123456 frequency (mhz) 0 -10 -20 -30 -40 -50 -60 attentuation (db) ntsc square pixel clk2 = 24.54mhz ntsc or pal clk2 = 27.00mhz rectangular pixel pal square pixel clk2 = 29.50mhz 0 0.2 0.4 0.6 0.8 1.0 1.2 frequency (mhz) 0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0 attentuation (db) ntsc square pixel clk2 = 24.54mhz ntsc or pal clk2 = 27.00mhz rectangular pixel -3.5 -4.0 1.4 1.6 pal square pixel clk2 = 29.50mhz hmp8156
14 the hmp8156 provides a write st atus bit for each captioning line. the encoder clears the writ e status bit to ?0? when cap- tioning is enabled and both bytes of the captioning data reg- ister have been written. the encod er sets the write status bit to ?1? after it outputs the data, indicating the registers are ready to receive new data. captioning information may be enabled for either line, both lines, or no lines. the captio ning modes are summarized in table 8. controlled edges the ntsc and pal video standards specify edge rates and rise and fall times for portions of the video waveform. the hmp8156 automatically implem ents controlled edge rates and rise and fall times on these edges: 1. analog horizontal sync (rising and falling edges) 2. analog vertical sync interval (rising and falling edges) 3. color burst envelope 4. blanking of analog active video 5. overlay with internal mixing 6. closed captioning information analog outputs the hmp8156 converts the video data into analog signals using four 10-bit dacs running at the clk2 rate. the dacs output a current proportional to the digital data. the full scale output current is determined by the reference voltage vref and an external resistor rset. the full scale output current is given by i fullscale (ma) = 3.6 * vref (v)/rset (k ? )(eq 1.) vref must be chosen such that it is within the part?s operat- ing range; rset must be chosen such that the maximum output current is not exceeded. if the vref pin is not connec ted, the hmp8156 provides an internal reference voltage. ot herwise, the applied voltage overdrives the internal referenc e. if an external reference is used, it must decoupled from any power supply noise. an example external reference circuit is shown in the applica- tions section. the hmp8156 generates 1v pp nominal video signals across 37.5 ? loads corresponding to doubly terminated 75 ? lines. the encoder may also drive larger loads. the full scale output current and load must be chosen such that the maxi- mum output voltage is not exceeded. output dac filtering since the dacs run at 2x the pixel sample rate, the sin(x)/x rolloff of the outputs is grea tly reduced, and there are fewer high frequency artifacts in the output spectrum. this allows using simple analog output filt ers. the analo g output filter should be flat to f s /4 and have good rejection at 3f s /4. example filters are shown in the applications section. composite + y/c output mode the hmp8156 provides three output modes: s-video, rgb, and power down. when s-video outputs are selected, the encoder outputs the luminanc e, modulated chrominance, and two copies of the composite video signals. all four out- puts are time aligned. to reduce power dissipation, the second composite output dac may be turned off. the output may be disabled in the host control register. composite + rgb output mode when analog rgb video is se lected, the hmp8156 trans- forms the filtered 8:8:8 ycbcr data into 8:8:8 rgb data. the transform matrix uses fixe d coefficients to generate pal video levels for interfacing to a european scart connector. the encoder will not generate pr oper video levels if rgb output is selected with ntsc format. table 8. closed captioning modes closed captioning enable bits output line(s) captioning register write status bit 284a 284b 21a 21b 284 21 00 none ignored ignored always 1 always 1 01 21 (ntsc) 18 (m pal) 22 (other pal) ignored caption data always 1 0 = loaded 1 = output 10 284 (ntsc) 281 (m pal) 335 (other pal) caption data ignored 0 = loaded 1 = output always 1 11 21, 284 (ntsc) 18, 281 (m pal) 22, 335 (other pal) caption data caption data 0 = loaded 1 = output 0 = loaded 1 = output hmp8156
15 the analog rgb outputs have a range of 0.3-1.0v with no blanking pedestal. composite sync information (0.0-0.3v) may be optionally added to the green output. closed cap- tioning data is not inclu ded on the rgb outputs. the hmp8156 also generates composite video when in rgb output mode. the analog composite video is output onto the ntsc/pal 1 pin. red information is output onto the ntsc/pal 2 pin, blue information is output onto the c pin, and green information is output onto the y pin. all four out- puts are time aligned. power down mode when the power down mode is enabled, all of the dacs are powered down (forcing their outputs to zero) and most of the internal clocks are stopped. the host processor may still read from and write to the internal control registers. host interfaces reset the hmp8156 resets to its default operating mode on power up, when the reset pin is asserted for at least four clk cycles, or when the software reset bit of the host control reg- ister is set. during the rese t cycle, the encod er returns its internal registers to their reset state and deactivates the i 2 c interface. i 2 c interface the hmp8156 provides a standard i 2 c interface and sup- ports fast-mode (up to 400 kbps) transfers. the device acts as a slave for receiving and transmitting data only. it will not respond to general calls or in itiate a transfer. the encoder?s slave address is either 0100 000x b when the sa input pin is low or 0100 001x b when it is high. (the ?x? bit in the address is the i 2 c read flag.) the i 2 c interface consists of the sda and scl pins. when the interface is not active, scl and sda must be pulled high using external 4-6k ? pull-up resistors. the i 2 c clock and data timing is shown in figures 15 and 16. figure 15. i 2 c serial timing flow figure 16. register write programming flow sda scl start condition s 1-7 address 8 r/w 9 ack 1-7 data 89 ack stop condition p s = start cycle p = stop cycle a = acknowledge from master from hmp8156 0x40 or data write data data data read register pointed to by subaddr register pointed to by subaddr na = no acknowledge 0x41 or 0100 000 or p na a s chip addr a sub addr a data data a ap a chip addr s a sub addr a chip addr s 0100 0010 0x43 0x42 optional frame may be repeated n times optional frame may be repeated n times hmp8156
16 during i 2 c write cycles, the first data byte after the slave address specifies the sub address, and is written into the address register. only the seven lsbs of the subaddress are used; the msb is ignored. any remaining data bytes in the i 2 c write cycle are written to th e control registers, begin- ning with the register specifie d by the address register. the 7-bit address register is incremented after each data byte in the i 2 c write cycle. data written to reserved bits within regis- ters or reserved registers is ignored. during i 2 c read cycles, data from t he control register speci- fied by the address register is output. the address register is incremented after each data byte in the i 2 c read cycle. reserved bits within registers return a value of ?0?. reserved registers return a value of 00 h . the hmp8156?s operating modes are determined by the contents of its internal regist ers which are accessed via the i 2 c interface. all internal registers may be written or read by the host processor at any time. however, some of the bits and words are read only or reserved and data written to these bits is ignored. table 9 lists the hmp8156?s internal registers. their bit descriptions are listed in tables 10-27. table 9. control register names sub address (hex) control register reset condition 00 01 02 03 04 05 06 07-0e 0f 10 11 12 13 14-1f 20 21 22 23 24 25 26-2f 30-7f product id output format input format video processing timing i/o 1 timing i/o 2 aux data enable reserved host control closed caption_21a closed caption_21b closed caption_284a closed caption_284b reserved start h_blank low start h_blank high end h_blank start v_blank low start v_blank high end v_blank reserved test and unused 56 h 00 h 06 h a0 h 00 h 00 h 00 h - 18 h 80 h 80 h 80 h 80 h - 4a h 03 h 7a h 03 h 01 h 13 h - - table 10. product id register sub address = 00 h bit number function description reset state 7-0 product id this 8-bit register specifies the last two digits of the product number. it is a read-only register. data written to it is ignored. 56 h table 11. output format register sub address = 01 h bit number function description reset state 7-5 video timing standard 000 = (m) ntsc 001 = (m) ntsc with a 0 ire setup; also called (nsm) ntsc 010 = (b, d, g, h, i) pal 011 = (m) pal 100 = (n) pal 101 = combination (n) pal; also called (cn) pal 110 = reserved 111 = reserved 000 b 4-3 output format these bits must be set to ?00? during (m, nsm) ntsc and (m, n, cn) pal modes. 00 = composite + y/c 01 = reserved 10 = composite + rgb (no sync on green) 11 = composite + rgb (with sync on green) 00 b 2-0 reserved 000 b hmp8156
17 table 12. input format register sub address = 02 h bit number function description reset state 7-5 input format 000 = 16-bit 4:2:2 ycbcr 001 = 8-bit 4:2:2 ycbcr 010 = 8-bit parallel itu-r bt.656 011 = 16-bit linear rgb 100 = 16-bit gamma-corrected rgb 101 = 24-bit linear rgb 110 = 24-bit gamma-corrected rgb 111 = blue screen 000 b 4 gamma select these bits are ignored except during linear rgb input modes. 0 = 1 / 2.2 1 = 1 / 2.8 0 b 3reserved 0 b 2-1 overlay mixing mode these bits must be set to ?11? in 24-bit rgb input modes. internal mixing should not be selected in bt.656 input mode. 00 = no mixing 01 = internal mixing 10 = external mixing 11 = disable overlays 11 b 0 input resolution this bit must be set to ?0? during bt.656 input mode. 0 = full resolution (2x upscaling disabled) 1 = sif resolution (2x upscaling enabled) 0 b table 13. video processing register sub address = 03 h bit number function description reset state 7 luminance processing 0 = none 1 = y lowpass filtering enabled 1 b 6 composite video limiting 0 = none 1 = lower limit of composite active video is about half the sync height 0 b 5 sch phase mode 0 = never reset sch phase 1 = reset sch phase every 4 (ntsc) or 8 (pal) fields 1 b 4-0 reserved 00000 b hmp8156
18 table 14. timing i/o register #1 sub address = 04 h bit number function description reset state 7blank timing select this bit is ignored unless blank is configured to be an output. 0 = data for the first active pixel of the scan line must arrive the clk cycle after the encoder negates blank . 1 = data for the first active pixel of the scan line must arrive immediately after the encoder negates blank . 0 b 6reserved 0 b 5blank output control 0 = blank is an input 1 = blank is an output 0 b 4blank polarity 0 = active low (low during blanking) 1 = active high (high during blanking) 0 b 3hsync and vsync output control 0 = hsync and vsync are inputs 1 = hsync and vsync are outputs 0 b 2hsync polarity 0 = active low (low during horizontal sync) 1 = active high (high during horizontal sync) 0 b 1 vsync polarity 0 = active low (low during vertical sync) 1 = active high (high during vertical sync) 0 b 0field polarity 0 = active low (low during odd fields) 1 = active high (high during odd fields) 0 b table 15. timing i/o register #2 sub address = 05 h bit number function description reset state 7-5 reserved 000 b 4 clk output control 0 = clk is an input 1 = clk is an output 0 b 3 aspect ratio mode this bit must be set to ?0? during bt.656 input mode. 0 = rectangular (bt.601) pixels 1 = square pixels 0 b 2-0 reserved 00 b table 16. auxiliary data enable register sub address = 06 h bit number function description reset state 7-6 closed captioning enable 00 = closed caption disabled 01 = closed caption enabled for odd fields: line 21 for ntsc, line 18 for (m) pal, or line 22 for (b, d, g, h, i, n, cn) pal 10 = closed caption enabled for ev en fields: line 284 for ntsc, line 281 for (m) pal, or line 335 for (b, d, g, h, i, n, cn) pal 11 = closed caption enabled for both odd and even fields 00 b 5-0 reserved 000000 b hmp8156
19 table 17. host control register sub address = 0f h bit number function description reset state 7 software reset setting this bit to ?1? initiates a softw are reset. it is automatically reset to a ?0? after the reset sequence is complete. 0 b 6 power down enable 0 = normal operation 1 = power down mode 0 b 5 ntsc/pal 2 output mode 0 = enabled 1 = disabled 0 b 4 closed caption line 21 write status 0 = caption_21a and caption_21b data registers contain unused data 1 = data has been output, host processor may now write to the registers 1 b 3 closed caption line 284 write status 0 = caption_284a and caption_284b dat a registers contain unused data 1 = data has been output, host processor may now write to the registers 1 b 2-0 reserved 000 b table 18. closed caption_21a data register sub address = 10 h bit number function description reset state 7-0 line 21 caption data (first byte) this register is cascaded with the closed caption_21b data register and they are read out serially as 16 bits during line 18, 21, or 22 if line 21 captioning is enabled. bit d0 of the 21a data register is shifted out first. 80 h table 19. closed caption_21b data register sub address = 11 h bit number function description reset state 7-0 line 21 caption data (second byte) this register is cascaded with the closed caption_21a data register and they are read out serially as 16 bits during line 18, 21, or 22 if line 21 captioning is enabled. bit d0 of the 21a data register is shifted out first. 80 h table 20. closed caption_284a data register sub address = 12 h bit number function description reset state 7-0 line 284 caption data (first byte) this register is cascaded with the closed caption_284b data register and they are read out serially as 16 bits during line 281, 284, or 335 if line 284 captioning is enabled. bit d0 of the 284a data register is shifted out first. 80 h hmp8156
20 table 21. closed caption_284b data register sub address = 13 h bit number function description reset state 7-0 line 284 caption data (second byte) this register is cascaded with the closed caption_284a data register and they are read out serially as 16 bits during line 281, 284, or 335 if line 284 captioning is enabled. bit d0 of the 284a data register is shifted out first. 80 h table 22. start h_blank low register sub address = 20 h bit number function description reset state 7-0 assert blank output signal (horizontal) this 8-bit register is cascaded with star t h_blank high register to form a 10-bit start_horizontal_blank register. it specifies t he horizontal count (in 1x clock cycles) at which to start ignoring pixel data eac h scan line. the leading edge of hsync is count 020 h . this register is ignored unless blank is configured as an output. 4a h table 23. start h_blank high register sub address = 21 h bit number function description reset state 7-2 reserved 000000 b 1-0 assert blank output signal (horizontal) this 2-bit register is cascaded with star t h_blank low register to form a 10-bit start_horizontal_blank register. it specifies t he horizontal count (in 1x clock cycles) at which to start ignoring pixel data eac h scan line. the leading edge of hsync is count 020 h . this register is ignored unless blank is configured as an output. 11 b table 24. end h_blank register sub address = 22 h bit number function description reset state 7-0 negate blank output signal (horizontal) this 8-bit register specifies the horizontal c ount (in 1x clock cycles ) at which to start inputting pixel data each scan line. the leading edge of hsync is count 000 h . this reg- ister is ignored unless blank is configured as an output. 7a h hmp8156
21 table 25. start v_blank low register sub address = 23 h bit number function description reset state 7-0 assert blank output signal (vertical) this 8-bit register is casc aded with start v_blank high register to form a 9-bit start_vertical_blank register. during normal oper ation, it specifies the line number (n) to start ignoring pixel input data (and what li ne number to start blanking the output video) each odd field; for even fields, it occurs on line (n + 262) or (n + 313). during sif input mode, the register value (n ) specifies the line number to start ignoring pixel input data each noninterlaced input fram e. the output video wi ll be blanked starting on line number (n) each odd field; for even fields , it occurs on line (n + 262) or (n + 313). the leading edge of vsync at the start of an odd field is count 000 h (note that this does not follow standard ntsc or pal line numbering). this register is ignored unless blank is configured as an output. 03 h table 26. start v_blank high register sub address = 24 h bit number function description reset state 7-1 reserved 0000000 b 0 assert blank output signal (vertical) this 1-bit register is cascaded with star t v_blank low register to form a 9-bit start_vertical_blank register. this register is ignored unless blank is configured as an output. 1 b table 27. end v_blank register sub address = 25 h bit number function description reset state 7-0 negate blank output signal (vertical) during normal operation, this 8-bit register s pecifies the line number (n) to start inputting pixel input data (and what line number to start generating active output video) each odd field; for even fields, it occurs on line (n + 262) or (n + 313). during sif input mode, the register value (n ) specifies the line number to start inputting pixel input data each noninterlaced input frame. the output video will be active starting on line number (n) each odd field; for even fields , it occurs on line (n + 262) or (n + 313). the leading edge of vsync at the start of an odd field is count 000 h (note that this does not follow standard ntsc or pal line numbering). this register is ignored unless blank is configured as an output. 13 h hmp8156
22 pinout hmp8156 (pqfp) top view pin descriptions pin name pin number input/ output description p0-p15 58, 55-43, 38, 37 i pixel input pins. see table 2. p16-p23 32-27, 23, 22 i overlay or pixel inputs. see table 2. field 34 o field output. the field output indicates that the encoder is outputting the odd or even video field. the polarity of field is programmable. hsync 35 i/o horizontal sync input/output. as an input, this pin must be asserted during the horizontal sync intervals. if it occurs early, the line time will be shortened. if it occurs late, the line time will be lengthen by holding the outputs at the front porch level. as an output, it is as- serted during the horizontal sync intervals. the polarity of hsync is programmable. vsync 36 i/o vertical sync input/output. as an input, this pin must be asserted during the vertical sync intervals. if it occurs early, the field time will be shortened. if it occurs late, the field time will be lengthened by holding the outputs at the bl anking level. as an output, it is assert- ed during the vertical sync intervals. the polarity of vsync is programmable. blank 33 i/o composite blanking input/output. as an input, this pin must be asserted during the hor- izontal and vertical blanking in tervals. as an output, it is asserted during the horizontal and vertical blanking intervals. the polarity of blank is programmable. clk 39 i/o 1x pixel clock input/output. as an i nput, this clock must be free-running and synchro- nous to the clock signal on the clk2 pin. as an output, this pin may drive a maximum of one ls ttl load. clk is generated by di viding clk2 by two or four, depending on the mode. 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 comp 1 comp 2 fs_adjust vref gnd gnd p0 vaa gnd p1 p2 p3 p4 p5 p6 p7 vaa vaa y/g c/b gnd vaa gnd ntsc/pal1 gnd gnd ntsc/pal2 gnd vaa gnd gnd vaa scl sda sa gnd p23 p22 vaa reset gnd p21 ol1/p17 m0/p19 ol2/p18 m1/p20 ol0/p16 gnd 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 p8 p9 p10 p11 p12 p13 gnd clk2 vaa clk p14 p15 vsync hsync field blank hmp8156
23 clk2 41 i 2x pixel clock input. this clock must be a cont inuous, free-running clock. scl 18 i i 2 c interface clock input. sa 19 i i 2 c interface address select input. sda 20 i/o i 2 c interface data input/output. the circuit for this pin should include a 4-6k ? pull up resistor connected to vaa. reset 25 i reset control input. a logical zero for a mi nimum of four clk cycles resets the device. reset must be a logical one for normal operation. y (g) 3 o luminance analog current output. this out put contains luminanc e video, sync, blank- ing, and closed captioning information. in analog rgb output mode, green analog video is generated. it is capable of driving a 37.5 ? load. if not used, it should be connected to gnd. c (b) 7 o chrominance analog current output. this output contains chrominance video, and blanking information. in analog rgb output mo de, blue analog video is generated. it is capable of driving a 37.5 ? load. if not used, it should be connected to gnd. ntsc/pal 1 11 o composite video analog current output. this output contains composite video, sync, blanking, and closed captioning informatio n. it is capable of driving a 37.5 ? load. if not used, it should be connected to gnd. ntsc/pal 2 (r) 15 o composite video analog current output. this output contains composite video, sync, blanking, and closed captioni ng information. in analog rgb output mode, red analog video is generated. it is capable of driving a 37.5 ? load. if not used, it should be con- nected to gnd. vref 61 i voltage reference input. an optional external 1.235v reference may be used to drive this pin. if left floating, the in ternal voltage reference is used. fs_adjust 62 full scale adjust control. a resist or (rset) connected between this pin and gnd sets the full-scale output current of each of the dacs. comp 1 64 compensation pin. a 0.1 f ceramic chip capacitor should be connected between this pin and vaa, as close to the device as possible. comp 2 63 compensation pin. a 0.1 f ceramic chip capacitor should be connected between this pin and vaa, as close to the device as possible. vaa +5v power. a 0.1 f ceramic capacitor, in parallel with a 0.01 f chip capacitor, should be used between each group of vaa pins and g nd. these should be as close to the device as possible. gnd ground pin descriptions (continued) pin name pin number input/ output description hmp8156
24 absolute maximum ratings thermal information v aa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +6v all signal pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .-0.5v to +6v analog output short circuit duration. . . . . . . . . . . . . . . . . . . .indefinite input current, all pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1ma vapor phase soldering, 1 minute. . . . . . . . . . . . . . . . . . . . . . 220 o c operating conditions temperature range . . . . . . . . . . . . . . . . . . . . . . . . . . . .0 o c to 70 o c thermal resistance (typical, note 1) ja o c/w pqfp package . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 maximum junction temperature. . . . . . . . . . . . . . . . . . . . . . . 150 o c maximum storage temperature range . . . . . . . . . - 65 o c to 150 o c maximum lead temperature (soldering 10s). . . . . . . . . . . . . 300 o c (lead tips only) caution: stresses above those listed in ?absolute maximum ratings ? may cause permanent damage to the device. this is a stress o nly rating and operation of the device at these or any other conditi ons above those indicated in the operational sections of this specification is not i mplied. note: 1. ja is measured with the component mount ed on an evaluation pc board in free air. electrical specifications v aa = +5v 5%, rset = 124 ? , vref_in = 1.225v, unless otherwise specified parameter test condition min typ max units dc parameters, digital inputs except clk2, sda, scl input logic low voltage, v il --0.8v input logic high voltage, v ih 2.0 - - v input logic low current, i il v in = 0.4v - - -1 a input logic high current, i ih v in = 2.4v - - 1 a input capacitance, c in -5-pf dc parameters, clk2 input input logic low voltage, v il --0.3 x v aa v input logic high voltage, v ih 0.7 x v aa --v input logic low current, i il v in = 0.5v - - -10 ma input logic high current, i ih v in = v aa -0.5v - - 10 ma input capacitance, c in -5-pf dc parameters, sda and scl inputs input logic low voltage, v il --0.3 x v aa v input logic high voltage, v ih 0.7 x v aa --v input logic low current, i il v in = 0.5v - - -1 ma input logic high current, i ih v in = v aa -0.5v - - 1 ma input capacitance, c in -5-pf dc parameters, digital outputs output logic low voltage, v il i ol = 2ma - - 0.4 v output logic high voltage, v ih i oh = -2ma 2.4 - - v output capacitance, c out -5-pf dc parameters, analog outputs dac resolution 10 10 10 bits integral nonlinearity, inl -2-lsb differential nonlinearity, dnl - 0.5 - lsb hmp8156
25 output current - - 34.8 ma output impedance - 100k - ohms output capacitance i out = 0ma, clk = 13.5mhz - 25 - pf output compliance range 0-1.4v video level error internal voltage reference external voltage reference note 2 - - 10 5 % % dac to dac matching --5% vref output voltage pin not driven, us ing internal reference 1.13 1.225 1.32 v vref output current pin not driven, using internal reference -50 - 50 a vref input voltage pin connected to external reference shown in figure 32 1.112 1.235 1.358 v vref input current pin connected to external reference shown in figure 32 -500 - 500 a ac parameters, analog outputs differential gain error using analog output filter shown in figure 33a. -1-% differential phase error - 1 - degree snr (weighted) -70-db hue accuracy - 2 - degree color saturation accuracy -2-% luminance nonlinearity -1-% residual subcarrier --60-db sch phase sch phase reset enabled -1.5 0 1.5 degree analog output skew, t ask --5ns analog output delay, t ad - - 12 ns dac-dac crosstalk --60-db glitch energy using analog output filter shown in figure 33a. includes clock and data feedthrough -35-pv-s ac parameters, pixel interface - inputs pixel setup time, t s 6- -ns pixel hold time, t h 0- -ns control setup time, t s 6- -ns control hold time, t h 0- -ns clk frequency - - 14.75 mhz clk high time, clk h 27.1 - 40.7 ns clk low time, clk l 27.1 - 40.7 ns clk2 frequency - - 29.5 mhz electrical specifications v aa = +5v 5%, rset = 124 ? , vref_in = 1.225v, unless otherwise specified (continued) parameter test condition min typ max units hmp8156
26 clk2 high time, clk2 h 13.6 - 20.3 ns clk2 low time, clk2 l 13.6 - 20.3 ns clk to clk2 setup time, clk su 6- -ns clk to clk2 hold time, clk h 0- -ns ac parameters, pixel interface - outputs control output delay, t d 3 - 12 ns clk2 to clk output delay, clk d 0 - 12 ns ac parameters, i 2 c interface all ac and dc parameters meet the fast-mode i 2 c bus interface specification. reset* pulse width low, t res 4- -clk cycles power supply characteristics dac psrr at dc note 4 - 50 - db power supply range, v aa 4.75 5.0 5.25 v normal supply current, i aa - - 260 ma power-down supply current, i aa note 3 - - 750 a power dissipation - 1100 1300 mw notes: 2. output level is dependent on the voltage on vref, the value of rset, and the load. 3. if using an external voltage reference, it is not power ed down. the internal voltage reference is always powered down. 4. the supply voltage rejection is the relative va riation of the full-scale output driving a 37.5 ? load for a 0.5% supply variation: psrr = 20 x log ( ? v aa / ? v out ). electrical specifications v aa = +5v 5%, rset = 124 ? , vref_in = 1.225v, unless otherwise specified (continued) parameter test condition min typ max units typical performance curves figure 17. noise spectrum (ntsc) figure 18. ntsc color bar vector scope plot 1.0 2.0 3.0 4.0 5.0 (mhz) average 0.0 -5.0 -10.0 -15.0 -20.0 -25.0 -30.0 -35.0 -40.0 -45.0 -50.0 -55.0 -60.0 -65.0 -70.0 -75.0 -80.0 -85.0 -90.0 -95.0 -100.0 field = 1 line = 47 amplitude (0db = 714mv p-p ) bandwidth 10khz to full wfm ---> pedestal noise level = -79.9db rms apl = 44.3% system line l 47 f1 angle (deg) 0.0 gain x1.000 0.000db 525 line ntsc burst from source setup 7.5% hmp8156
27 figure 19. ntsc fcc color bar figure 20. luminance non linearity (n tsc) figure 21. line frequency (ntsc) typical performance curves (continued) 97.9 99.1 99.1 99.8 100.0 105.0 104.5 104.0 103.5 103.0 102.5 102.0 101.5 101.0 100.5 100.0 99.5 99.0 98.5 98.0 97.5 97.0 96.5 96.0 95.5 95.0 1st 2nd 3rd 4th 5th average field = 1 line = 47 luminance non linearity (%) wfm ---> 5 step pk-pk = 2.1 0.00 (%) -0.4 -0.2 0.0 0.2 0.4 line frequency 15.734 (khz) field frequency 59.94 (hz) line frequency error average off (%) hmp8156
28 figure 22. h sync jitter in a frame (ntsc) figure 23. sch phase measurement figure 24. noise spectrum (pal) figur e 25. pal color bar vector scape plot typical performance curves (continued) line jitter (line 20 to 250) 2ns p-p 525 line ntsc mean sch 0.8 degrees average average apl = 40.0% system line l 72 angle (deg) 0.0 gain x1.000 0.000db 625 line pal burst from source display +v and -v hmp8156
29 figure 26. colourbar (pal) figure 27. luminance non linearity (pal) figure 28. line frequency (pal) typical performance curves (continued) line = 72 wfm ---> colour bar 97.9 99.1 99.1 99.8 100.0 105.0 104.5 104.0 103.5 103.0 102.5 102.0 101.5 101.0 100.5 100.0 99.5 99.0 98.5 98.0 97.5 97.0 96.5 96.0 95.5 95.0 1st 2nd 3rd 4th 5th average line = 72 luminance non linearity (%) wfm ---> 5 step pk-pk = 1.4 0.00 (%) -0.4 -0.2 0.0 0.2 0.4 line frequency 15.625 (khz) field frequency 50.00 (hz) line frequency error average off (%) hmp8156
30 applications information pal teletext teletext encoding may be implemented on any line by driv- ing the pixel inputs with appropriate data. for ycbcr input modes, cb and cr should equal 128 to disable the color information. for rgb input modes, r, g, and b should always have the same value to disable the color information. vertical blanking must be negated on the first scan line con- taining teletext information. if there are unused scan lines between teletext data and active video, blank must remain off and the pixel inputs should be set to the black level. video test signals video test signals may be generated by driving the pixel inputs with appropriate data. mo st of the video test signals require using ycbcr pixel data. vertical blanking must be negated on the first scan line con- taining video test signals. if there are unused scan lines between test signal data and active video, blank must remain off and the pixel inputs should be set to the black level. pcb considerations a pcb board with a minimum of 4 layers is recommended, with layers 1 and 4 (top and bottom) for signals and layers 2 and 3 for power and ground. the pcb layout should implement the lowest possible noise on the power and ground planes by pro- viding excellent decoupling. pcb trace lengths between groups of v aa and gnd pins should be as short as possible. component placement the optimum layout places th e hmp8156 at the edge of the pcb and as close as possible to the video output connector. external components should be positioned as close as pos- sible to the appropriate pin, ideally such that traces can be connected point to point. chip capacitors are recommended where possible, with radial lead ceramic capacitors the sec- ond-best choice. traces containing digital signals should not be routed over, under, or adjacent to the anal og output traces to minimize crosstalk. if this is not possible, coupling can be minimized by routing the digital signals at a 90 degree angle to the ana- log signals. the analog output tr aces should also not overlay the hmp8156 and v cc power planes to maximize high-fre- quency power supply rejection. power and ground planes a common ground plane for all devices, including the hmp8156, is recommended. ho wever, placing the encoder on an electrically connected gnd peninsula reduces noise levels. all gnd pins on the hmp8156 must be connected to the ground plane. typical pow er and ground planes are shown in figure 31. the hmp8156 should have its own power plane that is iso- lated from the common power pl ane of the board, with a gap between the two power planes of at least 1/8 inch. all v aa pins of the hmp8156 must be connected to this hmp8156 power plane. the hmp8156 power plane should be connected to the board?s normal v cc power plane at a single point though a low-resistance ferrite bead, such as a ferroxcube 5659065- 3b, fair-rite 2743001111, or tdk bf45-4001. the ferrite bead provides resistance to swit ching currents, improving the performance of hmp8156. a si ngle, large capacitor should also be used between the hmp8156 power plane and the ground plane to control low-frequency power supply ripple. figure 29. h sync jitter in a frame (pal) figure 30. sch phase measurement typical performance curves (continued) hmp8156
31 for proper operation, power supp ly decoupling is required. it should be done using a 0.1 f ceramic capacitor in parallel with a 0.01 f chip capacitor for each group of v aa pins to ground. these capacitors should be located as close to the v aa and gnd pins as possible, using short, wide traces. if a separate linear regulator is used to provide power to the hmp8156 power plane, the power-up sequence should be designed to ensure latchup will not occur. a separate linear reg- ulator is recommended if the power supply noise on the v aa pins exceeds 200mv. about 10% of the noise (that is less than 1mhz) on the v aa pins will couple onto the analog outputs. figure 31a. v cc and v aa planes figure 31b. common ground plane figure 31. example power and ground planes lp vcc vaa ferrite bead bulk area capacitor 8156 pcb analog filters conn. lp gnd 8156 pcb analog filters conn. hmp8156
32 external reference voltage if an external reference voltage is used, its circuitry should receive power from the same plane as the hmp8156. the external vref must also be stable and well decoupled from the power plane. an example vref circuit using a band gap reference diode is shown in figure 32. analog output filters the various video standards specify the freqency response of the video signal. the hm p8156 uses 2x oversampling dacs to simplify the reconstruc tion filter required. example post filters are shown in figur e 33. the analog output filters should be as close as possible to the hmp8156. evaluation kits there are two evaluation platforms available. the hmp8156eval1 is a small daughter card containing the encoder, voltage references and bypassing, analog output filters and connectors, a bt.656 interface and connector, and a 50 pin two row header. the header allows connecting the pixel and control pins of th e encoder into an existing sys- tem. the analog outputs allow the encoder?s performance to be observed and measured. the hmp8156eval2 is a standard size pc add in card with an isa bus interface and application software. the hmp8156eval2 kit is a comple te system which allows dem- onstrating all of the encoder?s operating modes. it has ana- log video inputs for composite, s-video, and component rgb signals. the analog signals are converted/decoded to the digital domain and input to the encoder. the board also provides a 3 megabyte video ram for image capture and display and a bt.656 connector and interface. figure 33a. high quality filter figure 33b. low cost filter figure 33. example post-filter circuits figure 32. external reference voltage circuit 0.01 f + 4.7 f vaa 1.235v icl8069 6.8k rl 75 1.0 h 82pf 2.2 h 330pf 1.0 h 39pf 330pf 75 rl 75 2.7 h 560pf 560pf 75 hmp8156
33 metric plastic quad flatpack packages (mqfp) d d1 e e1 -a- pin 1 a2 a1 a 12 o -16 o 12 o -16 o 0 o -7 o 0.40 0.016 min l 0 o min plane b 0.005/0.009 0.13/0.23 with plating base metal seating 0.005/0.007 0.13/0.17 b1 -b- e 0.008 0.20 a-b s d s c m 0.076 0.003 -c- -d- -h- q64.14x14 (jedec ms-022be issue b) 64 lead metric plastic quad flatpack package symbol inches millimeters notes min max min max a - 0.124 - 3.15 - a1 0.004 0.010 0.10 0.25 - a2 0.100 0.108 2.55 2.75 - b 0.012 0.018 0.30 0.45 6 b1 0.012 0.016 0.30 0.40 - d 0.672 0.682 17.08 17.32 3 d1 0.546 0.556 13.88 14.12 4, 5 e 0.673 0.681 17.10 17.30 3 e1 0.547 0.555 13.90 14.10 4, 5 l 0.029 0.040 0.73 1.03 - n64 647 e 0.032 bsc 0.80 bsc - rev. 1 4/99 notes: 1. controlling dimension: millimeter. converted inch dimensions are not necessarily exact. 2. all dimensions and toleranc es per ansi y14.5m-1982. 3. dimensions d and e to be determined at seating plane . 4. dimensions d1 and e1 to be determined at datum plane . 5. dimensions d1 and e1 do not include mold protrusion. allowable protrusion is 0.25mm (0.010 inch) per side. 6. dimension b does not include dambar protrusion. allowable dambar protrusion shall be 0.08mm (0.003 inch) total. 7. ?n? is the number of terminal positions. -c- -h- hmp8156 all intersil u.s. products are manufactured, asse mbled and tested utilizing iso9000 quality systems. intersil corporation?s quality ce rtifications can be viewed at www.intersil.com/design/quality intersil products are sold by description only. intersil corpor ation reserves the right to make changes in circuit design, soft ware and/or specifications at any time without notice. accordingly, the reader is cautioned to verify that data sheets are current before placing orders. information furnishe d by intersil is believed to be accurate and reliable. however, no responsibility is assumed by intersil or its subsidiaries for its use; nor for any infringements of paten ts or other rights of third parties which may result from its use. no license is granted by implication or otherwise under any patent or patent rights of intersil or its subsidiari es. for information regarding intersil corporation and its products, see www.intersil.com

▲Up To Search▲

Price & Availability of HMP8156A

	To Download HMP8156A Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .