View M66332FP_2355967.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

1 mitsubishi digital assp ? M66332FP facsimile image data processor mitsubishi digital assp ? M66332FP facsimile image data processor description the m66332 is a facsimile image processing controller that converts analog signals that are photoelectrically converted by an image sensor into bi-level signals. it has image processing functions such as peak detection, uniformity correction, g correction, mtf compensation, de- tector of background and object levels, dither control, sepa- ration of image data area, scale down, and area specification. this controller has a built-in 5-bit flash type a-d converter and interface circuits to image sensor, analog signal process- ing circuit, and codec (coder & decoder) to simplify control of the readout mechanism. features 0 high speed scan (max. 2 ms/line, typ. 5 ms/line) 0 a3 (8 pixels/mm) line sensor attachment 0 image sensor (ccd,cis) control signal generation ccd: sh, ck1, ck2, rs contact sensor (cis): sh, ck1 (or ck2) 0 analog signal processing circuit control signal generation clamp, s/h, agc, dsch 0 built-in 5-bit flash type a-d converter 0 bi-level data external input/output interface serial output ( ? m66330) 8-bit mpu bus output with external dma control signal 0 image data processing g correction uniformity correction (block correction in units of 8 pixels) mtf compensation (1 dimension) detector of background and object level (programmable) dithering control ? dither method (16 levels using 4 4 matrix) separation of image data area (1 dimension) scale down a3 ? b4, a3 ? a4, b4 ? a4 0 5v single power supply application facsimiles nc: no connection outline 56p6n-a pin configuration (top view) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 17 18 19 20 21 22 23 24 25 26 27 28 56 55 54 53 52 51 50 49 48 47 46 45 ck2 ck1 rs clamp s/h gnd test sysck test v cc svid sclk stim srdy ptim agc d7 d6 d5 d4 d3 d2 d1 d0 gnd v cc a3 a2 a1 a0 dgnd dv cc M66332FP codec interface analog signal prosessing interface test pin system clock sensor interface mpu interface (data) mpu interface (address) sh v cc dak drq gnd gnd reset wr rd cs v cc gnd dma interface sensor interface mpu interface dsch av cc agnd v bl v ml1 gnd v ml2 v ml3 v wl nc asig test analog signal processing interface black basic supply voltage middle basic supply voltage 1 middle basic supply voltage 2 middle basic supply voltage 3 white basic supply voltage test pin test pin analog signal processing interface processing interface sensor interface analog signal
2 mitsubishi digital assp ? M66332FP facsimile image data processor block diagram table 1 image processing functions remarks ? operated with system clock and pre_data (registers 2, 3) ? built-in sram as correction memory (304 words 5bits) (read/write allowed from mpu) ? no need for compensation memory ? built-in sram as dither memory (16 words 4bits) (read/write allowed from mpu) ? no need for processing memory ? apply external voltage (resistor connection is also allowed) to a-d converter middle basic supply voltage pins. ? built-in 5-bit flash a-d converter image processing function read width resolution read speed uniformity correction mtf compensation simple bi-level conversion pseudo half-tone separation of image data area scale down g correction image sensor control signal analog signal processing specifications ? a4, b4, a3 ? 8 pixels/mm (primary scanning direction) ? 5ms/line typ. 2ms/line maximum ? white correction only ? block correction in units of 8 pixels ? 50% correction range ? laplacian filter circuit for 3 1 pixels in current line (1 dimension) ? floating threshold method using background and object level detection circuit ? dither method: 16 levels (4 4matrix) ? detection by brightness difference in 5 1 pixels area in current line ? selection method ? scale down: a3 ? b4 set to 13/15; b4 ? a4, 9/11; a3 ? a4, 12/17 ? logarithmic correction ? control signal generation for contact sensor (cis) and scale down ccd ? generate control signals for external clamp circuit, sample/hold circuit, and agc circuit 15 ptim 56 sh 2 ck1 1 ck2 3 rs 4 clamp 5 s/h 16 agc 17 dsch 27 asig 20 v bl 21 v ml1 23 v ml2 24 v ml3 25 v wl 8 sysck 18 av cc 29 dv cc 10 35 46 55 44 d7 37 d0 34 a3 31 a0 49 wr 48 rd 47 cs 50 reset 53 drq 54 dak 11 svid 12 sclk 13 stim 14 srdy 52 51 45 36 22 6 30 19 5bit a-d converter (flash type) uniformity correction correction memory (304 words 5bits) adc reference voltage sensor interface analog signal interface sensor control analog control to each block image processing sequence control signal system clock adc analog vcc adc logic vcc v cc codec interface cur out/ scale down convert to bi-level dma control separation of image data area detection of image data area simple bi-level conversion/background and object level detection mtf compensation mpu bus interface collective dithering (16 levels) sram 16 words 4bits ~~ dma interface mpu interface mpu interface (address) mpu interface (data) agnd gnd dgnd adc analog gnd adc logic gnd
3 mitsubishi digital assp ? M66332FP facsimile image data processor pin descriptions description ccd: shift pulse signal to transmit photo charges from the sensor to the transfer unit. cis: start signal for the sensor read circuit. ccd: clock pulse signal for sequentially transmitting the transfer unit signal charge of the sensor. cis: clock pulse signal for the sensor read circuit shift register. reverse of ck1. pulse to reset the voltage of the ccd sensor floating capacitor to initial status. read roller pulse motor control signal. clamp pulse to set the dark level of the sensor to reference voltage of the digital circuit. sample-hold signal to smooth out sensor image signal waveform. analog signals. external agc circuit gain down signal. external agc circuit gain up signal. data transmission ready signal from codec. data transmission bound signal for codec. clock signal for transmitting image data to codec. serial output of image data to codec. h: black; l: white. dma request signal to external dma controller for parallel output of image data through mpu bus. dma acknowledge signal from external dma controller for the above drq signal. system clock input pin. system reset signal. resets counter, register, f/f, and latch, sets internal memory in standby mode, and halts clock generation circuit. chip select signal used by mpu to access m66332. set to h in operating mode (agc, unif, scan). control signal used by mpu to read data from m66332. control signal used by mpu to write data to m66332. address signals used to access m66332 internal registers. 8-bit bidirectional buffer. block sensor interface analog circuit interface codec interface dma interface clock mpu interface pin names sh ck1 ck2 rs ptim clamp s/h asig agc dsch srdy stim sclk svid drq dak sysck reset cs rd wr a0~a3 d0~d7 i/o o o o o o o o i o o i o o o o i i i i i i i i/o
