Part Number Hot Search : 
RT9273 TP868C12 HT86AR72 47000 91Z3P 5233B TL4810 100LS
Product Description
Full Text Search
 

To Download ML6440CH Datasheet File

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


  Datasheet File OCR Text:
  july 1999 preliminary ml6440 * multi-standard 8-bit adaptive digital input comb filter 1 general description the ml6440 smartcomb? is a single-chip, 8-bit digital comb filter that incorporates line delays and adaption algorithms for ntsc and pal at both ccir601 and square pixel rates. the ml6440 implements bandsplit filters and a proprietary adaption and decision logic block that allows for optimum combing over a wide range of video sources. the ml6440 contains all the necessary circuitry to provide high quality combed output of luminance and chrominance in the y/c format. internal filters with integrated adaption and compensation circuits provide filtered outputs with optimal video bandwidth and resolution while suppressing cross-color (rainbow), cross- luminance (dot crawl) and other corrupting artifacts that can reduce video compression efficiency. block diagram features n smartcomb algorithm for automatic or manual selection adaption thresholds n 3-line comb with 60+ line frame adjust n comb/notch thresholds set dynamically over 60+ lines automatically n 12-bit processing minimizes truncation errors and maintains signal-to-noise performance n optional 8-bit composite or separated y/c digital inputs n applications: digital tv, line doubler, imaging n separate comb/notch filter thresholds for luma and chroma channels n optional pin controls or two-wire serial control interface n operating power dissipation less than 700mw n no external components, except diode and caps * this part is end of life as of august 1, 2000 ci2/di2 ci3/di3 ci4/di4 ci5/di5 ci6/di6 ci7/di7 s clk s data s /p smartcomb? comb adaption and decision logic bandsplit filters (notch/delay) 39 16 v cc av cc v cc 5 c7 c6 c5 c4 c3 40 41 42 43 44 c2 c1 c0 1 2 3 y7 y6 y5 y4 y3 30 31 32 33 34 y2 y1 y0 35 36 37 oe 27 29 28 26 gnd gnd gnd 38 17 4 10 11 12 ci0/di0 ci1/di1 8 9 clk rst 7 6 13 14 15 yi2/cv2 yi3/cv3 yi4/cv4 yi5/cv5 yi6/cv6 yi7/cv7 20 21 22 yi0/cv0 yi1/cv1 18 19 23 24 25 chrominance output buffer luminance output buffer filter and comb control clock buffer 1hdl reset logic 1hdl high band low band high band low band total band chroma bypass high band low band ccir601/sq. pixel mode comb mode 0 comb mode 1 threshold control 0 threshold control 1 threshold control 2 y+c/cv input mode pal/ntsc mode luma bypass digital digital digital control mux d2 d3 d4 d5 d6 d7 d0 d1
ml6440 2 pin configuration top view ml6440 44-pin tqfp (h44-14) y4 y5 y6 y7 s clk s data oe s /p yi7/cv7 yi6/cv6 yi5/cv5 c2 c1 c0 gnd v cc rst clk ci0/di0 ci1/di1 ci2/di2 ci3/di3 12 13 14 15 16 c3 c4 c5 c6 c7 v cc gnd y0 y1 y2 y3 ci4/di4 ci5/di5 ci6/di6 ci7/di7 av cc gnd yi0/cv0 yi1/cv1 yi2/cv2 yi3/cv3 yi4/cv4 17 18 1 2 3 4 5 6 7 8 9 10 11 44 43 42 41 40 33 32 31 30 29 28 27 26 25 24 23 39 38 19 37 20 36 21 35 22 34 1 2 3 4 5 6 7 8 9 10 11 44 43 42 41 40 33 32 31 30 29 28 27 26 25 24 23 39 38 37 36 35 34 12 13 14 15 16 17 18 19 20 21 22
ml6440 3 pin name function 1 c2 chrominance output 2 c1 chrominance output 3 c0 chrominance output 4 gnd digital ground pin 5v cc digital supply pin 6 rst reset input active low. resets comb logic including the internal data register. required at power up. 7 clk ttl compatible clock reference 8 ci0/di0 (lsb) input chrominance signal (pal/ ntsc control pin in control pin mode: register bit d0) 9 ci1/di1 input chrominance signal (square pixel/ccir control pin in control pin mode: register bit d1) 10 ci2/di2 input chrominance signal. (comb mode 0 control pin in control pin mode: register bit d2) 11 ci3/di3 input chrominance signal (comb mode 1 control pin in control pin mode: register bit d3) 12 ci4/di4 input chrominance signal (adaption threshold 0 control pin in control pin mode: register bit d4) 13 ci5/di5 input chrominance signal (adaption threshold 1 control pin mode: register bit d5) 14 ci6/di6 input chrominance signal (adaption threshold 2 control pin mode: register bit d6) 15 ci7/di7 (msb) input chrominance (y+c/yi control pin in control pin mode: register bit d7) 16 av cc analog supply pin. bypass to ground with 1f ceramic capacitor 17 gnd ground pin for analog delay line 18 yi0/cv0 ttl compatible (lsb) input composite video signal or y in the y+c bypass mode 19 yi1/cv1 luma or composite video input signal pin description pin name function 20 yi2/cv2 luma or composite video input signal 21 yi3/cv3 luma or composite video input signal 22 yi4/cv4 luma or composite video input signal 23 yi5/cv5 luma or composite video input signal 24 yi6/cv6 luma or composite video input signal 25 yi7/cv7 luma or composite video (msb) input signal 26 s /p serial/parallel program mode. if high, allows 8-bit parallel control using the eight digital chrominance input pins. data clocks in on the positive edge transition. if low, serial port active. 27 oe output enable. (y[7:0] and c[7:0]) if low, outputs high impedance. 28 s data serial data input 29 s clk serial clock input. positive-edge clocks. 30 y7 ttl compatible luminance output (msb) 31 y6 luminance output 32 y5 luminance output 33 y4 luminance output 34 y3 luminance output 35 y2 luminance output 36 y1 luminance output 37 y0 luminance output (lsb) 38 gnd digital ground pin 39 v cc digital supply pin 40 c7 ttl compatible chrominance output (lsb) 41 c6 chrominance output 42 c5 chrominance output 43 c4 chrominance output 44 c3 chrominance output
ml6440 4 electrical characteristics unless otherwise specified, v cc = 5v 5%, c l = 50pf, t a = operating temperature range (notes 1, 2) symbol parameter conditions min typ max units supply i cc supply current v cc =5.25v, clk=14.75 mhz, 60 90 ma pal square pixel av cc analog supply voltage recommend operation v cc C0.6 v cc C0.4 v logic v il low level input voltage 0.8 v v ih high level input voltage v cc C 1.5 v i il low level input current 10 a i ih high level input current 10 a c in input capacitance 5pf v ol low level output voltage i ol = C2ma 0.4 v v oh high level output voltage i oh = 2ma v cc C 1.0 v output current 3-state mode 10 a c out output capacitance 3-state mode 5 pf system timing f clk clk input frequency square pixel pal 14.75 mhz square pixel ntsc 12.70 mhz ccir601 13.50 mhz t su setup time to rising clk edge f clk = 14.75mhz 10 ns clock low duration f clk = 14.75mhz 30 ns clock high duration f clk = 14.75mhz 45 ns input rise time f clk = 14.75mhz, 10% to 90% 20 ns input fall time f clk = 14.75mhz, 90% to 10% 20 ns data valid after rising clk edge f clk = 14.75mhz 20 ns 3-state delay time, output enable f clk = 14.75mhz 20 ns 3-state delay time, output disable f clk = 14.75mhz 20 ns output rise time f clk = 14.75mhz, 10% to 90% 20 ns output fall time f clk = 14.75mhz, 90% to 10% 20 ns absolute maximum ratings absolute maximum ratings are those values beyond which the device could be permanently damaged. absolute maximum ratings are stress ratings only and functional device operation is not implied. v cc ............................................................................. 7v analog & digital i/o ............. gnd C 0.3v to v cc + 0.3v input current .......................................................... 20a junction temperature ............................................. 150c storage temperature range...................... C65c to 150c lead temperature (soldering, 10 sec) ..................... 260c thermal resistance ( q ja ) ...................................... 67c/w operating conditions temperature range ........................................ 0c to 70c v cc range ............................................... 4.75v to 5.25v
ml6440 5 electrical characteristics (continued) symbol parameter conditions min typ max units video signal processing effective luminance bandwidth ntsc/pal, comb on at C3db (note 3) ntsc/ccir601, comb off at C3db 2.5 mhz ntsc/sq. pixel , comb off at C3db 2.5 mhz pal/ccir601 , comb off at C3db 2.5 mhz pal /sq. pixel, comb off at C3db 2.5 mhz effective chrominance bandwidth ntsc/pal, comb on at C3db (note 3) centered at f sc ntsc/ccir601, comb off at C3db 1.0 mhz ntsc/sq. pixel, comb off at C3db 1.0 mhz pal/ccir601, comb off at C3db 1.0 mhz pal /sq. pixel, comb off at C3db 1.0 mhz comb filter characteristics snr c signal to noise ratio, chrominance spurious luma artifact ntsc 48 db 13.5mhz pal 45 db 12.27mhz snr l signal to noise ratio, luminance spurious subcarrier ntsc 48 db artifact 13.5mhz pal 45 db 12.27mhz comb notch depth at f sc ntsc/pal C35 db comb notch bandwidth at C30db 500 khz
ml6440 6 electrical characteristics (continued) symbol parameter conditions min typ max units serial bus logic input low level input voltage 0 0.8 v high level input voltage v cc C 0.8 v cc v low level input current v in = 0v 1.0 a high level input current v in = v cc d 1.0 a input impedance f clk = 100khz 1 m w input capacitance (c in )2pf system timing s clk frequency (f clock ) 100 khz input hysteresis (v hys ) 0.2 v spike suppression (t spike ) max length for zero response 50 ns power setup time to valid data inputs vcc settled to within 1% 10 ms wait time from stop to start on s data (t wait ) 1.3 s hold time for start on s data (t hd/start ) 0.6 s setup time for start on s data (t su/start ) 0.6 s min low time on s clk (t low ) 1.3 s min high time on s clk (t hi ) 0.6 s hold time on s data (t hd/data ) 5.0 s setup time on (t su/data ) fast mode (note 4) 100 ns slow mode (note 4) 250 ns rise time for s clk & s data (t lh ) 30 300 ns fall time for s clk & s data (t hl ) 30 300 ns setup time for stop on s data (t su/stop ) 0.6 s note 1: limits are guaranteed by 100% testing, sampling, or correlation with worst case test conditions. note 2: supply voltage fed through 7.5 w resistor to all v cc pins. note 3: no bandlimiting is performed on the signal bandwidth when the comb is in the on state. note 4: parameter is luma dependent
ml6440 7 functional description the ml6440 is an adaptive digital comb filter that offers automatic self-adjusting or manual thresholds to handle vertical and horizontal contouring during y/c separation. the ml6440 implements a smartcomb algorithm where full bandwidth video and artifact suppression techniques are used to automatically select suppression thresholds for all video content. video artifacts such as hanging dots, dot crawl, and cross luminance are dramatically reduced. the algorithm functionally performs a statistical analysis of the video content of over 60 lines. it then calculates and selects the best threshold for artifact reduction at 12-bit resolution. the ml6440 is a comb for ntsc (525 lines) and pal (625 lines) with the ability to handle ccir601 and square pixel sampling rates. ccir601 is the existing standard for digitized video of either ntsc or pal at a 13.5mhz rate, and is used by mpeg encoders and decoders for image compression. the comb algorithms in the comb filter substantially reduce rainbow and dot-crawl artifacts which interfere with efficient image compression. square pixel is video digitized at 12.27mhz for ntsc, and 14.75mhz for pal, so that each sample corresponds to a single pixel in computer displays without the effort or expense of a multi-phase image resizing filter done in hardware or software. using the ml6440 comb filter requires no external components for operation. the comb filter can be bypassed to accommodate selection of external s-video signals. the comb filter is controllable via parallel pin or serial interface. smartcomb operation the user supplies 8-bit video as either digital y/c or digital composite video inputs and receives delayed and combed data at the y/c digital outputs. use of y/c input assumes that combing is unnecessary and sends the data through the same delay as the composite video path to prevent shifting of the video position on the screen. this makes the ml6440 useful in source selection such as s-video inputs and composite video. composite video (cv), applied to the ml6440 with comb operation enabled, permits the separation of luminance and chrominance with minimal artifacts. the smartcomb controls the application of the comb filter and bandsplit-notch filters on a pixel-by-pixel basis. the digital video image is sampled, and an analysis is performed. smartcomb step 1: looks for differences in the low frequency spectra (vertical detail). smartcomb step 2: searches for changes in phase angle/saturation of the chroma sub-carrier, which represents changes in color in the vertical direction. finally, smartcomb step 3: an averaging circuit is used as "smoothing function" to eliminate dithering in the horizontal dimension caused by excessive filter switching. figure 1 provides a simplified block diagram of the smartcomb architecture. this architecture implements several complex computations to determine the amount of notch vs. comb filtering. three 8-bit detectors are used to examine vertical and horizontal detail for luma and chroma signals on a pixel-by-pixel basis for 3-line analysis. these data are then fed into a weighting function to determine the best filtering approach over the existing 3-lines of video. furthermore, this data is then compared figure 1. smartcomb architecture: simplified block diagram c out y out cv in 1h 1h hi-band low band hi-band low band hi-band low band luma vertical detail detector 8-bit chroma vertical detail detector 8-bit luma horizontal detail detector 8-bit comb/notch control function 24-bit, 3-line comb/notch threshold select comb/notch control function 12-bit, frame, 60-line manual threshold select mixer 8-bit fader comp automatic threshold select video signals
ml6440 8 bit d7 d6 d5 d4 d3 d2 d1 d0 modefunction y+c or cv adaption adaption adaption comb comb square pixel/ pal select threshold threshold threshold mode mode ccir601 or ntsc 2 1 0 1 0 select select value y+c = 1 at2 = 0 at1 = 0 at0 = 0 cm1 = 0 cm0 = 0 sq. pixel = 1 pal = 1 cv = 0 or 1 or 1 or 1 or 1 or 1 ccir601 = 0 ntsc = 0 parallel mode ci7/di7 ci6/di6 ci5/di5 ci4/di4 ci3/di3 ci2/di2 ci1/di1 ci0/di0 input pin preamble data address (b3) hex data (d[7:0] 2 ) table 1. serial data format table 2. control register format d[7:0] (serial or parallel mode) table 3. clock frequency requirements standard/rate ccir601 square pixel ntsc (f sc = 3.58mhz) 13.5mhz 12.27mhz pal (f sc = 4.43mhz) 13.5mhz 14.75mhz comb mode, d[3,2] ntsc pal 00 y adaptive, c comb 100% active y adaptive, c adaptive 01 y adaptive, c adaptive y adaptive, c comb 100% active 10 y 100% comb active, c 100% comb active y 100% comb active, c 100% comb active not recommended due reduced performance 11 bandsplitting only (no combing) bandsplitting only (no combing) table 4. adaption threshold bit table adaption threshold d[6,5,4] adaption threshold level 000 automatic threshold for minimal comb artifact generation for various video source material. 001 threshold for computer graphics applications, minimizing dot crawl on bold lettering. 010 to 110 various degrees of thresholds. 111 threshold for better reproduction of natural photographic images. table 5. comb mode bit table
ml6440 9 figure 4. typical serial bus command figure 2. definition of start & stop on serial data bus figure 3. definition of data format on serial data bus s data s clk start stop t rise t fall t set/start all other s data transitions must occur while s clk is low start: a falling edge on the s data while s clk is held high stop: a rising edge on the s data while s clk is held high s data stop msb msb a1 a0 a6 a7 0 1 2 7 8 9 10 11 16 17 18 d7 d6 d1 d0 s clk: s clk: s data: s clk: s clk: s data: s clk: s data: 9th pulse strobes address decoder rising edge enables data transfer value set to a6, device address (msb-1) falling edge disables data transfer rising edge enables data transfer value set to a7, device address msb falling edge in prep for first address transfer falling edge w/s clk high means start of sequence s data: s data: s clk: s clk: s data: s clk: s clk: s data: rising edge with s clk hi = stop value set low in prep for stop 18th pulse strobes data shift register rising edge enables data transfer value set to d6, data msb-1 falling edge disables data transfer rising edge enables data transfer value set to d7, data msb s clk s data s clk 1 0 0 1 ? ? d7 d6 d5 d4 d3 d2 d1 d0 0 1 1 1 0 1 234 5 9abcd e fgh i s clk: s data: s clk: address decode strobed on 9th clock [data is dont care during strobe] 1011 0011 shifted on next 8 clocks falling edge in prep for device address transfer s data : control register, load msb first final clock strobes data into register device addr = b3 control register strobe (ack) 68 7
ml6440 10 to the video content of the last 60+ lines (frames) of video a frame weighing function to determine the best filtering approach of the current video content. finally, 1 of 7 thresholds can be chosen to adjust either real-time (automatically) by the ml6440 itself or fixed manually by design. an internal 12-bit analysis is used for the filtering algorithm. an additional fader circuit is included to "smooth" the transitions between comb and notch filtering. in ntsc, every other line of chroma is inverted while there are no changes in luma (assuming there is no change in color between adjacent lines). this permits data to be added from the first stored line of data with the third line of active high-band video data, each with a scaling of 0.25, and add the second stored line of data with a scaling of 0.50. the result contains no chroma, only luma information. subtracting this information from the high- band signal yields only chroma information. in pal, each line of data rotates the phase of the chroma by 90 degrees. therefore, by adding the first line of stored high-band information from the active third line, the chroma is canceled, leaving only luma information. once again, this information is subtracted with the appropriate amplitude scaling from the original highband information to yield only chroma. in bandsplit (notch) filtering and luma peaking, the digital bandsplit filter is used as an alternate filter when the smartcomb algorithm determines that the comb filter output signal would be contaminated with cross chroma artifacts. to maintain a reasonable amount of apparent resolution in the luminance signal, luma peaking is applied. this helps maintain image quality in the portion of the video image where comb filtering would generate artifacts. the ml6440 smartcomb provides a real-time filtering algorithm solution (automatic mode) for a wide variety of video sources without the shortcomings of using fixed thresholds (manual mode). the automatic threshold tracking circuit responds quickly, adjusting itself in about 1/300th of a second (about 60 lines). the transition from the comb to the digital bandsplit is made with a continuous digital fader to prevents switching glitches. the bandsplit-notch and comb filters were designed to reduce color bar peak-to-peak dot edge crawl to <7.5% for 75% color bars. peak-to-peak cross color on the white bar edge was designed to be <1% with 75% color bars. serial program mode in serial program mode, the comb filter can be controlled with the serial interface. control inputs may be clocked in serially with s data and s clk , when s /p is low. serial bus transfers require an 8-bit address followed by 8-bits of data for the internal data register. the pattern is described in tables 1 through 5. the serial bus control in the ml6440 has one level of addressing followed by control register programming. figure 2 shows the physical waveforms generated in order to address the ml6440. there are six basic parts of the waveform: 1. start indication: clock cycle 0 2. device address shifted in: clock cycle 1 thru 8 3. device address strobed /decoded: clock cycle 9 4. data shifted in: clock cycle 10 thru 17 5. data strobed to appropriate register: clock cycle 18 6. stop indication: clock cycle 19 parallel program mode in parallel program mode, the ci/di[7:0] pins act as programming pins for the internal 8-bit register (d[7:0]) such that ci7/di7=d7, ci6/di6=d6, ... ci0/di0=d0. in the parallel mode, the s /p serves as a rising edge-triggered strobe pin for the internal registers of d[7:0]. when s /p goes from low to high, data applied to the inputs (ci/ di[7:0]), is loaded into the internal 8-bit register d[7:0] directly. a transition on the s /p pin back to low will reactivate the serial mode. if the s /p is held at a high, the serial mode is inactive. control register, d[7:0] the internal register, d[7:0], controls the following functions within the ml6440. again, this register can be programmed through either a serial or parallel interface. the rst acts as a reset pin, forcing all the bits of the functional description (continued) figure 5: reset timing power on at least 5s high at least 1s low rst
ml6440 11 internal data register low (d[7:0]=0) which defaults to automatic combing of ntsc/ccir601 composite video. the y+c/cv selection bit, d7, controls whether y+c or cv format is the source for the comb filter. setting this bit low selects the composite input pins, which is combed by default. setting the bit high selects the y+c inputs as the source for video (see bypass section). this source is not combed and the y+c inputs are not activated until the s /p pin is rising-edge triggered. adaption thresholds 2 through 0 are d6 through d4, respectively. these set the adaption behavior from automatic (d[6:4]=000) to one of seven settings ranging from fine tuning of computer graphics to photographic images. see table 4. comb modes 1 and 0, (d3 and d2, respectively) are used to enable and disable adaption separately on y and c data. for ntsc, if d[3:2]=<00>, then y is adaptive and c is always combed. for ntsc, if d[3:2]=<01>, both y and c are adaptive. for pal, if d[3:2]=<00>, then both y and c are adaptive. for pal, if d[3:2]=<01>, then y is adaptive and c is combed. for ntsc and pal, if d[3:2]=<10>, then y and c are forced to comb mode. for ntsc and pal, if d[3:2]=<11>, then both y and c are in bandsplit filtering mode only. see table 5. the d1 bit sets the choice of data rates between ccir601 and square pixel. ccir601 is active low. square pixel is active high. the last bit, d0, select the standard, ntsc or pal. ntsc is active low. pal is active high. see table 2. the remaining pin function controls are independent of the data register: clk, rst , and oe. the clk pin requires a clock at the rates listed in table 2. its duty cycle must meet the duration minimum for high and low. the rst pin resets the comb logic including the internal data register on active low but does not clear the line delays. this part requires a timed reset pulse. on power up or at any time a return to automatic combing of ntsc/ ccir601 composite video is required the reset pin must have been high for at least 5 micro seconds. then taken low for at least one micro second before returning high again to resume normal operation. neither pulse width nor rise and fall times are critical. in most applications a microprocessor will be available to provide timing and control waveforms. it will provide this reset pulse and if needed set the appropriate comb filter mode. in non processor systems an r c network can provide this reset pulse. see figure 5. chroma and luma bypass operation this mode can be activated by setting the internal bit d7 = 1 either through the serial or parallel programming modes. once the register is set (d7 = 1), the comb filter will look at the inputs for luma on the cv[7:0] pins for a luma bypass to the outputs of the y[7:0] pins respectively; as well as inputs for chroma on the ci/di[7:0] pins for a chroma bypass to the outputs of the c[7:0] pins respectively. the comb filter will continue to look at the inputs as long as d7 is set high. data presented to the inputs (ci/di[7:0]) is not interrupted in the color processing path when s /p is held high or low. therefore, it is recommended that s /p activity be limited to sync or blanking intervals in the video to avoid unwanted visual artifacts during register programming. clock timing figure 6 depicts digital video input and output timing for valid data. functional description (continued) y or c data 60% 20ns 15ns 0ns y or c data y/c output cv input 13.5mhz clock cv data 40% cv data figure 6: clock timing
ml6440 12 typical applications figure 7 depicts an application block diagram of the ml6440. the ml6440 is easy to use in most standard video applications in conjunction with an 8-bit a/d converter or digital data for analog and digital applications. high speed digital layout should be observed. pay special attention that the outputs are not loaded beyond a 2ma load. an a/d, if used, should have at least seven effective bits at nyquist rates for good quality video. please note that startup of the part will require at least two lines to flush out the line delays and about 10 lines for the adaption to adjust. the automatic adaption thresholding should not be discernible in most video applications, but can be defeated with the fixed adaption levels. in applications using video tape recorders the comb may not be effective and can make results worse if time jitter exceeds the clock period 70ns. jitter of this magnitude would misalign the line delays, making the combing and adaption incorrect. in this case combing should be turned off. power supply requirements the av cc must be fed through a schottky diode of about 0.4v drop from a 5v supply. alternately, a 7.5 w resistor can be used instead of the diode. the av cc pin also needs a 1f surface mount capacitor to ground. see figure 7. bypass caps must be within 3mm of the chip pins they bypass. direct traces are necessary. digital video values of sync tip and peak white as recommended by smpte and itu-r bt.601, the minimum value of digitized sync tip is decimal 16. the maximum value of digitized video is decimal 255 (235 recommended is by smpte which allows for overshoots in video). normally, peak white is at 235; this avoids clipping due to overshoots. figure 8 presents micro linears entire solution for optimized video application designs. the ml6430 can be used as high performance line lock clock thats capable of generating horizontal and vertical syncs and a variety of system clocks using common reference frequency crystals. the ml6401 is a low cost a/d converter. the ml6421 is a three channel filter with sinx/x equalization. the ml6424 is a precision filter used as a clamp circuit and anti- aliasing filter. (see micro linear's ml6440eval for more information) test patterns when viewing test patterns, the fixed thresholds may give better performance, since the automatic adaptive threshold algorithm was optimized for live video. there are various specific application uses of the ml6440 smartcomb : ? large screen tvs ? hdtv ? video projection ? mpeg encoding ? video conferencing ? imaging and video capture ? format converters ? time base correctors ? professional video ? line doublers and quadrapulers figure 9 illustrates the use of the ml6440 in a big screen tv application. in this application, the ml6440 offers automatic self-adjusting thresholds to optimize y/c separation while minimizing artifacts for a wide variety of live video material without fixed threshold limitations. the ml6440 supports s-video bypass or luma and chroma separation enhancement, simplifying design for high- performance video source selection. figure 10 illustrates the use of the ml6440 in hdtv and ntsc application. in hybrid tv sets, where both ntsc and hdtv will be prevalent, the ml6440 smartcomb can serve as a high-performance y/c ntsc-separator in a ntsc-hdtv video receiver.
ml6440 13 typical applications (continued) figure 7. typical application circuit 1 2 3 4 5 6 7 8 9 10 11 44 43 42 41 40 33 32 31 30 29 28 27 26 25 24 23 39 38 37 36 35 34 12 13 14 15 16 17 18 19 20 21 22 parallel or chroma bypass input 5v reset genlock, system clock cv in serial/parallel ( s /p) output enable (oe) serial data in (s data) serial clock (s clock) a/d clamp filter 8 8 filter filter d/a d/a chroma luma ml6440 smartcomb 1f 1f 5v 5v 1f * * alternative power supply 1f 7.5 w pin 16 8
ml6440 14 figure 8. micro linear application solution ml6421 ml6424 filter clock y c cv 8 composite video ml6401 a/d d/a d/a d/a filter filter filter ml6431 genlock 8 8 8 s video luma chroma composite video ml6440 comb filter figure 9. application diagram: hdtv and ntsc television figure 10. application diagram: high-end television/monitor to display tuner hdtv detector ntsc detector genlock ml6440 smartcomb chrominance demodulator hdtv pll mux matrix digital mpeg decoder if yuv yuv to crt serial bus control genlock ml6440 smartcomb mux a/d a/d chrominance demodulator matrix cv in /y in c in luma r g b chroma u v composite video in composite video from tv tuner s video luma s video chroma typical applications (continued)
ml6440 15 performance data comb filter performance ? unrestricted bandwidth. no band-limiting is performed on the signal when the comb is on hence preventing a signal loss. ? high frequency peaking allows for gentle transitions from comb to notch modes. ? spurious sub-carrier supression by 48db removes color sub-carrier which otherwise can cause large area dot crawl notch filter performance ? when invoked by adaption threshold logic, hanging dot suppression is 35db. adaptive threshold performance ? seven fixed values to select threshold, the 8th value is for automatic self-adjusting threshold mode. ? adaptive thresholds improve around 10% per threshold level. however, adaptive thresholds can surpass optimum threshold thus causing smearing effects. ? d[6:4] = 000 is the automatic self-adjusting threshold mode. d[6:4] = 001 is the low threshold and d[6:4] = 111 is the highest threshold, d[6:4] = 010 to 110 is therefore the intermediate thresholds. figures 11 through 20 show the bandwidth of notch, chroma bandpass, and comb filters for ntsc and pal video. these curves were taken from a vm700 using a sweep frequency pattern using a tg2000 on the ml6440eval board. figure 11 shows the ntsc luma filter at ccir clock rates. this shows significant notch from 3.3mhz to 4mhz sufficient for hanging dot suppression. figure 12 shows the ntsc chroma bandpass filter for ccir clock rates where the filter at 3db ranges from 3mhz to 4.5mhz. figure 13 shows the luma notch filter for ntsc for hanging dot suppression for square pixel clock rates. this notch ranges from 3.3mhz to 4mhz. figure 14 shows the ntsc chroma bandpass filter running at square pixel clock rates where it at 3db ranges from 3mhz to 4.5mhz. figure 15 shows pal luma response for hanging dot suppression at square pixel clock rates where notch ranges from 3.3mhz to 4mhz. figure 16 shows the response of the pal chroma bandpass filter where at 3db down ranges from 3mhz to 4.2mhz. figure 17 shows the pal notch filter for luma hanging dot suppression at ccir clock rates where the notch ranges from 4mhz to 4.3mhz at zero amplitude. figure 18 shows the bandpass filter for pal at the ccir clock rate where at 3db down ranges from 3.3mhz to 4.7mhz. figure 19 shows the ntsc luma comb response over a line-to-line in phase frequency sweep to 6mhz for ccir clock rates. figure 20 shows the ntsc luma response over a line-to-line in phase frequency sweep to 6mhz for square pixel clock rate. figure 11. ntsc luma notch filter ccir amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35 figure 12. ntsc chroma band phase filter, ccir amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35
ml6440 16 figure 15. pal luma notch filter, square pixel clock figure 16. pal chroma band pass filter, square pixel clock figure 17. pal luma notch filter figure 13. ntsc luma notch filter, square pixel figure 14. ntsc chroma band pass at square pixel clock amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35 amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35 amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35 amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35 amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35 figure 18. pal chroma pand pass, ccir clock rate amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35
ml6440 17 figure 19. ntsc comb luma, ccir clock rate figure 20. ntsc luma comb, square pixel amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35 amplitude (v) frequency (mhz) 1.5 1.0 0.5 0 C0.5 02 6 4 1 35
ml6440 18 ds6440-01 physical dimensions inches (millimeters) ordering information part number temperature range package ML6440CH (eol) 0c to 70c 44 pin tqfp (h44-14) 2092 concourse drive san jose, ca 95131 tel: (408) 433-5200 fax: (408) 432-0295 www.microlinear.com 0.048 max (1.20 max) seating plane 0.630 bsc (16.00 bsc) 0.551 bsc (14.00 bsc) 1 0.551 bsc (14.00 bsc) 0.630 bsc (16.00 bsc) 12 34 23 0.039 bsc (1.00 bsc) pin 1 id 0.014 - 0.020 (0.36 - 0.51) 0.037 - 0.041 (0.95 - 1.05) 0.018 - 0.030 (0.45 - 0.75) 0.003 - 0.008 (0.09 - 0.20) 0o - 7o package: h44-14 44-pin (14 x 14 x 1mm) tqfp ? micro linear 2000. is a registered trademark of micro linear corporation. all other trademarks are the property of their respective owners. products described herein may be covered by one or more of the following u.s. patents: 4,897,611; 4,964,026; 5,027,116; 5,281,862; 5,283,483; 5,418,502; 5,508,570; 5,510,727; 5,523,940; 5,546,017; 5,559,470; 5,565,761; 5,592,128; 5,594,376; 5,652,479; 5,661,427; 5,663,874; 5,672,959; 5,689,167; 5,714,897; 5,717,798; 5,742,151; 5,747,977; 5,754,012; 5,757,174; 5,767,653; 5,777,514; 5,793,168; 5,798,635; 5,804,950; 5,808,455; 5,811,999; 5,818,207; 5,818,669; 5,825,165; 5,825,223; 5,838,723; 5.844,378; 5,844,941. japan: 2,598,946; 2,619,299; 2,704,176; 2,821,714. other patents are pending. micro linear makes no representations or warranties with respect to the accuracy, utility, or completeness of the contents of this publication and reserves the right to make changes to specifications and product descriptions at any time without notice. no license, express or implied, by estoppel or otherwise, to any patents or other intellectual property rights is granted by th is document. the circuits contained in this document are offered as possible applications only. particular uses or applications may invalidate some of the specifications and/or product descriptions contained herein. the customer is urged to perform its own engineering review before deciding on a particular application. micro linear assumes no liability whatsoever, and disclaims any express or implied warranty, relating to sale and/or use of micro linear products including liability or warranti es relating to merchantability, fitness for a particular purpose, or infringement of any intellectual property right. micro linea r products are not designed for use in medical, life saving, or life sustaining applications.


▲Up To Search▲   

 
Price & Availability of ML6440CH

All Rights Reserved © IC-ON-LINE 2003 - 2022  

[Add Bookmark] [Contact Us] [Link exchange] [Privacy policy]
Mirror Sites :  [www.datasheet.hk]   [www.maxim4u.com]  [www.ic-on-line.cn] [www.ic-on-line.com] [www.ic-on-line.net] [www.alldatasheet.com.cn] [www.gdcy.com]  [www.gdcy.net]


 . . . . .
  We use cookies to deliver the best possible web experience and assist with our advertising efforts. By continuing to use this site, you consent to the use of cookies. For more information on cookies, please take a look at our Privacy Policy. X