4 mitsubishi digital assp ? M66332FP facsimile image data processor pin descriptions (continued) description plus supply voltage. plus supply voltage for a-d converter analog units. plus supply voltage for a-d converter logic units. gnd pin. ground for a-d converter analog units. ground for a-d converter digital units. a-d converter white basic supply voltage pin. a-d converter black basic supply voltage pin. middle basic supply voltage pin. v ml1 =(v wl C v bl ) /4 middle basic supply voltage pin. v ml2 =2 (v wl C v bl ) /4 middle basic supply voltage pin. v ml3 =3 (v wl C v bl ) /4 test input pin. fix to l. test output pin. keep open. block others pin names v cc av cc dv cc gnd agnd dgnd v wl v bl v ml1 v ml2 v ml3 test(in) test(out) i/o CC CC CC CC CC CC CC CC CC CC CC CC CC functional description the following items which are necessary to use the image processing functions of the m66332 are described. (1) operating mode (2) line period and read sequence (3) image processing function (4) sensor unit/analog signal processing unit interface (5) codec interface (6)read/write to dither memory and uniformity correction memory (7) reset (8) image quality control using registers
5 mitsubishi digital assp ? M66332FP facsimile image data processor when the sensor is cis: (creation and transmission of uniformity correction data) unif mode this mode is started when the unif command in register 0 is set to h. when the sensor is cis, if white cor- rection is started with the unif com- mand, peak detection (16 line periods) and uniformity white correc- tion data creation (8 line period) are started. to exit this operating mode, wait 30 line periods (at least 24 line periods) from the start and set the unif com- mand to l. data transfer the uniformity correction data per- taining to white correction created in unif mode are transferred to the backup memory. (read operation) agc mode peak detection is performed for 16 line periods when the agc com- mand in register 0 is set to h. to exit this operating mode, wait 20 line periods (at least 16 lines) from the start and set the agc command to l. scan mode the read operation is started by set- ting the scan command in register 0 to h. set the scan command to l to exit this operating mode. the signal functions and data flow in each mode are shown on pages 4C123 and 4C124. flowcharts are shown on pages 4-158 to 4C160. (1) operating mode the m66332 performs three basic operations. ? peak value detection: the peak value of the analog signal output from the analog signal processing circuits is matched to the white reference voltage (v wl ) of the m66332 internal a-d converter. (see also figs. 19 to 22 in the m66333fp document.) ? uniformity correction data creation: white reference data is created for sensor unit uniformity correction and written to the correction memory (sram: 304 words 5bits). ? read operation: a document is read and the image is pro- cessed to output bi-level data as serial or parallel output. these three basic operations are performed in the following sequence depending on whether the sensor is ccd or cis. the sensor is selected with register 0 (sens). when the sensor is ccd: unif mode operation is started by setting the unif command in register 0 to h. if the sensor is ccd, peak detection (16 line periods) and white unifor- mity correction data creation (8 line periods) are performed consecu- tively. to exit this operating mode, wait 30 line periods (at least 24 lines) from the start and set the unif command to l. scan mode the read operation is started by set- ting the scan command in register 0 to h. set the scan command to l to exit this operation mode. v v v
6 mitsubishi digital assp ? M66332FP facsimile image data processor operation during peak detection data flow in creation of uniformity correction data 5bit a-d converter (flash type) uniformity correction correction memory (304 words 5bits) sensor control analog control to each block image processing sequence control signal cut out/ scale down convert to bi-level dma control separation of image data area detection of image data area simple bi-level conversion/background and object level detection resolution compensation mpu bus interface collective dithering (16 levels) sram 16 words 4bits sysck image sensor analog signal processing circuit ptim sh ck1 ck2 rs clamp s/h agc dsch mpu d0~d7 codec stim sclk svid srdy dma drq dak v bl v ml1 v ml2 v ml3 v wl asig reset cs,rd wr a0~a3 5bit a-d converter (flash type) uniformity correction correction memory (304 words 5bits) sensor control analog control to each block image processing sequence control signal cut out/ scale down convert to bi-level dma control separation of image data area detection of image data area simple bi-level conversion/background and object level detection mtf compensation mpu bus interface collective dithering (16 levels) sram 16 words 4bits sysck image sensor analog signal processing circuit ptim sh ck1 ck2 rs clamp s/h agc dsch mpu d0~d7 codec stim sclk svid srdy dma drq dak v bl v ml1 v ml2 v ml3 v wl asig reset cs,rd wr a0~a3
7 mitsubishi digital assp ? M66332FP facsimile image data processor date flow during read operation (for serial output) date flow during read operation (for parallel output) 5bit a-d converter (flash type) uniformity correction correction memory (304 words 5bits) sensor control analog control to each block image processing sequence control signal cut out/ scale down convert to bi-level dma control separation of image data area detection of image data area simple bi-level conversion/ background and object level detection mtf compensation mpu bus interface collective dithering (16 levels) sram 16 words 4bits sysck image sensor analog signal processing circuit ptim sh ck1 ck2 rs clamp s/h agc dsch mpu d0~d7 codec stim sclk svid srdy dma drq dak v bl v ml1 v ml2 v ml3 v wl asig reset cs,rd wr a0~a3 : image date : correction, compensation date 5bit a-d converter (flash type) uniformity correction correction memory (304 words 5bits) sensor control analog control to each block image processing sequence control signal cut out/ scale down convert to bi-level dma control separation of image data area detection of image data area simple bi-level conversion/ background and object level detection resolution compensation mpu bus interface collective dithering (16 levels) sram 16 words 4bits sysck image sensor analog signal processing circuit ptim sh ck1 ck2 rs clamp s/h agc dsch mpu d0~d7 codec stim sclk svid srdy dma drq dak v bl v ml1 v ml2 v ml3 v wl asig reset cs,rd wr a0~a3 : image date : correction, compensation date
8 mitsubishi digital assp ? M66332FP facsimile image data processor (2) line period and read sequence figure 1 shows the relationship between the m66332 line period and the read sequence. ? 1 line period (1/acck): defines the processing time per line for m66332. the line period is determined from the line period counter registers 2 and 3 (pre_data) and pixel transmis- sion clock (adck). adck is 1/ 16th of sysck. 1 line period (1/acck) [ns] = line period counter pixel transmission clock period [ns] = (pre_data + 1) 1/adck [ns] = (pre_data + 1) 16/sysck [ns] the line period counter is counted down with the pixel transmission clock after loading the pre_data value and gener- ates the following addresses. ? sensor start pulse (sh): image sensor start pulse. the po- sition of the start pulse is deter- mined by the value in register 4 (st_pl) which is the offset from the uniformity correction range (unifg). set st_pl to the following values according to the type of image sensor. ccd: st_pl = sensor dummy pixel + 2 cis: st_pl =2 fig. 1 line period and read sequence line period (acck) pre_data load 0 down count relationship with registers registers 2, 3 (pre_data) register 0 (sens_w) register 4 (st_pl) register 0 (sens_w) register 0 (sens_w) register 1 (source) register 1 (source, dest , redu) 1 line period st_pl sensor start pulse (sh) uniformity correction range (unifg) agc range (agcg) source document read width pulse motor control (ptim) defines the uniformity correction range. this range corresponds to the sensor width (a3 to a4). refer to table 2 for the relation- ship between sensor width and uniformity correction range. defines the peak detection range. this range corresponds to the sensor width (a3 to a4). auto gain control is performed for the entire width (solid line) of the sensor in agc mode and for the range inside the sensor width (dotted line) in scan mode. refer to table 2 for the relation- ship between sensor width and agc range. defines the source document read width. if the document width is less than the sensor width, the document should be centered on the sensor because the read range is set from the center of the sensor. refer to table 3 for the relation- ship between sensor width and source document read width. generates the pulse motor con- trol signals for the read roller. ? uniformity correction range (unifg): ? agc range (agcg): ? source document read width: ? pulse motor control signal (ptim):
9 mitsubishi digital assp ? M66332FP facsimile image data processor fig. 2 codec interface and read sequence sensor width gate signal uniformity correction range (unifg) agc range (agcg) a3 2487/55 2487/55 2370/162 b4 2279/231 2279/231 2194/306 a4 2119/391 2119/391 1760/740 agc mode scan mode table 2 sensor width and gate signal range sensor width source document size a3 b4 a4 a3 2487/55 2278/230 2118/390 b4 CC 2278/230 2118/390 a4 CC CC 2118/390 table 3 source document read width according to sensor width and source document size 1 line period (acck) sensor start pulse (sh) srdy (sscan) (bcscan) stim sclk svid ptim : output bound (sscan), (bcscan) : internal signals xy x / y x : left end address y : right end address
10 mitsubishi digital assp ? M66332FP facsimile image data processor (3) image processing function the m66332 converts image signals from the image sensor to bi-level signals. bi-level conversion can be either simple bi- level conversion or pseudo half-tone conversion which con- verts image shades into bi-levels. the signal output from the image sensor must be corrected and compensated to reduce distortion and degradation be- fore it can be converted to bi-level signals. furthermore, for reduction in transmission time, separation of image data area and optimum bi-level conversion must be performed. the functions necessary for image processing are described below. ? peak detection ? uniformity correction ? mtf compensation ? background and object level detection (simple bi-level con- version) ? pseudo half-tone dither method ? separation of image data area ? image scale down/area specification ? peak detection the a-d converter of the m66332 is used with its reference voltages (v wl , v bl ) fixed. normallly, v wl is set to v cc and v bl , is set to 0v to keep the dynamic range of the a-d con- verter wide. peak detection must be performed for analog sig- nals to match them with the full scale value of the a-d converter before they are input to the a-d converter. peak detection is performed by reading white data in agc mode, one of the three m66332 operating modes (agc, unif, scan). in agc mode, 8-line period worth of dsch signal to raise gainfor gain controland 16-line period worth of agc sig- nal to lower gainfor the overflowing of the a-d converter are generated after agc command start (register 0: agc) as shown in fig. 3. this changes the gain as shown in fig. 4. fig. 3 peak detection fig. 4 changes in gain during peak detection v wl v bl peak detection (preprocessing) after completing peak detection preprocessing white data 1 line match the output level of the last pixel of a line to v wl v wl v bl peak gain control after completing gain control for peak value 1 line match the peak sensor output within a line tov wl 12345678910111213141516 peak detection line period dsch (raise gain) agc a-d converter overflow (lower gain) peak detection preprocessing peak gain control : generated in pixel units only when the a-d converter overflows.
11 mitsubishi digital assp ? M66332FP facsimile image data processor ? uniformity correction uniformity correction corrects the drop in lighting level at both ends of the light source, shading distortion due to drop in lighting level at the rim of the lens, and high frequency dis- tortion caused by the scattering of pixel-unit image sensor characteristic (see fig. 5). the m66332 creates uniformity correction data in unif mode, one of the three operating modes (agc, unif, scan), handling 8 pixels as a unit as shown in fig. 6. the created data is written to the internal correction memory (sram: 304 words 5 bits). in scan mode, the correction data is read from the internal correction memory to successively correct the input image data in pixel units. fig. 5 image sensor white data output waveform fig. 6 creation of uniformity correction data black level high frequency distortion shading distortion 1 line white level primary scanning direction pixels correction data 3 correction data 2 correction data 1 189161724 for pixels from 1 to 8, uniformity correction is carried out by correction data 1. for pixels from 9 to 16, uniformity correction is carried out by correction data 2. for pixels from 17 to 24, uniformity correction is carried out by correction data 3.
12 mitsubishi digital assp ? M66332FP facsimile image data processor ? correction the m66332 performs entire pixel correction for 50% correc- tion range as shown in fig. 7. correction is not possible if the white correction data exceeds the 50% correction range as shown in fig. 7. therefore, be sure to keep the input signal within the correction range. fig. 7 uniformity correction v wl v bl analog signal input 1 line 50% 2 5 -1 0 2 4 -1 white correction white data v wl v bl 1 line 50% 2 5 -1 0 2 4 -1 white data white data exceeding correction range portion exceeding correction range analog signal input if correction range is exceeded (during white correction)
13 mitsubishi digital assp ? M66332FP facsimile image data processor ? mtf compensation as shown in fig. 8, characters and photos that have been photoelectrically converted by the sensor unit are character- ized by a drop in resolution. the mtf compensation per- formed by the m66332 enhances the high frequency compo- nents with a laplacian filter to maintain the resolution of the image data and creates a perception of increased dynamic range. fig. 8 mtf compensation photoelectric conversion source document (characters) photoelectric conversion image signal mtf compensation compensated data source document (photo) photoelectric conversion image signal mtf compensation compensated data mtf compensation a = primary scanning compensation coefficient, register 5 (mtf) x = x + a ((x C a) + (x C b)) g f fCg axb secondary scanning direction window 3 1 primary scanning direction pixel
14 mitsubishi digital assp ? M66332FP facsimile image data processor ? background and object level detection the m66332 uses the floating threshold method rather than the fixed threshold method. this method successively gener- ates a threshold for optimum simple bi-level conversion of the target pixel. therefore, a threshold matching the picture data is generated without modifying the image data. this value is used as the threshold of the bi-level area when simple bi-level conversion or image separation is selected as bi-level conversion mode. : register 5 (mode) background level counter if an image data greater (brighter) than the current counter value is input, this counter is incremented to approach the image data. if an image data less (darker) than the current counter value is input, this counter is decremented to approach the image data. the count up/down speed can be set with the following regis- ter. : register 9 (max_up, max_down) the lower limit of the background level can be set with the fol- lowing register. : register b (ll_max) object level counter if an image data greater (brighter) than the current counter value is input, this counter is incremented to approach the image data. it an image data less (darker) than the current counter value is input, the image data is set to this counter. the count down speed can be set with the following register. : register 9 (min_up) the upper limit of the character level can be set with the fol- lowing register. : register a (ul_min) fig. 9 background-object level background level counter image data object level counter threshold generation compare input data object level threshold level background level white level black level threshold level = (background level - object level) k + object level k = bi-level threshold coefficient: register 9 (slice) bi-level data
15 mitsubishi digital assp ? M66332FP facsimile image data processor ? dither method the m66332 has a built-in 16 words 4 bits sram which is used as a collective dithering memory. during initialization, threshold values are written in the dither memory, matching the desired dither pattern into 4 4 dither matrix. : register e (dith_d) fig. 10 shows some examples of dither patterns. refer to the section on dither memory and uniformity correc- tion memory read/write for details on how to read/write the dither memory. this is used when dither method and image data area sepa- ration are selected for bi-level conversion mode during read. : register 5 (mode) fig. 10 collective dither pattern 0 12 3 15 8 4 11 7 2 14 1 13 10 6 9 5 4 15 9 14 10 0 3 8 5 2 1 13 11 6 12 7 11 12 7 3 4 0 8 15 6 2 10 13 9 14 5 1 a)diffused type b) concentrated type c) mesh type 4 4 matrix sequence table 4 scanning line density and dither matrix size primary/secondary scanning line (line/mm) 8 3.85 8 7.7 level CC 16 matrix size CC 4 4 scanning line density normal fine
16 mitsubishi digital assp ? M66332FP facsimile image data processor ? separation of image data area in order to perform bi-level conversion appropriate for the im- age, a black and white image is separated into bi-level con- version area and gradation conversion area. simple bi-level conversion is applied to the bi-level conversion area and dither method is applied to the gradation area. : register 5 (mode) fig. 11 separation of image data area black and white image bi-level conversion area gradation conversion area simple bi-level conversion pseudo half-tone conversion character photo there is no significant change in illumination of the gradation conversion area (photo) when a black and white image is viewed through a 5 1 window. this characteristic is used to distinguish between gradation conversion area and bi-level conversion area. x secondary scanning direction 5 1 window primary scanning direction pixel current line lmax: maximum illumination in window lmin : minimum illumination in window identification equation 1 lmax - lmin > a (illumination difference in bi-level conversion area) : register 6 difference (sepa_a) identification equation 2 lmin > b (area is entirely white) : register 7 minimum (sepa_b) identification equation 3 lmax < c (area is entirely black) : register 8 maximum (sepa_c) simple bi-level conversion if identification equation 1, 2, or 3 is satisfied. pseudo half-tone conversion when none of the identification equations is satisfied. white level = 15 input data black level = 0 a lmin lmax difference lmax C lmin minimum lmin b maximum lmax c area process example
17 mitsubishi digital assp ? M66332FP facsimile image data processor ? image scale down/area specification scale down function the image data input from the analog signal processing cir- cuit can be scaled down (a3 ? b4, a3 ? a4, b4 ? a4) by leaving out pixels in the primary scanning direction for bi-level conversion. : register 1 (source, dest, redu) scale down in secondary scanning direction can be per- formed in the same rate by mpu program. table 5 scaling rate in out b4 a4 a4 CC 1 b4 1 9/11 a3 13/15 12/17 area specification function when area specification is selected, bi-level conversion is performed only in the specified area from the center of the source document as shown fig. 12. : register 1 (source, dest, redu) fig. 12 cut out function source document width specified area
18 mitsubishi digital assp ? M66332FP facsimile image data processor (4) sensor unit/analog signal processing unit interface ccd-bit clamp type sh f 1 f 2 rs sensor output (os) s / h clamp out (asig) (a-d converter clock) register 4: stpl = dummy pixel + 2 dummy area valid pixel area when st_pl is odd ck1= f 1 ck2= f 2 when st_pl is even ck1= f 2 ck2= f 1 sh f 1 f 2 rs s / h clamp out (asig) (a-d converter clock) (os) 12 2 2 16 16 8 4 4 16 3 4 9 13 21 n n 16 unit: 1/sysck
19 mitsubishi digital assp ? M66332FP facsimile image data processor ccd-line clamp type note: line clamp uses sensor output equivalent to (dummy area C8) pixels from the first pixel after sh. sh f 1 f 2 rs sensor output (os) s / h clamp out (asig) (a-d converter clock) register 4: stpl = dummy pixel + 2 dummy area valid pixel area when st_pl is odd ck1= f 1 ck2= f 2 when st_pl is even ck1= f 2 ck2= f 1 (line clamp area) 8: fixed sh f 1 f 2 rs s / h clamp out (asig) (a-d converter clock) (os) 12 2 2 16 16 8 4 4 16 3 4 9 n n 3 4 9 unit: 1/sysck
20 mitsubishi digital assp ? M66332FP facsimile image data processor cis type note: clamp: in case of cis, check with the sensor manufacturer for the use of clamp. sh and ck1, ck2: sh can be selected with register 5 and ck can be selected with ck1 and ck2 (2 choices each) to provide interface with various types of cis. sh ck2 sensor output (sig) s / h clamp out (asig) (a-d converter clock) ck1 resister 4: stpl = 2 16 6 8 2 6 8 2 16 16 16 4 1 16 2 1 16 2 1 n n sh ck2 (sig) s / h clamp out (asig) (a-d converter clock) ck1 select with register 5 sh_w select with register 5 s/h_w unit: 1/sysck
21 mitsubishi digital assp ? M66332FP facsimile image data processor (5) codec interface serial output note: handshaking of three lines srdy, sh, and stim, which are interface to the codec, is the same as serial output. a b 10 4 12 10 2 2 srdy sh stim sclk svid unit: 1/sysck 1 2 3 4 5 6 7 8 n+8 n+7 n+6 n+5 n+4 n+3 n+2 n+1 n nC1 nC1 nC2 nC3 nC4 nC5 nC6 nC7 nC8 d7 d6 d5 d4 d3 d2 d1 d0 rd dak drq svid sclk pixel note: a is determined by register 4 (st_pl), and b is determined by register 1 (source, dest, redu). parallel output
22 mitsubishi digital assp ? M66332FP facsimile image data processor (6) read/write to dither memory, uniformity correction memory the following figures show the sequence for writing and read- ing dither patterns in the 16 words 4 bits collective dithering sram built in the m66332. dither memory write (mpu ? m66332) dither memory read (m66332 ? mpu) 4 4 matrix a3 a7 a11 a15 a2 a6 a10 a14 a1 a5 a9 a13 a0 a4 a8 a12 clear d4 (po) in register 1 to 0 in order to set the mpu bus (d7 C d0) to dither matrix memory data output mode. set d7 (reset) in register 0 to 1 in order to reset the dither memory address counter. select dith_d with register e and write data (0) on the mpu bus (d5 C d0). increment the address counter of the dither memory at the rising edge of wr. (during write) select dith_d with register e and read data (0) in dither memory to the mpu bus (d5 C d0). increment the address counter of the dither memory at the rising edge of rd. (during read) dither matrix address 1 h 0 h e h e h d4 = 0 d7 = 1 data (0) data (1) cs a3 ~ a0 wr d7 ~ d0 (input) initial setting (1) initial setting (2) memory address (0) memory address (1) 1 2 3 3 ~ 1 h 0 h e h e h data (0) data (1) cs a3 ~ a0 wr d7 ~ d0 (output) initial setting (1) initial setting (2) memory address (0) memory address (1) 1 2 4 4 ~ d4 = 0 d7 = 1 d7 ~ d0 (input) rd
23 mitsubishi digital assp ? M66332FP facsimile image data processor the m66332 can read/write uniformity correction data in the external correction sram through the mpu bus. this en- ables the uniformity correction data to be temporarily saved in backup memory during power off. the following figures show the uniformity correction data read/write sequence. uniformity correction memory write (mpu ? m66332) uniformity correction memory read (m66332 ? mpu) the last 5 digits (a4 C a0) of an address in the unif memory are written in regis- ter 2. the initial 4 digits (a8 C a5) of the address in the unif memory and d_load = 1 (d6) are written in register 3. steps and identifies the address in the unif memory. the unif memory is selected with register f, and data on the mpu bus (d4 C d0) is written at the identified address. the unif memory is selected with register f, and data stored at the identified ad- dress is read to the mpu bus (d4 C d0). initial setting: d7 (um_r/w) and d4 (p0) of register 1 are set to 1 and 0, re- spectively, to select read/write mode of uniformity correction memory. closing setting : d7 (um_r/w) of register 1 is set to 0 while d4 (p0) is set to that taken in operation, to cancel read/write mode of uniformity correc- tion memory. address space sensor width a3 b4 a4 left end address 310 284 264 right end address 7 29 49 3 h 2 h 1 h f h 1 h d7 = 1 data d7 = 0 d4 = 0 2 1 3 cs a3~a0 wr d7~d0 (input) initial setting closing setting repeats steps 1 , 2 , and 3 . d7 = 1 d4 = 0 3 h 2 h 1 h f h 1 h data d7 = 0 1 2 4 cs a3~a0 wr d7~d0 (output) initial setting closing setting repeats steps 1 , 2 , and 4 . d7~d0 (input) rd a4 a3 a2 a1 a0 0 0 0 d7 d6 d5 d4 d3 d2 d1 d0 r2 0 * 0 0 a8 a7 a6 a5 r3 d_load : 0 for normal 1 for data load a8~a0 : unif memory address
24 mitsubishi digital assp ? M66332FP facsimile image data processor dither memory write/read uniformity correction memory write/read 5bit a-d converter (flash type) uniformity correction correction memory (304 words 5bits) sensor control analog control to each block image processing sequence control signal cut out/ scale down convert to bi-level dma control separation of image data area detection of image data area simple bi-level conversion/ background and object level detection mtf compensation mpu bus interface collective dithering (16 levels) sram 16 words 4bits sysck image sensor analog signal processing circuit ptim sh ck1 ck2 rs clamp s/h agc dsch mpu d0~d7 codec stim sclk svid srdy dma drq dak v bl v ml1 v ml2 v ml3 v wl asig reset cs,rd wr a0~a3 : dither date 5bit a-d converter (flash type) uniformity correction correction memory (304 words 5bits) sensor control analog control to each block image processing sequence control signal cut out/ scale down convert to bi-level dma control separation of image data area detection of image data area simple bi-level conversion/ background and object level detection mtf compensation mpu bus interface collective dithering (16 levels) sram 16 words 4bits sysck image sensor analog signal processing circuit ptim sh ck1 ck2 rs clamp s/h agc dsch mpu d0~d7 codec stim sclk svid srdy dma drq dak v bl v ml1 v ml2 v ml3 v wl asig reset cs,rd wr a0~a3 : uniformity correction date
25 mitsubishi digital assp ? M66332FP facsimile image data processor (7) reset the m66332 has three types of reset. each reset function is described below. hard reset: initializes the circuit. hard reset also performs the following soft reset and standby reset. soft reset: used when cancelling a line read operation in the middle during scan mode. read operation is resumed starting from the next line. standby : used as standby mode. the internal clock is stopped by stopping the clock generator which generates the internal clock from the system clock. therefore, the internal circuit is stopped and power is saved. the period counter and register statuses are saved and the internal memory is placed in standby mode. table 6 reset function function reset type hard reset reset soft reset register 0 (reset) standby register1 (stnby) (8) image quality control using registers ? mtf compensation if the sensor has high resolution, resolution compensation need not be performed for half-tone area. mtf compensation should be performed for bi-level area regardless of the sensor resolution in order to achieve good object reproduction. ? simple bi-level conversion, background and object level detection set the background level detection and object level detec- tion counters as follows in order to obtain clear output of objects that do not have completely white background and that are not entirely black. ????? fast max_up > max_down > min_up the output becomes darker as bi-level conversion threshold coefficient (slice) is increased. select a large slice value for light source document. stop clock generator operation reset period counter initialize internal f/f initialize register ? pseudo half-tone conversion, dither method select collective dithering (16 gradations using 4 4 dither matrix) for fine mode. refer to the section on image pro- cessing function for details on providing dither pattern threshold. ? separation of image data area the optimum parameter is selected to perform the best bi- level conversion for each area: simple bi-level conversion for the object and pseudo half-tone conversion for half-tone. stop line read
26 mitsubishi digital assp ? M66332FP facsimile image data processor table 7 shows the recommended values for parameters re- lated to picture quality. use these values as reference to determine the optimum pa- rameter. separation of image data area sepa c CC CC 01 h sepa b CC CC 0d h sepa a CC CC 06 h dither pattern g correction no g =0.9 v ml1 =1.1v v ml2 =2.2v v ml3 =3.5v g =0.9 v ml1 =1.1v v ml2 =2.2v v ml3 =3.5v CC 4 4 diffusion pattern, g = 0.8 4 4 diffusion pattern, g = 0.8 ll max 0a h CC 0a h ul min 04 h CC 04 h min up nor- mal CC nor- mal background and object level max down nor- mal CC nor- mal max up nor- mal CC nor- mal resolution compensa- tion mtf 1/2 mon mon ye s ye s ye s uniformity correc- tion image simple bi-level conversion dithering separation of image data area 8 4 7 3 2 c 2 b 6 3 9 5 0 12 3 15 8 4 11 7 2 14 1 13 10 6 9 5 dither pattern ( g = 0.8) table 7 recommended parameter values 1 a 2 d slice 5/8 CC 5/8
27 mitsubishi digital assp ? M66332FP facsimile image data processor usage precautions ? peak detection in scan mode in scan mode, successive peak detection is performed for the image data being read as shown for the agc range (dotted line) in fig. 1. this enables better picture reproduction when picture data brighter than the white reference used during peak detec- tion is input in scan mode. this is especially effective for sensor units such as cis that do not have a built-in white reference. ? read operation with cis sensor if the sensor is cis, it is possible to select whether or not to use white correction in scan mode. do not select white correction for the input of analog signals already processed by entire pixel correction. ? collective dithering thresholds written in dither matrix should be between 1 and 15 excluding 0 as shown in fig. 13. as the m66332 carries out block correction in 8-bit units for uniformity correction, a cis sensor may generate back- ground noises due to irregularity of pixels. it is possible to remove noises and gain a fine image quality by reducing the maximum threshold value as shown in fig. 14. ? g correction g correction is performed to simulate the sensitivity charac- teristics (exponential nature) of the human eye in order to make the image data more similar to natural image. g = 0.45 is said to be the optimum correction when using a thermal head printer. the m66332, due to its capacity to handle 4-bit internal data, performs g correction by means of both collective dith- ering and the middle reference voltage pins (v ml1 , v ml2 , and v ml3 ) of the a-d converter. ( g correction by collective dithering ) g correction is realized applying a g characteristic to the threshold value to be written in the dither matrix as shown in fig. 15. the example given in fig. 15 is an approximation of g characteristic, g , to 0.8. ( g correction by the middle reference voltage pins of the a- d converter) the example shown in fig. 16 is an approximation of g char- acteristic, g , to 0.9 , which is carried out by applying v ml1 = 1.1v, v ml2 = 2.2v, and v ml3 = 3.5v to the middle reference pins of the a-d converter. fig. 23 in the M66332FP leaflet shows an example of cir- cuits for applying voltages to middle reference voltage pins. 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 0123456789101112131415 threshold threshold sequence 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 0123456789101112131415 threshold threshold sequence fig. 13 thresholds for collective dithering : example 1 fig. 14 thresholds for collective dithering : example 2 fig. 16 an example of g correction by middle reference pins fig. 15 an example of g correction by dither matrix 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 0123456789101112131415 threshold threshold sequence g ; 0.8 white 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 black 0 black = 0v v ml1 =1.1v v ml2 =2.2v v ml3 =3.5v white = 5v analog input a-d converter output g ; 0.9 g =1
28 mitsubishi digital assp ? M66332FP facsimile image data processor ? time function when time = 1 is set in register 1, the processing time per line is doubled to 2 line periods, data is read once every two line periods and processed. when the read and write motors operate simultaneously during copy operation, this command can be used to reduce the processing speed to 1/2 in order to reduce the power load. 1 line period (acck) sensor start pulse (sh) srdy stim ptim (sscan) (bcscan) n period n + 1 period n + 2 period note: (sscan), (bcscan): internal signals read is performed when (sscan) is h. data read is processed when (bcscan) is h. fig. 17 when processing speed is 1/2
29 mitsubishi digital assp ? M66332FP facsimile image data processor register structure explanation address 0 h r/w w d7 0 1 reset system reset normal mode reset mode d6 0 1 sens sensor type ccd cis reset while write pulse is l when d7 = 1 d5 0 0 1 1 sens_w sensor width a4 b4 a3 CC d4 0 1 0 1 d3 0 1 umode cis uniformity correction mode with white correction no white correction ? for selecting with or without correction in scanning (with cis only) d2 0 1 agc agc mode stop start ? controls agc mode start/stop. d1 0 1 unif unif mode stop start ? controls unif mode start/stop. d0 0 1 scan scan mode stop start ? controls scan mode start/stop. reset d7 sens d6 d5 d4 umode d3 agc d2 unif d1 scan d0 sens_w (default is 00 h )
30 mitsubishi digital assp ? M66332FP facsimile image data processor explanation address 1 h r/w w d7 0 1 um_r/w uniformity correction memory read/ write normal unif memory read/write d6 0 1 stnby standby mode normal standby mode um_r/w d7 stnby d6 d5 d4 d3 d2 dest d1 redu d0 source time p_0 (default is 00 h ) ? standby mode stops the clock generation circuit. the period counter and register status are saved and the internal memory is placed in standby mode. d5 0 1 time line time 1 line period 2 line period d4 0 1 p_o parallel output without parallel output parallel output ? when read and write operations are performed together as in copy operation, the power load can be reduced by selecting 2 line period. the processing speed drops to 1/2 when 2 line period is selected. ? d0 is output in lsb format and d7, in msb format. when scan data is output in scan mode, d7 is output in lsb (left) format and d0, is output in msb (right) format. d3 0 0 1 1 source source width a4 b4 a3 CC d2 0 1 0 1 d1 0 1 dest destination width a4 b4 d0 0 1 redu scale down/cut out cut out scale down scaling rate a3 ? b4 b4 ? a4 a3 ? a4 13/15 9/11 12/17 ? refer to image scale down/ area specification for scale down/cut out.
31 mitsubishi digital assp ? M66332FP facsimile image data processor 3 h w explanation address 2 h r/w w d7 d6 d5 d4 d3 d2 d1 d0 pre_data <7:0> (default is 00 h ) 0 d7 d_load d6 d5 d4 d3 d2 d1 d0 pre_data <12:8> (default is 00 h ) d4~d0: pre_data <12:8> pre data of line period counter (upper part) ? 1 line period is determined from pre_data and pixel transmission clock frequency (adck). adck is 1/16 of system clock. refer to line period and read sequence section. d3~d0: if register 3 d_load = 1 these bits will be the address denoted by upper 4 digits (a8~a5) used for read/write operations on the uniformity correction memory. d7~d0: pre_data <7:0> pre data of line period counter (lower part) d7~d4: if register 3 d_load = 1 these bits will be the address denoted by lower 5 digits (a4~a0) used for read/write operations on the uniformity correction memory. d6 0 1 uniformity correction memory address setting mode. normal data load ? this bit is for address setting for the access form mpu to the uniformity correction memory. set this bit to normal during access operation. d6~d0: st_pl <6:0> start pulse of line sensor st_pl = (sensor dummy pixel + 2) refer to line period and read sequence section. 4 h w d7 d6 d5 d4 d3 d2 d1 d0 st_pl <6:0> (default is 00 h )
32 mitsubishi digital assp ? M66332FP facsimile image data processor explanation address 5 h r/w w d7 d6 d5 d4 d3 d2 d1 d0 s/h_w (default is 00 h ) sh_w clamp mode mtf 7 h w 6 h w d6 0 1 s/h_w pulse width of s/h normal (sysclk period 4) normal 0.5 d5 0 1 sh_w sh pulse width for cis normal reverse of (normal 2) d4 0 1 clamp clamp method of analog circuit bit clamp line clamp ? in case of ccd, there is only one sh pulse width and this register is ignored. ? line clamp does not apply to cis. d3 0 0 1 1 mode bi-level mode simple bi-level dither separation (simple bi-level + dither) CC d2 0 1 0 1 d1 0 0 1 1 d0 0 1 0 1 mtf main coefficient of mtf compensation non(0) a little less (1/4) middle (1/2) a little over (1) d7 d6 d5 d4 d3 d2 d1 d0 (default is 00 h ) sepa_a d3~d0: sepa_a separation of image data area (difference) d7 d6 d5 d4 d3 d2 d1 d0 (default is 00 h ) sepa_b d3~d0: sepa_b separation of image data area (min.)
33 mitsubishi digital assp ? M66332FP facsimile image data processor explanation address 8 h r/w w d7 d6 d5 d4 d3 d2 d1 d0 (default is 00 h ) sepa_c d3~d0: sepa_c separation of image data area (max.) 9 h w d7 d6 d5 d4 d3 d2 d1 d0 slice (default is 00 h ) max_up max_down min_up d7 0 0 1 1 slice detector of background and object levels (slice) normal (4/8) light (3/8) dark (5/8) darker (6/8) d6 0 1 0 1 d5 0 0 1 1 max_up detector of background level (up counter clk) normal (t = (1 pixel period) 32) slow (t = (1 pixel period) 64) fast (t = (1 pixel period) 16) faster (t = (1 pixel period) 8) d4 0 1 0 1 d3 0 0 1 1 max_down detector of background level (down counter clk) normal (t = (1 pixel period) 128) slow (t = (1 pixel period) 256) fast (t = (1 pixel period) 64) faster (t = (1 pixel period) 32) d2 0 1 0 1 d1 0 0 1 1 min_up detector of object level (up counter clk) normal (t = (1 pixel period) 512) slow (t = (1 pixel period) 1024) fast (t = (1 pixel period) 256) faster (t = (1 pixel period) 128) d0 0 1 0 1 t: counter clock period
34 mitsubishi digital assp ? M66332FP facsimile image data processor explanation address a h r/w w b h w d3~d0: ll_max detector of background and object levels (lower limit of background level) d7 d6 d5 d4 d3 d2 d1 d0 (default is 07 h ) ll_max d3~d0: ul_min detector of background and object levels (upper limit of object level) d7 d6 d5 d4 d3 d2 d1 d0 (default is 06 h ) ul_min d7 d6 d5 d4 d3 d2 d1 d0 dith_d d3~d0: dith_d internal dither memory data ? refer to the section on dither memory and uniformity correction memory read/write for information concerning read/write method. d4~d0: unif_d internal uniformity correction data ? refer to the section on dither memory and uniformity correction memory read/write for information concerning read/write method. d7 d6 d5 d4 d3 d2 d1 d0 unif_d e h r/w f h r/w
35 mitsubishi digital assp ? M66332FP facsimile image data processor conditions ratings C0.3 ~ +7.0 C0.3 ~ v cc + 0.3 0 ~ v cc v cc C0.3 ~ v cc +0.3 C0.3 ~ av cc +0.3 C0.3 ~ av cc +0.3 C0.3 ~ av cc +0.3 C0.3 ~ av cc +0.3 C65 ~ 150 symbol v cc v i v o av cc v wl v bl v ml v ain t stg parameter supply voltage input voltage output voltage analog supply voltage reference voltage (white) reference voltage (black) reference voltage (middle) analog input voltage storage temperature range unit v v v v v v v v ?c absolute maximum rating (ta = C20 ~ 75?c, unless otherwise noted) recommended operating conditions symbol v cc gnd v i av cc a gnd v wl v bl v ain t opr parameter supply voltage gnd voltage input voltage analog supply voltage analog gnd voltage (note) reference voltage (white) reference voltage (black) analog input voltage asig operating temperature range limits min. 4.5 0.0 4.5 3 0.0 v bl C20 typ. 5.0 0.0 5.0 0.0 0.0 max. 5.5 v cc 5.5 av cc 1.0 v wl 75 unit v v v v v v v v ?c note: connect agnd with gnd externally.
36 mitsubishi digital assp ? M66332FP facsimile image data processor symbol v ih v il v t+ v tC v h v oh v ol v oh v ol i ih i il i ozh i ozl i ain r l s inl i ccs i cca i cc parameter h input voltage l input voltage positive-going threshold voltage negative-going threshold voltage hysteresis voltage h output voltage l output voltage h output voltage l output voltage h input current l input current off-state h output current off-state l output current analog input current reference resistance a-d converter non-linear error (note 1) quiescent supply current (standby) (note 2) quiescent supply current (active state) (note 2) dynamic supply current electrical characteristics (ta = C20 ~ 75?c, v cc = 5 v 10%, unless otherwise noted) max. 0.8 2.4 0.55 0.55 1.0 C1.0 5.0 C5.0 10 1.0 20 40 unit v v v v v v v v v m a m a m a m a m a k w lsb ma ma ma typ. 0.2 1.0 0.5 10 15 40 min. 2.0 0.6 v cc C0.8 v cc C0.8 limits test conditions i oh =C12ma i ol =12ma i oh =C4ma i ol =4ma v cc =5.5v v i =5.5v v cc =5.5v v i =0v v cc =5.5v v i =5.5v v cc =5.5v v i =0v v cc =5.0v v cc =5.5v v i =v cc , gnd v cc =5.5v v i =v cc , gnd v cc =5.5v v i =v cc ,gnd sysck, srdy, dak, cs, rd, wr, a0~a3, d0~d7 reset d0~d7 drq, sh, ck1, ck2, rs, ptim, clamp, s/h, agc, dsch, stim, sclk, svid sysck, srdy, dak,reset, cs, rd, wr, a0~a3 d0~d7 asig (standby) sysck=8mhz note 1: the a-d converter has a 5-bit resolution. 2: current flowing in the reference resistor in the a-d converter is not included.
37 mitsubishi digital assp ? M66332FP facsimile image data processor symbol t c(sys) t w+(sys) t wC(sys) t r(sys) t f(sys) t w(rd) t su(cs-rd) t su(a-rd) t su(dak-rd) t h(rd-cs) t h(rd-a) t h(rd-dak) t w(wr) t su(cs-wr) t su(a-wr) t su(d-wr) t h(wr-cs) t h(wr-a) t h(wr-d) t h(stim-srdy) timing requirements (ta = C20 ~ 75?c, v cc = 5 v 10%, unless otherwise noted) max. 20 20 unit ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns typ. 125 62.5 62.5 min. 100 20 20 20 10 10 10 100 20 20 50 20 10 0 0 limits test conditions period high-level pulse width low-level pulse width rise time fall time pulse width setup time setup time setup time hold time hold time hold time pulse width setup time setup time setup time hold time hold time hold time hold time cs a0~a3 dak cs a0~a3 dak cs a0~a3 d0~d7 cs a0~a3 d0~d7 srdy system clock read pulse write pulse stim parameter
38 mitsubishi digital assp ? M66332FP facsimile image data processor unit ns ns ns symbol t pzl(rd-d) t pzh(rd-d) t plz(rd-d) t phz(rd-d) t phl(rd-drq) parameter output enable time to low-level and high-level (rd-d) output disable time from low-level and high-level (rd-d) high-level to low-level output propagation time (rd-drq) switching characteristics (ta = C20 ~ 75?c, v cc = 5 v 10%, unless otherwise noted) max. 75 50 50 typ. min. 10 limits test circuit parameter t plh, t phl t plz t phz t pzl t pzh sw2 open open closed open closed sw1 open closed open closed open system clock t c(sys) t w+(sys) t w-(sys) 1.3v 1.3v sysck 1.3v t f(sys) 90% 10% t r(sys) 90% 10% 3v 0v output input v cc r l =1k w r l =1k w sw2 sw1 v cc c l 50 w gnd dut p.g (1)the pulse generator (pg) has the following characteristics (10%~90%) : t r = 3 ns, t f = 3 ns (2)the capacitance c l = 150pf includes stray wiring capacitance and the probe input capacitance. test conditions c l =150pf c l =50pf
39 mitsubishi digital assp ? M66332FP facsimile image data processor mpu interface (1) read timing (m66332 ? mpu) t su(cs-rd) 1.3v 1.3v t h(rd-cs) 1.3v 1.3v t h(rd-a) t w(rd) t su(a-rd) 1.3v 1.3v t pzl(rd-d) t plz(rd-d) 50% 10% 50% 90% t pzh(rd-d) t phz(rd-d) 3v 3v 0v 3v v oh 0v 0v v ol cs a0~a3 rd d0~d7 d0~d7 t su(cs-wr) 1.3v 1.3v t h(wr-cs) 1.3v 1.3v t h(wr-a) t w(wr) t su(a-wr) 1.3v 1.3v 3v 3v 0v 3v 0v 0v cs a0~a3 wr d0~d7 3v 0v 1.3v 1.3v t su(d-wr) t h(wr-d) valid data (2) write timing (mpu ? m66332)
40 mitsubishi digital assp ? M66332FP facsimile image data processor dma timing read timing (m66332 ? system bus) codec interface 1.3v 3v 0v srdy stim v oh v ol 50% sclk svid t h(stim-srdy) v oh v ol v oh v ol 120/sysck 50% 1.3v t phl(rd-drq) 50% 1.3v t h(rd-dak) t w(rd) t su(dak-rd) 1.3v 1.3v t pzl(rd-d) t plz(rd-d) 50% 10% 50% 90% t pzh(rd-d) t phz(rd-d) v oh 3v 0v 3v v oh v ol 0v v ol drq dak rd d0~d7 d0~d7
41 mitsubishi digital assp ? M66332FP facsimile image data processor flowchart read operation (sensor: ccd) start power on soft reset s/h, sh pulse width clamp output format period counter start pulse image data area separation parameter background and object level parameter dither pattern write reset standby mode start uniformity correction end uniformity correction document width, specification width bi-level mode image processing function start document read end document read complete? complete? transfer? a a end power off set standby mode power off? next document? register 0 register 5 register 1 registers 2, 3 register 4 registers 6, 7, 8 registers a, b register e register 1 register 0 register 0 register 1 register 5 registers 5, 9 register 0 agc + white compensation agc: 16 times white compensation : 8 times wait 1 line period 30 (24 or more) next document turn on light stabilize (white reference) register 0 y n n y n y initial settings white compensation setting for each document read a page next document y n n y register 1 wait 1 line period 2 turn off light
42 mitsubishi digital assp ? M66332FP facsimile image data processor uniformity correction data creation, transmission (sensor: cis) start power on soft reset s/h, sh pulse width clamp output format period counter start pulse end uniformity correction start uniformity correction complete? n y end transmit correction data register 0 register 5 register 1 registers 2, 3 register 4 initial settings register 0 uniformity correction /white transmit uniformity correction data agc + white compensation agc: 16 times white compensation : 8 times wait 1 line period 30 (24 or more) turn on light stabilize register 0 registers 1, 2, 3, f turn off light
43 mitsubishi digital assp ? M66332FP facsimile image data processor read operation (sensor: cis) start power on soft reset s/h, sh pulse width clamp output format period counter start pulse image data area separation parameter background and object level parameter dither pattern write write to uniformity correction memory start agc end agc document width, specification width bi-level mode image processing function start document read end document read complete? complete? transmit? a a end power off set standby mode power off? next document? register 0 register 5 register 1 registers 2, 3 register 4 registers 6, 7, 8 registers a, b register e register f register 0 register 0 register 1 register 5 registers 5, 9 register 0 wait 1 line period x 20 (16 or more) register 0 y n n y n y initial settings agc settings for each document read a page next document y n n y register 1 wait 1 line period 2 turn off light reset standby mode register 1 complete? n y turn on light stabilize (white reference) next document (agc: 16 times)

▲Up To Search▲

Price & Availability of M66332FP

	To Download M66332FP Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .