HMP8117 ® Data Sheet April 19, 2007 FN4643.3 NTSC/PAL Video Decoder Features The HMP8117 is a high quality NTSC and PAL video decoder with internal A/D converters. It is compatible with NTSC M, PAL B, D, G, H, I, M, N, and combination N (NC) video standards. • (M) NTSC and (B, D, G, H, I, M, N, NC) PAL Operation - Optional Auto Detect of Video Standard - ITU-R BT.601 (CCIR601) and Square Pixel Operation Both composite and S-video (Y/C) input formats are supported. A 2-line comb filter plus a user-selectable chrominance trap filter provide high quality Y/C separation. User adjustments include brightness, contrast, saturation, hue, and sharpness. Vertical blanking interval (VBI) data, such as Closed Captioning, Wide Screen Signalling and Teletext, may be captured and output as BT.656 ancillary data. Closed Captioning and Wide Screen Signalling information may also be read out via the I2C interface. The Videolyzer™ feature provides approved Macrovision™ copy-protection bypass and detection. Ordering Information PART NUMBER TEMP RANGE (°C) PART MARKING • Videolyzer Feature - Macrovision™ Bypass and Detection • Digital Anti-Alias Filter • Power Down Mode • Digital Output Formats - VMI Compatible - 8-bit, 16-bit 4:2:2 YCbCr - 15-bit (5, 5, 5), 16-bit (5, 6, 5) RGB - Linear or Gamma-Corrected - 8-bit BT.656 • Analog Input Formats - Three Analog Composite Inputs - Analog Y/C (S-video) Input • “Raw” (Oversampled) VBI Data Capture PACKAGE PKG DWG. # HMP8117CN HMP8117CN 0 to +70 80 Ld PQFP Q80.14x20 (Note 2) HMP8117CNZ (Note 1) HMP8117CNZ 0 to +70 80 Ld PQFP Q80.14x20 (Note 2) (Pb-free) HMPVIDEVAL/ISA Evaluation Board: ISA Frame Grabber (Note 3) • “Sliced” VBI Data Capture Capabilities - Closed Captioning - Widescreen Signalling (WSS) - BT.653 System B, C and D Teletext - North American Broadcast Teletext (NABTS) - World System Teletext (WST) • 2-Line (1H) Comb Filter Y/C Separator • Fast I2C Interface NOTES: 1. Intersil Pb-free plus anneal products employ special Pb-free material sets; molding compounds/die attach materials and 100% matte tin plate termination finish, which are RoHS compliant and compatible with both SnPb and Pb-free soldering operations. Intersil Pb-free products are MSL classified at Pb-free peak reflow temperatures that meet or exceed the Pb-free requirements of IPC/JEDEC J STD-020. 2. PQFP is also known as QFP and MQFP. 3. Evaluation Board descriptions are in the Applications section. • Pb-Free Plus Anneal Available (RoHS Compliant) Applications • Multimedia PCs • Video Conferencing • Video Compression Systems • Video Security Systems • LCD Projectors and Overhead Panels • Related Products - NTSC/PAL Encoders: HMP8156, HMP8170 1 CAUTION: These devices are sensitive to electrostatic discharge; follow proper IC Handling Procedures. 1-888-INTERSIL or 1-888-468-3774 | Intersil (and design) is a registered trademark of Intersil Americas Inc. Copyright Intersil Americas Inc. 2003, 2006, 2007. All Rights Reserved All other trademarks mentioned are the property of their respective owners. Functional Block Diagram SEE ANALOG FRONT END BLOCK DIAGRAM SEE DIGITAL PROCESSING BLOCK DIAGRAM EXTERNAL ANTI-ALIAS FILTER MACROVISION DETECT VBI DETECTION & DECODING LOGIC 2 YOUT COMPOSITE/LUMA CVBS1 CVBS2 CVBS3(Y) INPUT CLAMP, MUX, COARSE AGC, DC-RESTORE YIN YIN[7:0] ADC P[15:0] HSYNC INPUT SAMPLE RATE CONVERTER Y/C SEPARATION AND USER ADJUSTS OUTPUT SAMPLE RATE CONVERTER OUTPUT VSYNC TIMING BLANK AND DATA FIELD FIFO DVALID CHROMA EXTERNAL ANTI-ALIAS FILTER C COARSE AGC DC-RESTORE VBIVALID CIN[7:0] ADC CCAP SYNC-TIP, BACKPORCH TIMING MICROPROCESSOR INTERFACE AND CONTROL RESET INTREQ SA SCL SCL HMP8117 FINE AGC, DC-RESTORE CONTROL LCAP FN4643.3 April 19, 2007 Analog Front End Block Diagram (EXTERNAL) INPUT VIDEO nmos 1.0μF 75 1.0μF VID1 (INTERNAL CLAMP) VAA 1.75V TO MUX + - 13 # PIN 50μA CVBS1 3 7 CLAMP 6 CLAMP BACKPORCH ENABLE GAIN CORRECTION LOGIC LOW-PASS FILTER (REMOVE Fsc) 4 4 TO 13 DECODER CHROMA ATTEN (BELOW) GAIN CONTROL SET POINT 11 CHROMA MULT. (BELOW) FINE ADJUST MULT. FACTOR 1.0μF CVBS2 1.0μF Y_IN 5 CLAMP M U X 13-STEP VARIABLE ATTENUATOR 2X 9 EXTERNAL ANTI-ALIAS FILTER YIN 8 BUF 75 1.75V DISABLE CHARGE PUMP MUX SELECT 100μA DISCHARGE 76 CHARGE 0.1μF LUMA DC-RESTORE LOGIC 100μA CHROMA ATTEN SYNC-TIP ENABLE CHROMA MULT. CHROMA INPUT C_IN C 19 (SIGNAL BIAS ~ 2.0V) 13-STEP VARIABLE ATTENUATOR 2X 17 BUF 10-BIT 10 ADC DIGITAL ANTIALIAS FILTER 1.5V 2.0V DISABLE CHARGE PUMP 100μA CCAP STORAGE CAP DISCHARGE 29 CHARGE (OFFSET CHROMA SIGNAL TO ADC MID-SCALE ~= 2.0V) 0.1μF CHROMA DC-RESTORE LOGIC CORRECTION MULTIPLIER 75 EXTERNAL ANTI-ALIAS FILTER A/D_TEST 8-BIT TRUNCATE 2.5V 1.0μF Y[7:0] 100μA C[7:0] POWERDOWN SYNC-TIP ENABLE # PIN # NOMINAL (NTSC) OPERATING CONDITION FN4643.3 April 19, 2007 6 7 CVBSX. SIZE = INPUT. SYNC TIP CLAMPED AT ~= 1.75 VDC. 76 LCAP. DC-SIGNAL OFFSET ~= 2.4 VDC. 9 8 YOUT /YIN . SIZE = ~1.0 VP-P, SYNC TIP OFFSET ~= 1.5 VDC. 29 CCAP. DC-SIGNAL OFFSET ~= 2.4 VDC. 19 C. SIZE = INPUT SIZE. PORCH OFFSET ~= 2.0 VDC. 28 RSET. DC-SIGNAL OFFSET ~= 1.2 VDC. 17 A/D_TEST. SIZE ~= F(LUMA AGC), PORCH OFFSET ~= 2.0VDC. 78 REF_CAP. DC-SIGNAL OFFSET ~= 2.5 VDC. BIAS/ INTERNAL REFERENCE 28 78 RSET REF_CAP 1.0μF 12.1k NOMINAL (NTSC) OPERATING CONDITION PIN # 5 DECODER PIN # HMP8117 (OFFSET LUMA SIGNAL TO LOWER ADC REF ~= 1.5V) STORAGE CAP DIGITAL ANTIALIAS FILTER 1.5V 1.75V REF LCAP 10-BIT 10 ADC CORRECTION MULTIPLIER 75 75 CVBS3(Y) 2.5V YOUT 8-BIT TRUNCATE VID2 Digital Processing Block Diagram CLK2 FREQ SELECT (24.54, 27.0 or 29.5MHz) HUE ADJUST CHROMA PLL LOOP FILTER 4 CHROMA PLL NCO CHROMA PHASE DETECTOR CLK2 TO 4FSC RATIO 4FSC CLOCK M U X CHROMA AGC AND SATURATION ADJUST CHROMA DATA C[7:0] C, CVBS INPUT CHROMA U, V DEMOD LINE LOCKED NCO VSYNC DETECT U/V TO CbCr COLOR SPACE CONVERTER, COLOR KILLER SAMPLE FIELD AND VSYNC TO FIFO CHROMA LP FILTER ENABLE RATE CHROMA TRAP CONVERTER M LUMA DATA U X Y SHARPNESS ADJUST VBI STATUS VBI DETECTION AND DECODING LOGIC MACROVISION DETECT HORIZONTAL AND VERTICAL SHARPNESS ADJUST OUTPUT SAMPLE RATE MV STATUS NORMAL LP FILTER M U X FILTER SELECT P[15:0] CbCr CONVERTER Y[7:0] LINE LOCKED PLL LOOP FILTER Y STANDARD SELECT SYNC STRIPPER, BRIGHTNESS DATA OUTPUT TIMING AND AND CONTRAST TIMING DATA ADJUST, FIFO RGB CONVERSION HSYNC VSYNC BLANK FIELD DVALID VBIVALID HMP8117 LINE DELAY COMB FILTER LOCK STATUS FN4643.3 April 19, 2007 HMP8117 Introduction Digitization of Video The HMP8117 is designed to decode baseband composite or S-video NTSC and PAL signals, and convert them to either digital YCbCr or RGB data. In addition to performing the basic decoding operations, these devices include hardware to decode different types of VBI data and to generate full-screen blue, black and color bar patterns. Prior to A/D conversion, the input signal is offset and scaled to known video levels. After digitization, sample rate converters and a comb filter are used to perform color separation and demodulation. Digital PLLs are used to synchronize to all NTSC and PAL standards. A chroma PLL is used to maintain color lock for chroma demodulation while a line-locked PLL is used to maintain vertical spatial alignment. The PLLs are designed to maintain lock in the presence of VCR head switches, VCR trick-mode and multi-path noise. The HMP8117 provides the Videolyzer feature for Macrovision (MV) copy-protection bypass and detection. A/D Conversion Each CVBSX video input channel has a video clamp circuit that is independent of PLL timing. The input clamp provides a coarse signal offset to position the sync tip within the A/D converter sampling range so that the AGC and DCRESTORE logic can operate. A/D Conversion Video data is sampled at the CLK2 frequency then processed by the input sample rate converter. The output levels of the ADC after AGC and DC restoration processing are: External Video Processing Before a video signal can be digitized the decoder has some external processing considerations that need to be addressed. This section discusses those external aspects of the HMP8117. Analog Video Inputs The HMP8117 supports either three composite or two composite and one S-video input. Three analog video inputs (CVBS 1-3) are used to select which one of three composite video sources are to be decoded. To support S-video applications, the Y channel drives the CVBS3(Y) analog input, and the C channel drives the C analog input. The analog inputs must be AC-coupled to the video signals, as shown in the Applications section. Anti-alias Filters Although a 23 tap digital halfband anti-alias filter is provided for each A/D channel, an external passive filter is recommended for optimum performance. The digital filter has a flat response out to 5.4MHz with an approximate -3dB bandwidth of 6.3MHz using a 27MHz input CLK2 sample rate. For the CVBSx inputs, the filter is connected between the YOUT and YIN pins. For the C (chroma) input, the antialias filter should be connected before the C input. Recommended filter configurations are shown on the reference schematic in Figure 20. These filters have flat response out to 4.2MHz with an approximate -3dB bandwidth of 8MHz. If upgrading from the HMP8115 or HMP8112A, the previous filter configurations may be used but with slightly degraded bandwidth. Alternative higher or lower performance filters configurations may substituted. (M) NTSC (M, N) PAL (B, D, G, H, I, NC) PAL white 196 196 black 66 59 blank 56 59 sync 0 0 AGC and DC Restoration The AGC amplifier attenuates or amplifies the analog video signal to ensure that the blank level generates code 56 or 59 depending on the video standard. The difference from the ideal blank level of 56 or 59 is used to control the amount of attenuation or gain of the analog video signal. To obtain a stable DC reference for the AGC, a digital low-pass filter removes the chroma burst from the input signal’s backporch. DC restoration positions the video signal so that the sync tip generates a code 0. The internal timing windows for AGC and DC restoration are show in Figure 3. The appropriate windows are automatically determined by the decoder when the input signal is auto-detected or manually selected. VIDEO INPUT AGC DC RESTORE FIGURE 1. AGC AND DC RESTORE INTERNAL TIMING 5 FN4643.3 April 19, 2007 HMP8117 Input Signal Detection Comb Filter If no input video signal is detected for 16 consecutive line periods, nominal video timing is generated for the previously detected or programmed video standard. A maskable interrupt is provided for the condition of “Input Signal Loss” allowing the host to enable blue field output if desired. A 2-line comb filter, using a single line delay, is used to perform part of the Y/C separation process. During S-video operation, the Y signal bypasses the comb filter; the C signal is processed by the comb filter since it is an integral part of the chroma demodulator. During PAL operation, the chroma trap filter should also be enabled for improved performance. Vertical Sync and Field Detection The vertical sync and field detect circuit uses a low time counter to detect the vertical sync sequence in the video data stream. The low time counter accumulates the low time encountered during any sync pulse, including serration and equalization pulses. When the low time count exceeds the vertical sync detect threshold, VSYNC is asserted immediately. FIELD is asserted at the same time that VSYNC is asserted. FIELD is asserted low for odd fields and high for even fields. Field is determined from the location in the video line where VSYNC is detected. If VSYNC is detected in the first half of the line, the field is odd. If VSYNC is detected in the second half of a line, the field is even. In the case of lost vertical sync or excessive noise that would prevent the detection of vertical sync, the FIELD output will continue to toggle. Lost vertical sync is declared if after 337 lines, a vertical sync period was not detected for 1 or 3 (selectable) successive fields as specified by bit 2 of the GENLOCK CONTROL register 04H. When this occurs, the PLLs are initialized to the acquisition state. Y/C Separation A composite video signal has the luma (Y) and chroma (C) information mixed in the same video signal. The Y/C separation process is responsible for separating the composite video signal into these two components. The HMP8117 utilizes a comb filter to minimize the artifacts that are associated with the Y/C separation process. Input Sample Rate Converter The input sample rate converter is used to convert video data sampled at the CLK2 rate to a virtual 4xfSC sample rate for comb filtering and color demodulation. An interpolating filter is used to generate the 4xfSC samples as illustrated in Figure 2. Since a single line store is used, the chroma will normally have a half-line vertical offset from the luma data. This may be eliminated, vertically aligning the chroma and luma samples, at the expense of vertical resolution of the luma. Bit 0 of the OUTPUT FORMAT register 02H controls this option. Chroma Demodulation The output of the comb filter is further processed using a patented frequency domain transform to complete the Y/C separation and demodulate the chrominance. Demodulation is done at a virtual 4xfSC sample rate using the interpolated data samples to generate U and V data. The demodulation process decimates by 2 the U/V sample rate. Output Sample Rate Converter The output sample rate converter converts the Y, U and V data from a virtual 4xfSC sample rate to the desired output sample rate (i.e., 13.5MHz). It also vertically aligns the samples based on the horizontal sync information embedded in the digital video data stream. The output sample rate is determined by the input video standard and the selected rectangular/square pixel mode. The output pixel rate is 1/2 of the CLK2 input clock frequency. The output format is 4:2:2 for all modes except the RGB modes which use a 4:4:4 output format. CLK2 Input The decoder requires a stable clock source for the CLK2 input. For best performance, use termination resistor(s) to minimize pulse overshoot and reflections on the CLK2 input. Since chroma demodulation uses the virtual 4xfSC , any jitter on CLK2 will be transferred as chrominance error on the output pixels. The CLK2 clock frequency must be one of the valid selections from Table 1 below based on the video standard and desired pixel mode. TABLE 1. VIDEO STANDARD CLOCK RATE SELECTION SUMMARY INCOMING VIDEO SAMPLES VALID CLK2 FREQUENCIES (MHz) TIME RESAMPLED VIDEO VIDEO FORMAT RECTANGULAR PIXEL MODE SQUARE PIXEL MODE (M) NTSC, (M) PAL 27.00 24.54 (B, D, G, H, I, N, NC) PAL 27.00 29.50 TIME 4xfSC FIGURE 2. SAMPLE RATE CONVERSION 6 The CLK2 should be derived from a stable clock source, such as a crystal. CLK2 must have at least a ±50ppm accuracy and at least a 60/40% duty cycle to ensure proper FN4643.3 April 19, 2007 HMP8117 operation. Use of a PLL to generate a “Line Locked” CLK2 input based on the input video is not recommend. (See the following section.) Cycle Slipping and Real-Time Pixel Jitter The decoder’s digital PLL allows it to maintain lock and provide high quality Y/C separation even on the poorest quality input video signals. However, this architecture does not provide a “Line Lock Clock” output and should not be used as a timing master for direct interface to another video encoder in a system. Since the decoder uses a fixed CLK2 input frequency, the output pixel rate must be periodically adjusted to compensate for any frequency error between CLK2 and the input video signal. This output pixel rate adjustment is referred to as cycle slipping. Since the decoder has an output data FIFO, all cycle slipping can be deferred until the next horizontal blanking interval. This guarantees a consistent number of pixels during the active video region. Due to cycle slipping, the output timing and data will exhibit a nominal real-time (line-to-line) pixel jitter of one CLK2 period. Although the sample rate converter maintains a 1/8 pixel vertical sample alignment, the output data must be routed to a frame buffer or video compression chip in order remove the effects of cycle slipping. (The frame buffer or compression chip serves as a time base corrector.) By directly interfacing the decoder to a video encoder, the output video signal will directly reflect the real-time pixel jitter effects of the decoder output timing. The jitter effects can be visualized on a CRT monitor using a static image containing patterns with sharp vertical edges. The edges will appear more “ragged” when compared to the input video signal. The severity of this visual effect relates directly to the frequency error between CLK2 and the input video signal. It is nearly impossible to completely match CLK2 with the input video signal. Therefore, a direct decoder to encoder interface is not recommended. The use of an external PLL to generate a “Line Locked” CLK2 input derived from the input video signal is also not recommended, since this will defeat the internal digital PLL and result in pixel decoding errors. Digital Processing of Video Once the luma and chroma have been separated the HMP8117 then performs programmable modifications (i.e. contrast, coring, color space conversions, color AGC, etc.) to the decoded video signal. UV to CbCr Conversion The baseband U and V signals are scaled and offset to generate a nominal range of 16-240 for both the Cb and Cr data. 7 Digital Color Gain Control There are four types of color gain control modes available: no gain control, automatic gain control, fixed gain control, and freeze automatic gain control. If “no gain control” is selected, the amplitude of the color difference signals (CbCr) is not modified, regardless of variations in the color burst amplitude. Thus, a gain of 1x is always used for Cb and Cr. If “automatic gain control” is selected, the amplitude of the color difference signals (CbCr) is compensated for variations in the color burst amplitude. The burst amplitude is averaged with the two previous lines having a color burst to limit lineto-line variations. A gain of 0.5x to 4x is used for Cb and Cr. If “fixed gain control” is selected, the amplitude of the color difference signals (CbCr) is multiplied by a constant, regardless of variations in the color burst amplitude. The constant gain value is specified by the COLOR GAIN register 1CH . A gain of 0.5x to 4x is used for Cb and Cr. Limiting the gain to 4x limits the amount of amplified noise. If “freeze automatic gain control” is selected, the amplitude of the color difference signals (CbCr) is multiplied by a constant. This constant is the value the AGC circuitry generated when the “freeze automatic gain” command was selected. Color Killer If “enable color killer” is selected, the color output is turned off when the running average of the color burst amplitude is below approximately 25% of nominal for four consecutive fields. When the running average of the color burst amplitude is above approximately 25% of nominal for four consecutive fields, the color output is turned on. The color output is also turned off when excessive phase error of the chroma PLL is present. If “force color off” is selected, color information is never present on the outputs. If “force color on” is selected, color information is present on the outputs regardless of the color burst amplitude or chroma PLL phase error. Y Processing The black level is subtracted from the luminance data to remove sync and any blanking pedestal information. Negative values of Y are supported at this point to allow proper decoding of “below black” luminance levels. Scaling is done to position black at 8-bit code 0 and white at 8-bit code 219. A chroma trap filter may be used to remove any residual color subcarrier from the luminance data. The center frequency of the chroma trap is automatically determined from the video standard being decoded. The chroma trap should be disabled during S-video operation to maintain maximum luminance bandwidth. Alternately, a 3MHz low-pass filter may be used to FN4643.3 April 19, 2007 HMP8117 remove high-frequency Y data. This may make a noisy image more pleasing to the user, although softer. RGB Output Format Processing Coring of the high-frequency Y data may be done to reduce low-level high frequency noise. The 4:2:2 YCbCr data is converted to 4:4:4 YCbCr data and then converted to either 15-bit or 16-bit gamma-corrected RGB (R′G′B′) data. While BLANK is asserted, RGB data is forced to a value of 0. Coring of the Y data may also be done to reduce low-level noise around black. This forces Y data with the following values to a value of 0: 15-Bit R′G′B′ coring = 1: ± 1 coring = 2: ± 1, ± 2 coring = 3: ± 1, ± 2. ± 3 High-frequency components of the luminance signal may be “peaked” to control the sharpness of the image. Maximum gain may be selected to occur at either 2.6MHz or the color subcarrier frequency. This may be used to make the displayed image more pleasing to the user. It should not be used if the output video will be compressed, as the circuit introduces high-frequency components that will reduce the compression ratio. The brightness control adds or subtracts a user-specified DC offset to the Y data. The contrast control multiplies the Y data by a user-specified amount. These may be used to make the displayed image more pleasing to the user. Finally, a value of 16 is added to generate a nominal range of 16 (black) to 235 (white). CbCr Processing The CbCr data is low-pass filtered to either 0.85MHz or 1.5MHz. Coring of the CbCr data may be done to reduce low-level noise around zero. This forces CbCr data with the following values to a value of 128. coring = 1: 127, 129 coring = 2: 126, 127, 129, 130 coring = 3: 125, 126, 127, 129, 130, 131 The saturation control multiplies the CbCr data by a userspecified amount. This may be used to make the displayed image more pleasing to the user. The CbCr data may also be optionally multiplied by the contrast value to avoid color shifts when changing contrast. The hue control provides a user-specified phase offset to the color subcarrier during decoding. This may be used to correct slight hue errors due to transmission. YCbCr Output Format Processing Y has a nominal range of 16 to 235. Cb and Cr have a nominal range of 16 to 240, with 128 corresponding to zero. Values less than 1 are made 1 and values greater than 254 are made 254. While BLANK is asserted, Y is forced to have a value of 16, with Cb and Cr forced to have a value of 128, unless VBI data is present. 8 The following YCbCr to R′G′B′ equations are used to maintain the proper black and white levels: R′ = 0.142(Y - 16) + 0.194(Cr - 128) G′ = 0.142(Y - 16) - 0.099(Cr - 128) - 0.048(Cb - 128) B′ = 0.142(Y - 16) + 0.245(Cb - 128) The resulting 15-bit R′G′B′ data has a range of 0 to 31. Values less than 0 are made 0 and values greater than 31 are made 31. The 15-bit R′G′B′ data may be converted to 15-bit linear RGB, using the following equations. Although the PAL specifications specify a gamma of 2.8, a gamma of 2.2 is normally used. The HMP8117 allows the selection of the gamma to be either 2.2 or 2.8, independent of the video standard. for gamma = 2.2: for R′G′B′ < 0.0812*31 R = (31)((R′/31)/4.5) G = (31)((G′/31)/4.5) B = (31)((B′/31)/4.5) for R′G′B′ >= 0.0812*31 R = (31)(((R′/31) + 0.099)/1.099)2.2 G = (31)(((G′/31) + 0.099)/1.099)2.2 B = (31)(((B′/31) + 0.099)/1.099)2.2 for gamma = 2.8: R = (31)(R′/31)2.8 G = (31)(G′/31)2.8 B = (31)(B′/31)2.8 16-Bit R′G′B′ The following YCbCr to R′G′B′ equations are used to maintain the proper black and white levels: R′ = 0.142(Y - 16) + 0.194(Cr - 128) G′ = 0.288(Y - 16) - 0.201(Cr - 128) - 0.097(Cb - 128) B′ = 0.142(Y - 16) + 0.245(Cb - 128) The resulting 16-bit R′G′B′ data has a range of 0 to 31 for R′ and B′, and a range of 0 to 63 for G′. Values less than 0 are made 0; R′ and B′ values greater than 31 are made 31, G′ values greater than 63 are made 63. The 16-bit R′G′B′ data may be converted to 16-bit linear RGB, using the following equations. Although the PAL specifications specify a gamma of 2.8, a gamma of 2.2 is normally used. The HMP8117 allows the selection of the gamma to be either 2.2 or 2.8, independent of the video standard. FN4643.3 April 19, 2007 HMP8117 for gamma = 2.2: Pixel Port Timing for R′B′ < 0.0812*31, G′ < 0.0812*63 R = (31)((R′/31)/4.5) G = (63)((G′/63)/4.5) B = (31)((B′/31)/4.5) The the timing and format of the output data and control signals is presented in the following sections. Refer to the section “CYCLE SLIPPING AND REAL-TIME PIXEL JITTER” for PLL and interface considerations. for R′B′ >= 0.0812*31, G′ >= 0.0812*63 R = (31)(((R′/31) + 0.099)/1.099)2.2 G = (63)(((G′/63) + 0.099)/1.099)2.2 B = (31)(((B′/31) + 0.099)/1.099)2.2 HSYNC and VSYNC Timing The HSYNC and VSYNC output timing is VMI v1.4 compatible. Figures 3-6 illustrate the video timing. The leading edge of HSYNC is synchronous to the video input signal and has a fixed latency due to internal pipeline processing. The pulse width of the HSYNC is defined by the END HSYNC register 36H , where the trailing edge of HSYNC has a programmable delay of 0-510 CLK2 cycles from the leading edge. for gamma = 2.8: R = (31)(R′/31)2.8 G = (63)(G′/63)2.8 B = (31)(B′/31)2.8 The leading edge of VSYNC is asserted approximately half way through the first serration pulse of each field. An accumulator is used to detect a low-time period within the serration pulse. Since the leading edge of VSYNC is detected, it should not be used for timing with respect to HSYNC or BLANK. Built-in Video Generation The decoder can be configured to output a full-screen of built-in blue, black or 75% color bar patterns. The type of pattern generated is determined by bits 2-1 of the OUTPUT FORMAT register 02H . When built-in video generation is not desired, the bits need to be set for normal operation to pass decoded video. The trailing edge of VSYNC implements the VMI handshake with HSYNC in order to determine field information without using the FIELD pin. For an odd field, the trailing edge of VSYNC is 5 ±1 CLK2 cycles after the trailing edge of the HSYNC that follows the last equalization pulse. Refer to Figures 3 and 5. For an even field, the trailing edge of VSYNC is 5 ±1 CLK2 cycles after the leading edge of the HSYNC that follows the last equalization pulse. Refer to Figures 4 and 6. If the decoder is currently locked to a video source on the input, the output data timing will be based on the input video source. If an input video source is not detected, internallygenerated output data timing will be used. The following table lists the data codes output for each built-in video pattern in YCbCr format. Field Timing TABLE 2. BUILT-IN VIDEO PATTERN DATA CODES PATTERN: COLOR Y Cb Cr 75% Color Bar: White Yellow Cyan Green Magenta Red Blue Black B4H A2H 83H 70H 54H 41H 23H 10H 80H 2CH 9CH 48H B8H 64H D4H 80H 80H 8EH 2CH 3AH C6H D4H 72H 80H Blue Screen: Blue 4BH D9H 88H Black Screen: Black 10H 80H 80H When field information can be determined from the input video source, the FIELD output pin reflects the video source field state. When field information cannot be determined from the input video source, the FIELD output pin alternates its state at the beginning of each field. FIELD changes state 5±1 CLK2 cycles before the leading edge of VSYNC. NTSC(M) LINE# 524 525 1 2 3 4 5 6 7 8 9 10 PAL(M) LINE# 521 522 523 524 525 1 2 3 4 5 6 7 VIDEO INPUT HSYNC VSYNC FIELD ‘EVEN’ FIELD ‘ODD’ FIELD FIGURE 3. NTSC(M) AND PAL(M) ODD FIELD TIMING 9 FN4643.3 April 19, 2007 HMP8117 NTSC(M) LINE# 262 263 264 265 266 267 268 269 270 271 272 273 PAL(M) LINE# 259 260 261 262 263 264 265 266 267 268 269 270 VIDEO INPUT HSYNC VSYNC FIELD ‘ODD’ FIELD ‘EVEN’ FIELD FIGURE 4. NTSC(M) AND PAL(M) EVEN FIELD TIMING LINE # 621 622 623 624 625 1 2 3 4 5 6 7 317 318 319 320 VIDEO INPUT HSYNC VSYNC FIELD ‘ODD’ FIELD ‘EVEN’ FIELD FIGURE 5. PAL(B, D, G, H, I, N, NC) ODD FIELD TIMING LINE # 309 310 311 312 313 314 315 316 VIDEO INPUT HSYNC VSYNC FIELD ‘ODD’ FIELD ‘EVEN’ FIELD FIGURE 6. PAL(B, D, G, H, I, N, NC) EVEN FIELD TIMING BLANK and DVALID Timing DVALID is asserted when P15-P0 contain valid data. The behavior of the DVALID output is determined by bit 4 (DVLD_LTC) and bit 5 (DLVD_DCYC) of the GENLOCK CONTROL register 04H for each video output mode. The BLANK output pin is used to distinguish the blanking interval period from active video data. The blanking intervals are programmable in both horizontal and vertical dimensions. Reference Figure 7 for active video timing and use Table 3 for typical blanking programming values. During active scan lines, BLANK is asserted when the horizontal pixel count matches the value in the START H_BLANK register 31H/30H . The pixel counter is 000H at the 10 leading edge of the sync tip after a fixed pipeline delay. Since blanking normally occurs on the front porch, (prior to count 000H) the START H_BLANK count must be programmed with a large value from the previous line. Refer to the Last Pixel Count from Table 3. BLANK is negated when the horizontal pixel count matches the value in the END H_BLANK register 32H. Note that horizontally, BLANK is programmable with two pixel resolution. START V_BLANK register 34H/33H and END V_BLANK register 35H determine which scan lines are blanked for each field. During inactive scan lines, BLANK is asserted during the entire scan line. Half-line blanking of the output video cannot be done. FN4643.3 April 19, 2007 HMP8117 NTSC M PAL B, D, G, H, I, N, NC LINES 1 - 22 NOT ACTIVE LINES 1 - 22 NOT ACTIVE ODD FIELD SYNC AND BACK PORCH 240 ACTIVE LINES PER FIELD (LINES 23-262) 288 ACTIVE LINES PER FIELD (LINES 23 - 310) VERTICAL BLANKING 480 ACTIVE LINES/FRAME (NTSC, PAL M) EVEN FIELD LINES 263 - 284 NOT ACTIVE LINES 311 - 335 NOT ACTIVE FRONT PORCH 240 ACTIVE LINES PER FIELD (LINES 285 - 524) LINE 525 NOT ACTIVE TOTAL PIXELS 858 (780) ACTIVE PIXELS 288 ACTIVE LINES PER FIELD (LINES 336 - 623) NUMBER OF PIXELS RECTANGULAR (SQUARE) NTSC 720 (640) 576 ACTIVE LINES/FRAME (PAL) LINES 624-625 NOT ACTIVE PAL 864 (944) TOTAL PIXELS 720 (768) ACTIVE PIXELS NOTE: 4. The line numbering for PAL (M) is the NTSC (M) line count minus 3 per the video standards. FIGURE 7. TYPICAL ACTIVE VIDEO REGIONS TABLE 3. TYPICAL VALUES FOR H_BLANK AND V_BLANK REGISTERS VIDEO STANDARD (MSB/LSB) ACTIVE PIXELS/ LINE TOTAL PIXELS/ LINE LAST PIXEL COUNT START H_BLANK (31H/30H) END H_BLANK (32H) START V_BLANK (34H/33H) END V_BLANK (35H) 720 720 858 864 857 (0359H) 863 (035FH) 842 (034AH) 852 (0354H) 122 (7AH) 132 (84H) 259 (0103H) 310 (0136H) 19 (13H) 22 (16H) 640 768 780 944 779 (030BH) 943 (03AFH) 758 (02F6H) 922 (039AH) 118 (76H) 154 (9AH) 259 (0103H) 310 (0136H) 19 (13H) 22 (16H) RECTANGULAR PIXELS NTSC (M), PAL (M) PAL (B, D, G, H, I, N, NC) SQUARE PIXELS NTSC (M), PAL (M) PAL (B, D, G, H, I, N, NC) TABLE 4. PIXEL OUTPUT FORMATS PIN NAME 8-BIT, 4:2:2, YCbCr 16-BIT, 4:2:2, YCbCr 15-BIT, RGB, (5,5,5) 16-BIT, RGB, (5,6,5) BT.656 P0 P1 P2 P3 P4 P5 P6 P7 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] Cb0, Cr0 [D0n+1] Cb1, Cr1 [D1n+1] Cb2, Cr2 [D2n+1] Cb3, Cr3 [D3n+1] Cb4, Cr4 [D4n+1] Cb5, Cr5 [D5n+1] Cb6, Cr6 [D6n+1] Cb7, Cr7 [D7n+1] B0 [D0n+1] B1 [D1n+1] B2 [D2n+1] B3 [D3n+1] B4 [D4n+1] G0 [D5n+1] G1 [D6n+1] G2 [D7n+1] B0 [D0n+1] B1 [D1n+1] B2 [D2n+1] B3 [D3n+1] B4 [D4n+1] G0 [D5n+1] G1 [D6n+1] G2 [D7n+1] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] P8 P9 P10 P11 P12 P13 P14 P15 Y0, Cb0, Cr0 [D0] Y1, Cb1, Cr1 [D1] Y2, Cb2, Cr2 [D2] Y3, Cb3, Cr3 [D3] Y4, Cb4, Cr4 [D4] Y5, Cb5, Cr5 [D5] Y6, Cb6, Cr6 [D6] Y7, Cb7, Cr7 [D7] Y0 [D0n] Y1 [D1n] Y2 [D2n] Y3 [D3n] Y4 [D4n] Y5 [D5n] Y6 [D6n] Y7 [D7n] G3 [D0n] G4 [D1n] R0 [D2n] R1 [D3n] R2 [D4n] R3 [D5n] R4 [D6n] 0 [D7n] G3 [D0n] G4 [D1n] G5 [D2n] R0 [D3n] R1 [D4n] R2 [D5n] R3 [D6n] R4 [D7n] YCbCr Data, Ancillary Data, SAV and EAV Sequences [D0-D7, where P8 corresponds to D0] NOTE: 5. Definitions in brackets are port definitions during raw VBI data transfers. Refer to the section on teletext for more information on raw VBI. 11 FN4643.3 April 19, 2007 HMP8117 Pixel Output Port Cr Y′...], with the first active data each scan line containing Cb data. The pixel output timing is shown in Figures 8 and 9. Pixel data is output via the P0-P15 pins. Refer to Table 4 for the output pin definition as a function of the output mode. Refer to the section “CYCLE SLIPPING AND REAL-TIME PIXEL JITTER” for PLL and interface considerations. BLANK, HSYNC, VSYNC, DVALID, VBIVALID, and FIELD are output following the rising edge of CLK2. When BLANK is asserted and VBIVALID is deasserted, the YCbCr outputs have a value of 16 for Y and 128 for Cb and Cr. The behavior of the DVALID output is determined by bit 4 (DVLD_LTC) of the GENLOCK CONTROL register 04H . 8-Bit YCbCr Output Each YCbCr data byte is output following each rising edge of CLK2. The YCbCr data is multiplexed as [Cb Y Cr Y′ Cb Y CLK DVALID BLANK P[15-8] Cb0 Y0 Cr0 Y1 Cb2 Y2 Cr2 Y3 Cb4 Y4 tDVLD NOTE: 6. Y0 is the first active luminance pixel data of a line. Cb0 and Cr0 are first active chrominance pixel data in a line. Cb and Cr will alternate every cycle due to the 4:2:2 subsampling. Pixel data is not output during the blanking period, but the values are forced to blanking levels. FIGURE 8. OUTPUT TIMING FOR 8-BIT YCbCr MODE (DVLD_LTC = 0) 16-Bit YCbCr, 15-Bit RGB, or 16-RGB Output is asserted and VBIVALID is deasserted, the YCbCr outputs have a value of 16 for Y and 128 for Cb and Cr; the RGB outputs have a value of 0. For 16-bit YCbCr, 15-bit RGB data, or 16-bit RGB output modes, the data is output following the rising edge of CLK2 with DVALID asserted. Either linear or gamma-corrected RGB data may be output. The pixel output timing is shown in Figures 10 to 13. The behavior of the DVALID output is determined by bit 4 (DVLD_LTC) and bit 5 (DLVD_DCYC) of the GENLOCK CONTROL register 04H. BLANK, HSYNC, VSYNC, DVALID, VBIVALID, and FIELD are output following the rising edge of CLK2. When BLANK CLK DVALID BLANK P[15-8] Cb0 Y0 Cr0 Y1 Cb2 Y2 Cr2 Y3 Cb4 Y4 tDVLD NOTES: 7. Y0 is the first active luminance pixel data of a line. Cb0 and Cr0 are first active chrominance pixel data in a line. Cb and Cr will alternate every cycle due to the 4:2:2 subsampling. Pixel data is not output during the blanking period, but the values are forced to blanking levels. 8. When DVLD_LTC is set to 1, the polarity of DVALID needs to be set to active low, otherwise DVALID will stay low during active video and be gated with the clock only during the blanking interval. FIGURE 9. OUTPUT TIMING FOR 8-BIT YCbCr MODE (DVLD_LTC = 1) 12 FN4643.3 April 19, 2007 HMP8117 CLK DVALID BLANK P15-P8 Y0 Y1 Y2 Y3 Y4 P7-P0 Cb0 Cr0 Cb2 Cr2 Cb4 tDVLD NOTES: 9. Y0 is the first active luminance pixel data of a line. Cb0 and Cr0 are first active chrominance pixel data in a line. Cb and Cr will alternate every cycle due to the 4:2:2 subsampling. 10. BLANK is asserted per Figure 7. FIGURE 10. OUTPUT TIMING FOR 16-BIT YCbCr MODE (DVLD_LTC = 0, DVLD_DCYC = 0) CLK DVALID P15-P11 [P14-P10] R0 R1 R2 R3 R4 P10-P5 [P9-P5] G0 G1 G2 G3 G4 P4-P0 B0 B1 B2 B3 B4 tDVLD NOTE: 11. BLANK is asserted per Figure 7. FIGURE 11. OUTPUT TIMING FOR 16-BIT [15-BIT] RGB MODE (DVLD_LTC = 0, DVLD_DCYC = 0) 13 FN4643.3 April 19, 2007 HMP8117 CLK DVALID P15-P8 P7-P0 Y0 Y1 Y2 Y3 Y4 Cb0 Cr0 Cb2 Cr2 Cb4 tDVLD NOTES: 12. Y0 is the first active luminance pixel of a line. Cb0 and Cr0 are first active chrominance pixels in a line. Cb and Cr will alternate every cycle due to the 4:2:2 subsampling. 13. BLANK is asserted per Figure 7. 14. DVALID is asserted for every valid pixel during both active and blanking regions. FIGURE 12. OUTPUT TIMING FOR 16-BIT YCbCr MODE (DVLD_LTC = 0, DVLD_DCYC = 1) CLK DVALID BLANK P15-P11 [P14-P10] R0 R1 R2 R3 R4 P10-P5 [P9-P5] G0 G0 G2 G2 G4 P4-P0 B0 B1 B2 B3 B4 tDVLD NOTES: 15. BLANK is asserted per Figure 7. 16. DAVLID is asserted for every valid pixel during both active and blanking regions. DVALID is not a 50% duty cycle synchronous output and will appear to jitter as the Output Sample Rate converter adjusts the output timing for various data rates and clock frequency inputs. FIGURE 13. OUTPUT TIMING FOR 16-BIT [15-BIT] RGB MODE (DVLD_LTC = 0, DVLD_DCYC = 1) 8-Bit BT.656 Output For the BT.656 output mode, data is output following each rising edge of CLK2. The BT.656 EAV and SAV formats are shown in Table 5 and the pixel output timing is shown in Figure 14. The EAV and SAV timing is determined by the programmed horizontal and vertical blank timing. During the blanking intervals, the YCbCr outputs have a value of 16 for Y and 128 for Cb and Cr, unless ancillary data is present. BLANK, HSYNC, VSYNC, DVALID, VBIVALID, and FIELD are output following the rising edge of CLK2. 14 FN4643.3 April 19, 2007 HMP8117 . CLK DVALID BLANK P[15-8] FF 00 00 Status Cb0 Y0 Cr0 Y1 Cb2 Y2 tDVLD NOTES: 17. Y0 is the first active luminance pixel data of a line. Cb0 and Cr0 are first active chrominance pixel data in a line. Cb and Cr will alternate every cycle due to the 4:2:2 subsampling. Pixel data is not output during the blanking period. 18. Notice that DVALID is not asserted during the preamble and that BLANK is still asserted. 19. See table 5 for Status bit definitions. FIGURE 14. OUTPUT TIMING FOR 8-BIT BT.656 MODE TABLE 5. BT.656 EAV AND SAV SEQUENCES PIXEL INPUT P15 P14 P13 P12 P11 P10 P9 P8 Preamble 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 F V H P3 P2 P1 P0 Status Word NOTES: 20. P3 = V xor H; P2 = F xor H; P1 = F xor V; P0 = F xor V xor H 21. F: “0” = field 1; “1” = field 2 22. V: “1” during vertical blanking 23. H: “0” at SAV (start of active video); “1” at EAV (end of active video) Advanced Features In addition to digitizing an analog video signal the HMP8117 has hardware to process different types of Vertical Blanking Interval (VBI) data as described in the following sections. “Sliced” VBI Data Capture The HMP8117 implements “sliced” data capture of select types of VBI data. The VBI decoders incorporate detection hysteresis to prevent them from rapidly turning on and off due to noise and transmission errors. In order to handle realworld signals, the VBI decoders also compensate for DC offsets and amplitude variations. Closed Captioning During closed captioning capture, the scan lines containing captioning information are monitored. If closed captioning is enabled and captioning data is present, the caption data is loaded into the caption data registers. DETECTION OF CLOSED CAPTIONING The closed caption decoder monitors the appropriate scan lines looking for the clock run-in and start bits used by captioning. If found, it locks to the clock run-in, the caption 15 data is sampled and loaded into shift registers, and the data is then transferred to the caption data registers. If the clock run-in and start bits are not found, it is assumed the scan line contains video data unless other VBI information is detected, such as teletext. Once the clock run-in and start bits are found on the appropriate scan line for four consecutive odd fields, the Closed Captioning odd field Detect status bit is set to “1”. It is reset to “0” when the clock run-in and start bits are not found on the appropriate scan lines for four consecutive odd fields. Once the clock run-in and start bits are found on the appropriate scan line for four consecutive even fields, the Closed Captioning even field Detect status bit is set to “1”. It is reset to “0” when the clock run-in and start bits are not found on the appropriate scan lines for four consecutive even fields. READING THE CAPTION DATA The caption data registers may be accessed in two ways: via the I2C interface or as BT.656 ancillary data. FN4643.3 April 19, 2007 HMP8117 CAPTIONING DISABLED ON BOTH LINES DETECTION OF WSS In this case, any caption data present is ignored. The WSS decoder monitors the appropriate scan lines looking for the run-in and start codes used by WSS. If found, it locks to the run-in code, the WSS data is sampled and loaded into shift registers, and the data is then transferred to the WSS data registers. The Caption odd field Read status bit and the Caption even field Read status bit are always a “0”. ODD FIELD CAPTIONING In this case, any caption data present on line 284 (or line 281 or 335 in the PAL modes) is ignored. Caption data present on line 21 (or line 18 or 22 in the PAL modes) is captured into a shift register then transferred to CLOSED CAPTION_ODD_A register 20H and CLOSED CAPTION_ODD_B register 21H. The Caption even field Read status bit is always a “0”. The Caption odd field Read status bit is set to “1” after data has been transferred from the shift register to the CLOSED CAPTION_ODD_A and CLOSED CAPTION_ODD_B registers. It is set to “0” after the data has been read out. EVEN FIELD CAPTIONING In this case, any caption data present on line 21 (or line 18 or 22 in the PAL modes) is ignored. Caption data present on line 284 (or line 281 or 335 in the PAL modes) is captured into a shift register then transferred to CLOSED CAPTION_EVEN_A register 22H and CLOSED CAPTION_EVEN_B register 23H. If the run-in and start codes are not found, it is assumed the scan line contains video data unless other VBI information is detected, such as teletext. Once the run-in and start codes are found on the appropriate scan line for four consecutive odd fields, the WSS Line 20 Detect status bit is set to “1”. It is reset to “0” when the run-in and start codes are not found on the appropriate scan lines for four consecutive odd fields. Once the run-in and start codes are found on the appropriate scan line for four consecutive even fields, the WSS Line 283 Detect status bit is set to “1”. It is reset to “0” when the clock run-in and start bits are not found on the appropriate scan lines for four consecutive even fields. READING THE WSS DATA The WSS data registers may be accessed in two ways: via the I2C interface or as BT.656 ancillary data. WSS DISABLED ON BOTH LINES The Caption odd field Read status bit is always a “0”. The Caption even field Read status bit is set to “1” after data has been transferred from the shift register to the CLOSED CAPTION_EVEN_A and CLOSED CAPTION_EVEN_B registers. It is set to “0” after the data has been read out. In this case, any WSS data present is ignored. ODD AND EVEN FIELD CAPTIONING In this case, any WSS data present on line 283 (or line 280 or 336 in the PAL modes) is ignored. WSS data present on line 20 (or line 17 or 23 in the PAL modes) is captured into a shift register then transferred to the WSS_ODD_A and WSS_ODD_B data registers. Caption data present on line 21 (or line 18 or 22 in the PAL modes) is captured into a shift register then transferred to the CLOSED CAPTION_ODD_A and CLOSED CAPTION_ODD_B registers. Caption data present on line 284 (or line 281 or 335 in the PAL modes) is captured into a shift register then transferred to the CLOSED CAPTION_EVEN_A and CLOSED CAPTION_EVEN_B registers. The Caption odd field Read status bit is set to “1” after data has been transferred from the shift register to the CLOSED CAPTION_ODD_A and CLOSED CAPTION_ODD_B registers. It is set to “0” after the data has been read out. The Caption even field Read status bit is set to “1” after data has been transferred from the shift register to the CLOSED CAPTION_EVEN_A and CLOSED CAPTION_EVEN_B registers. It is set to “0” after the data has been read out. Widescreen Signalling (WSS) During WSS capture (ITU-R BT.1119 and EIAJ CPX-1204), the scan lines containing WSS information are monitored. If WSS is enabled and WSS data is present, the WSS data is loaded into the WSS data registers. 16 The WSS odd field Read status bit and the WSS even field Read status bit are always a “0”. ODD FIELD WSS The WSS even field Read status bit is always a “0”. The WSS odd field Read status bit is set to “1” after data has been transferred from the shift register to the WSS_ODD_A and WSS_ODD_B registers. It is set to “0” after the data has been read out. EVEN FIELD WSS In this case, any WSS data present on line 20 (or line 17 or 23 in the PAL modes) is ignored. WSS data present on line 283 (or line 280 or 336 in the PAL modes) is captured into a shift register then transferred to the WSS_EVEN_A and WSS_EVEN_B data registers. The WSS odd field Read status bit is always a “0”. The WSS even field Read status bit is set to “1” after data has been transferred from the shift register to the WSS_EVEN_A and WSS_EVEN_B registers. It is set to “0” after the data has been read out. FN4643.3 April 19, 2007 HMP8117 ODD AND EVEN WSS Real-Time Control Interface (RTCI) information. Teletext and RTCI data is only available as BT.656 ancillary data. WSS data present on line 20 (or line 17 or 23 in the PAL modes) is captured into a shift register then transferred to the WSS_ODD_A and WSS_ODD_B registers. WSS data present on line 283 (or line 280 or 336 in the PAL modes) is captured into a shift register then transferred to the WSS_EVEN_A and WSS_EVEN_B registers. VBIVALID Output Timing The VBIVALID output is asserted when outputting Closed Captioning, Wide Screen Signalling, Teletext or RTCI data as BT.656 ancillary data. It is asserted during the entire BT.656 ancillary data packet time, including the preamble. The WSS odd field Read status bit is set to “1” after data has been transferred from the shift register to the WSS_ODD_A and WSS_ODD_B registers. It is set to “0” after the data has been read out. BT.656 Closed Captioning and Wide Screen Signalling Table 6 illustrates the format when outputting the caption data registers as BT.656 ancillary data. The ancillary data is present during the horizontal blanking interval after the line containing the captioning information. The WSS even field Read status bit is set to “1” after data has been transferred from the shift register to the WSS_EVEN_A and WSS_EVEN_B registers. It is set to “0” after the data has been read out. Table 7 illustrates the format when outputting the WSS data registers as BT.656 ancillary data. The ancillary data is present during the horizontal blanking interval after the line containing the WSS information. BT.656 Ancillary Data Through the BT.656 interface the HMP8117 can generate non-active video data which contains CC, WSS, teletext or CLK VBIVALID P[15-8] 00 FF FF DATA ID BLK # # BYTES/4 BYTE #1 BYTE #2 BYTE #3 BYTE #4 tDVLD NOTES: 24. BT.656 VBI ancillary starts with a 00H, FFH and FFH sequence which is opposite to the SAV/EAV sequence of FFH, 00H and 00H. 25. During active VBI data intervals, DVALID is deasserted and BLANK is asserted. FIGURE 15. OUTPUT TIMING FOR BT.656 VBI DATA TRANSFERS (CC, WSS, TELETEXT, RTCI) TABLE 6. READING THE CLOSED CAPTION DATA AS BT.656 ANCILLARY DATA PIXEL OUTPUT P15 P14 P13 P12 P11 P10 P9 P8 Preamble 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Data ID P14 ep 1 1 0 0 0 0 = odd field data 1 = even field data Data Block Number P14 ep 0 0 0 0 0 1 Data Word Count P14 ep 0 0 0 0 0 1 Caption Data P14 ep 0 0 bit 15 bit 14 bit 13 bit 12 P14 ep 0 0 bit 11 bit 10 bit 9 bit 8 P14 ep 0 0 bit 7 bit 6 bit 5 bit 4 P14 ep 0 0 bit 3 bit 2 bit 1 bit 0 P14 bit 6 bit 5 bit 4 bit 3 bit 2 bit 1 bit 0 CRC NOTES: 26. ep = even parity for P8-P13. 27. CRC = Sum of P8-P14 of Data ID through last user data word. Preset to all zeros, carry is ignored. 17 FN4643.3 April 19, 2007 HMP8117 TABLE 7. OUTPUTTING THE SLICED WSS DATA AS BT.656 ANCILLARY DATA PIXEL OUTPUT P15 P14 P13 P12 P11 P10 P9 P8 Preamble 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Data ID P14 ep 1 1 0 0 1 0 = odd field data 1 = even field data Data Block Number P14 ep 0 0 0 0 0 1 Data Word Count P14 ep 0 0 0 0 1 0 WSS Data P14 ep 0 0 0 0 bit 13 bit 12 P14 ep 0 0 bit 11 bit 10 bit 9 bit 8 P14 ep 0 0 bit 7 bit 6 bit 5 bit 4 P14 ep 0 0 bit 3 bit 2 bit 1 bit 0 P14 ep 0 0 0 0 bit 5 bit 4 P14 ep 0 0 bit 3 bit 2 bit 1 bit 0 P14 ep 0 0 0 0 0 0 P14 ep 0 0 0 0 0 0 P14 bit 6 bit 5 bit 4 bit 3 bit 2 bit 1 bit 0 WSS CRC Data CRC NOTES: 28. ep = even parity for P8-P13. 29. WSS CRC data = “00 0000” during PAL operation. 30. CRC = Sum of P8-P14 of Data ID through last user data word. Preset to all zeros, carry is ignored. Teletext ACCESSING THE TELETEXT DATA The HMP8117 supports ITU-R BT.653 625-line and 525-line teletext system B, C and D capture. NABTS (North American Broadcast Teletext Specification) is the same as BT.653 525-line system C, which is also used to transmit Intel Intercast™ information. WST (World System Teletext) is the same as BT.653 system B. Figure 16 shows the basic structure of a video signal that contains teletext data. The teletext data must be output as BT.656 ancillary data. The I2C interface does not have the bandwidth to output teletext information when needed. The scan lines containing teletext information are monitored. If teletext is enabled and teletext data is present, the teletext data is output as BT.656 ancillary data. DETECTION OF TELETEXT Table 8 illustrates the teletext BT.656 ancillary data format and Figure 15 depicts the portion of the incoming teletext signal which is sliced and output as part of the ancillary data stream. The teletext data is present during the horizontal blanking interval after the line containing the teletext information. The actual BT.656 bytes that contain teletext data only contain 4 bits of the actual data packet. Note that only the data packet of Figure 16 is sent as ancillary data; the clock run-in is not included in the data stream. The teletext decoder monitors the scan lines, looking for the 16-bit clock run-in (sometimes referred to as the clock synchronization code) used by teletext. If found, it locks to the clock run-in, the teletext data is sampled and loaded into shift registers, and the data is then transferred to internal holding registers. If the clock run-in is not found, it is assumed the scan line contains video data unless other VBI information is detected, such as WSS. If a teletext clock run-in is found before line 23 or line 289 for NTSC and (M) PAL, or line 336 for (B, D, G, H, I, N, NC) PAL, the VBI Teletext Detect status bit is immediately set to “1”. If not found by these lines, the status bit is immediately reset to “0”. 18 FN4643.3 April 19, 2007 HMP8117 CLOCK RUN-IN DATA PACKET Bit 0 MSB NOTES: 31. The MSB is bit number: 271 for system C, 279 for system B 525-line and 343 for system B 625-line. 32. The clock run-in is 16 bits wide for both systems and is not included in the BT.656 ancillary data stream. 33. The bit rate is 5.727272 Mbits/s for system B and C on 525/60 systems and 6.9375 and 5.734375 Mbits/second respectively for 625/50 systems. 34. Teletext VBI Video Signal FIGURE 16. TELETEXT VBI VIDEO SIGNAL TABLE 8. OUTPUTTING THE SLICED TELETEXT DATA AS BT.656 ANCILLARY DATA PIXEL INPUT P15 P14 P13 P12 P11 P10 P9 P8 Preamble 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Data ID P14 ep 1 1 0 1 0 0 Data Block Number P14 ep 0 0 0 0 0 1 Data Word Count P14 ep 0 1 0 1 1 0 Teletext Data (B, 625-line = 43 bytes) (B, 525-line = 35 bytes) (C = 34 bytes) P14 ep 0 = 525-line 1 = 625-line 0 = system B 1 = system C bit 343 bit 342 bit 341 bit 340 P14 ep 0 0 bit 339 bit 338 bit 337 bit 336 : Reserved CRC P14 ep 0 0 bit 7 bit 6 bit 5 bit 4 P14 ep 0 0 bit 3 bit 2 bit 1 bit 0 P14 ep 0 0 0 0 0 0 P14 ep 0 0 0 0 0 0 P14 bit 6 bit 5 bit 4 bit 3 bit 2 bit 1 bit 0 NOTES: 35. ep = even parity for P8-P13. 36. CRC = Sum of P8-P14 of Data ID through last user data word. Preset to all zeros, carry is ignored. 37. For 525-line system B, bits 280-343 are “0”. 38. For system C, bits 272-343 are “0”. “RAW” VBI DATA CAPTURE “Raw” data capture of VBI data during blanked scan lines may be optionally implemented. In this instance, the active line time of blanked scan lines are sampled at the CLK2 rate, and output onto the pixel outputs. This permits software decoding of the VBI data to be done. The line mask registers specify on which scan lines to generate “raw” VBI data. If the RAW VBI All bit is enabled, all the video lines are treated as raw VBI data, excluding the equalization and serration lines. The start and end timing of capturing “raw” VBI data on a scan line is determined by the Start and End Raw VBI 19 Registers. This allows the proper capture of “raw” VBI data regardless of the BLANK# output timing for active video. The blanking level is subtracted from the “raw” VBI data samples, and the result is output onto the pixel outputs. Note both “sliced” and “raw” VBI data may be available on the same line. During NTSC operation, the first possible line of VBI data is lines 10 and 272, and the last possible lines are the last blanked scan lines. Lines 1-9 and 264-271 are always blanked. FN4643.3 April 19, 2007 HMP8117 During PAL (B, D, G, H, I, N, NC) operation, the first possible line of VBI data are lines 6 and 318, and the last possible lines are the last blanked scan lines. Lines 623-5 and 311317 are always blanked. TABLE 9. OUTPUTTING RTCI AS BT.656 ANCILLARY DATA PIXEL INPUT P15 P14 P13 P12 P11 P10 P9 P8 Preamble 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Data ID P14 ep 1 1 0 1 0 1 Data Block Number P14 ep 0 0 0 0 0 1 Data Word Count P14 ep 0 0 0 0 1 1 HPLL Increment P14 ep 0 0 0 0 0 0 P14 ep 0 0 0 0 0 0 P14 ep 0 0 0 0 0 0 P14 ep 0 0 0 0 0 0 P14 ep PSW 0 bit 31 bit 30 bit 29 bit 28 P14 ep F2 = 0 F1 = 0 bit 27 bit 26 bit 25 bit 24 FSCPLL Increment : CRC P14 ep 0 0 bit 7 bit 6 bit 5 bit 4 P14 ep 0 0 bit 3 bit 2 bit 1 bit 0 P14 bit 6 bit 5 bit 4 bit 3 bit 2 bit 1 bit 0 NOTES: 39. ep = even parity for P8-P13. 40. CRC = Sum of P8-P14 of Data ID through last user data word. Preset to all zeros, carry is ignored. During PAL (M) operation, the first possible line of VBI data is lines 7 and 269, and the last possible lines are the last blanked scan lines. Lines 523-6 and 261-268 are always blanked. Real Time Control Interface The Real Time Control Interface (RTCI) outputs timing information for a NTSC/PAL encoder as BT.656 ancillary data. This allows the encoder to generate “clean” output video. RTCI information via BT.656 ancillary data is shown in Table 9. If enabled, this transfer occurs once per line and is completed before the start of the SAV sequence. The PSW bit is always a “0” for NTSC encoding. During PAL encoding, it indicates the sign of V (“0” = negative; “1” = positive) for that scan line. Host Interface All internal registers may be written to or read by the host processor at any time, except for those bits identified as read-only. The bit descriptions for the control registers are listed beginning with Table 10. The HMP8117 supports the fast-mode (up to 400kbps) I2C interface consisting of the SDA and SCL pins. The device 20 acts as a slave for receiving and transmitting data over the serial interface. When the interface is not active, SCL and SDA must be pulled high using external 4kΩ pull-up resistors. The SA input pin determines the slave address for the HMP8117. If the SA pin is pulled low, the address is 1000100xB . If the SA pin is pulled high through a 10kΩ pullup resistor, the address is 1000101xB . (This ‘x’ bit in the address is the I2C read flag.) Data is placed on the SDA line when the SCL line is low and held stable when the SCL line is pulled high. Changing the state of the SDA line while SCL is high will be interpreted as either an I2C bus START or STOP condition as indicated by Figure 18. During I2C write cycles, the first data byte after the slave address is treated as the control register sub address and is written into the internal address register. Any remaining data bytes sent during an I2C write cycle are written to the control registers, beginning with the register specified by the address register as given in the first byte. The address register is then auto-incremented after each additional data byte sent on the I2C bus during a write cycle. Writes to reserved bits within registers or reserved registers are ignored. FN4643.3 April 19, 2007 HMP8117 In order to perform a read from a specific control register within the HMP8117, an I2C bus write must first be performed to properly setup the address register. Then an I2C bus read can be performed to read from the desired control register(s). As a result of needing the write cycle for a read cycle there are actually two START conditions as shown in Figure 19. The address register is then auto-incremented after each byte read during the I2C read cycle. Reserved registers return a value of 00H . tSU:DATA tBUF SDA tHD:DATA SCL tLOW tHIGH tR tF tSU:STOP FIGURE 17. I2C TIMING DIAGRAM SDA SCL 1-7 S START CONDITION 8 ADDRESS 9 R/W 1-7 ACK 8 DATA 9 P ACK STOP CONDITION FIGURE 18. I2C SERIAL DATA FLOW DATA WRITE 1000 1000 S CHIP ADDR FROM MASTER A SUB ADDR A 0x88 DATA READ S DATA A REGISTER POINTED TO BY SUB ADDR DATA A P FROM HMP8117 S = START CYCLE P = STOP CYCLE A = ACKNOWLEDGE NA = NO ACKNOWLEDGE OPTIONAL FRAME MAY BE REPEATED n TIMES 1000 1000 (R/W) CHIP ADDR A SUB ADDR A S CHIP ADDR 0x89 0x88 A DATA REGISTER POINTED TO BY SUB ADDR A DATA NA P OPTIONAL FRAME MAY BE REPEATED n TIMES FIGURE 19. REGISTER WRITE/READ FLOW 21 FN4643.3 April 19, 2007 HMP8117 Control Registers R TABLE 10. CONTROL REGISTER SUMMARY SUBADDRESS CONTROL REGISTER RESET/ DEFAULT VALUE USE VALUE COMMENTS 00H Product ID 01H Input Format 19H Defaults to auto-detect of input video standard. 02H Output Format 00H Defaults to 16-bit YCbCr data format. 03H Output Control 00H 04H Genlock Control 09H Defaults to 27MHz CLK2, Rectangular Pixel Mode 05H Analog Input Control 10H Defaults to input signal select = CVBS1. 06H Color Processing 52H 08H Luma Processing 04H 0AH Sliced VBI Data Enable 00H 0BH Sliced VBI Data Output 00H 0CH VBI Data Status 00H 0EH Video Status 00H 0FH Interrupt Mask 00H 10H Interrupt Status 00H 11H Raw VBI Control 00H 12H Raw VBI Start Count 7AH 14H/13H 16H or 17H Raw VBI Stop Count MSB/LSB Raw VBI Line Mask_7_0 FEH 16H Raw VBI Line Mask_15_8 1FH 17H Raw VBI Line Mask_18_16 00H 18H Brightness 00H 19H Contrast 80H 1AH Hue 00H 1BH Saturation 80H 1CH Color Gain Adjust 40H 1DH Video Gain Adjust 80H 1EH Sharpness 10H 1FH Host Control 00H 20H-23H Closed Caption Data Registers 80H 24H-29H WSS Data & CRC Registers 31H/30H Start H_BLANK MSB/LSB 34H/33H End H_BLANK Start V_BLANK MSB/LSB C0H Set Bits 7-6 to enable data and timing outputs. 03H/4AH 15H 32H Returns last two digits of part number in hex format. Set bit 7 for Soft Reset. Set bit 6 for Power Down. 00H 03H/4AH Table 3 BLANK programming changes for each video standard. 7AH Table 3 (same as above) 01H/02H Table 3 (same as above) 35H End V_BLANK 12H Table 3 (same as above) 36H End HSYNC 30H Table 3 (same as above) 90H A wider window tolerates poorly timed video sources. 37H HSYNC Detect Window 20H 41H MV Control 26H 42H Reserved 00H 30H Set bits 5-4 to 11B for optimum performance. 50H Programmable Fractional Gain 0CH 21H A slower PFG improves AGC stability. 51H MV Stripe Gate 14H 52H Reserved 02H 22H Set bit 5 to “1” for optimum performance. 53H AGC Hysteresis 00H F0H Larger hysteresis improves AGC stability. 7FH Device Revision 01H Production baseline revision is 01H. Sub-Addresses: 40H, 43H-4FH are reserved. Reads from these registers may return non-zero values. Sub-Addresses: 07H, 09H, 0DH, 2AH-2FH, 38H-3FH and 54H-7EH are unused. Reads from these registers return 00H. Writes are ignored. 22 FN4643.3 April 19, 2007 HMP8117 TABLE 11. PRODUCT ID REGISTER SUB ADDRESS = 00H BIT NUMBER 7-0 FUNCTION DESCRIPTION Product ID This 8-bit register specifies the last two digits of the product number. Data written to this readonly register is ignored. RESET STATE 17H TABLE 12. INPUT FORMAT REGISTER SUB ADDRESS = 01H BIT NUMBER 7 6-5 FUNCTION DESCRIPTION Reserved RESET STATE 0B Video Timing Standard These bits are read only unless bit 4 = “0”. 00 = (M) NTSC 01 = (B, D, G, H, I, N) PAL 10 = (M) PAL 11 = Combination (N) PAL; also called (NC) PAL 00B 4 Auto Detect Video Standard 0 = Manual selection of video timing standard 1 = Auto detect of video timing standard 1B 3 Setup Select Typically, this bit should be a “1” during (M) NTSC and (M, N) PAL operation. Otherwise, it should be a “0”. 0 = Video source has a 0 IRE blanking pedestal 1 = Video source has a 7.5 IRE blanking pedestal 1B 2-1 0 Reserved 00B Adaptive Sync Slice This bit specifies whether to use fixed or adaptive sync slicing. Adaptive sync slicing Enable automatically determines the midpoint of the sync amplitude to determine timing. 0 = Fixed sync slicing 1 = Adaptive sync slicing 1B TABLE 13. OUTPUT FORMAT REGISTER SUB ADDRESS = 02H BIT NUMBER FUNCTION DESCRIPTION RESET STATE 7-5 Output Color Format 000 = 16-bit 4:2:2 YCbCr100 = 16-bit RGB 001 = 8-bit 4:2:2 YCbCr101 = reserved 010 = 8-bit parallel BT.656110 = reserved 011 = 15-bit RGB111 = reserved 000B 4-3 RGB Gamma Select These bits are ignored except during RGB output modes. 00 = Linear RGB (gamma of input source = 2.2) 01 = Linear RGB (gamma of input source = 2.8) 10 = Gamma-corrected RGB (gamma = gamma of input source) 11 = reserved 00B 2-1 Output Color Select 00 = Normal operation10 = Output black field 01 = Output blue field11 = Output 75% color bars 00B Reserved Set to “0” for proper operation. Vertical Pixel Siting control is not supported. 0B 0 23 FN4643.3 April 19, 2007 HMP8117 TABLE 14. OUTPUT CONTROL REGISTER SUB ADDRESS = 03H BIT NUMBER FUNCTION DESCRIPTION RESET STATE 7 Video Data Output Enable This bit is used to enable the P0-P15 outputs. 0 = Outputs 3-stated. 1 = Outputs enabled 0B 6 Video Timing Output Enable This bit is used to enable the HSYNC, VSYNC, BLANK, FIELD, VBIVALID, DVALID, and INTREQ outputs. 0 = Outputs 3-stated. 1 = Outputs enabled 0B 5 FIELD Polarity 0 = Active low (low during odd fields). 1 = Active high (high during odd fields) 0B 4 BLANK Polarity 0 = Active low (low during blanking). 1 = Active high (high during blanking) 0B 3 HSYNC Polarity 0 = Active low (low during horizontal sync). 1 = Active high (high during horizontal sync) 0B 2 VSYNC Polarity 0 = Active low (low during vertical sync). 1 = Active high (high during vertical sync) 0B 1 DVALID Polarity 0 = Active low (low during valid pixel data). 1 = Active high (high during valid pixel data) 0B 0 VBIVALID Polarity 0 = Active low (low during VBI data). 1 = Active high (high during VBI data) 0B TABLE 15. GENLOCK CONTROL REGISTER SUB ADDRESS = 04H BIT NO. FUNCTION DESCRIPTION RESET STATE 7 Aspect Ratio Mode 0 = Rectangular (BT.601) pixels 1 = Square pixels 0B 6 Freeze Output Timing Enable Setting this bit to a “1” freezes the output timing at the end of the field. Resetting this bit to a “0” resumes normal operation at the start of the next field. 0 = Normal operation 1 = Freeze output timing 0B 5 DVALID Duty Cycle Control (DVLD_DCYC) This bit is ignored during the 8-bit YCbCr and BT.656 output modes. During 16-bit YCbCr, 15-bit RGB, or 16-bit RGB output modes, this bit is defined as: 0 = DVALID has 50/50 duty cycle at the pixel output data rate 1 = DVALID goes active based on line-lock. This will cause DVALID to not have a 50/50 duty cycle. This bit is intended to be used in maintaining backward compatibility with the HMP8112A DVALID output timing. 0B 4 DVALID Line Timing During 16-bit YCbCr, 15-bit RGB, or 16-bit RGB output modes, this bit is defined as: 0 = DVALID present only during active video time on active scan lines Control (DVLD_LTC) 1 = DVALID present the entire scan line time on all scan lines During the 8-bit YCbCr and BT.656 output modes, this bit defines the DVALID output as: 0 = Normal timing 1 = DVALID signal ANDed with CLK2 0B 3 Missing HSYNC Detect Select This bit specifies the number of missing horizontal sync pulses before entering horizontal lock acquisition mode. 0 = 12 pulses 1 = 1 pulse 1B 2 Missing VSYNC Detect Select This bit specifies the number of missing vertical sync pulses before entering vertical lock acquisition mode. 0 = 3 pulses 1 = 1 pulse 0B 1-0 CLK2 Frequency This bit indicates the frequency of the CLK2 input clock. 00 = 24.54MHz10 = 29.5MHz 01 = 27.0MHz11 = Reserved 01B 24 FN4643.3 April 19, 2007 HMP8117 TABLE 16. ANALOG INPUT CONTROL REGISTER SUB ADDRESS = 05H BIT NO. FUNCTION DESCRIPTION RESET STATE 7-6 Lock Loss Video Gain Select If bits 5-4 do not equal “01”, these bits indicate what mode the AGC circuitry will be after loss of sync. If bits 5-4 equal “01”, these bits are ignored. 00 = Automatic gain control: bits 5-4 will be reset to “01” 01 = Maintain fixed gain: bits 5-4 will not be changed 10 = Normal AGC switching to fixed gain after lock achieved: bits 5-4 will not be reset to “01” unless they indicated “freeze automatic gain control” 11 = reserved 00B 5-4 Video Gain Control Select 00 = Fixed 1x gain 01 = Automatic gain control 10 = Fixed gain control. (Use gain factor from Video Gain Adjust register 1DH.) 11 = Freeze automatic gain control 01B Digital Anti-Alias Filter Control 0 = Internal digital anti-alias filter is active. 1 = Internal digital anti- alias filter is bypassed. (Not Recommended) 0B Video Signal Input Select 000 = CVBS1 001 = CVBS2 010 = CVBS3 011 = S-video 1XX = reserved 3 2-0 000B TABLE 17. COLOR PROCESSING REGISTER SUB ADDRESS = 06H BIT NO. FUNCTION DESCRIPTION RESET STATE 7-6 Digital Color Gain Control Select 00 = No gain control (gain = 1x) 01 = Automatic gain control 10 = Fixed gain control. (Use gain factor from Color Gain Adjust register 1CH.) 11 = Freeze automatic gain control 01B 5-4 Color Killer Select 00 = Force color on 01 = Enable color killer 10 = reserved 11 = Force color off 01B 3-2 Color Coring Select Coring may be used to reduce low-level noise in the CbCr signals. 00 = No coring 01 = 1 code coring 10 = 2 code coring 11 = 3 code coring 00B 1 Contrast Control Select This bit specifies whether the contrast control affects just the Y data (“0”) or both the Y and CbCr data (“1”). To avoid color shifts when changing contrast, this bit should be a “1”. 0 = Contrast controls only Y data 1 = Contrast controls Y and CbCr data 1B 0 Color Low-Pass Filter Select This bit selects the bandwidth of the CbCr data. 0 = 850kHz 1 = 1.5MHz 0B 25 FN4643.3 April 19, 2007 HMP8117 TABLE 18. LUMA PROCESSING REGISTER SUB ADDRESS = 08H BIT NO. FUNCTION DESCRIPTION RESET STATE 7-6 Y Filtering Select The chroma trap filter may be used to remove any residual color subcarrier information from the Y channel. During S-video operation, it should be disabled. During PAL operation, it should be enabled. The 3MHz low-pass filter may be used to remove high-frequency noise. 00 = No filtering 01 = Enable chroma trap filter 10 = Enable 3.0MHz low-pass filter 11 = reserved 00B 5-4 Black Level Y Coring Select Coring may be used to reduce low-level noise around black in the Y signal. 00 = No coring 01 = 1 code coring 10 = 2 code coring 11 = 3 code coring 00B 3-2 High Frequency Y Coring Select Coring may be used to reduce high-frequency low-level noise in the Y signal. 00 = No coring 01 = 1 code coring 10 = 2 code coring 11 = 3 code coring 01B 1-0 Sharpness Frequency Select Specifies the amount of sharpness to be applied per the Sharpness Adjust register 1EH. 00 = Bypass sharpness control10 = Maximum gain at color FSC 01 = Maximum gain at 2.6MHz11 = reserved 00B TABLE 19. SLICED VBI DATA ENABLE REGISTER SUB ADDRESS = 0AH BIT NO. FUNCTION DESCRIPTION RESET STATE 7-6 Sliced 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, NC) PAL 10 = Closed caption enabled for even fields: line 284 for NTSC, line 281 for (M) PAL, or line 335 for (B, D, G, H, I, N, NC) PAL 11 = Closed caption enabled for both odd and even fields 00B 5-4 Sliced WSS Enable 00 = WSS disabled 01 = WSS enabled for odd fields: line 20 for NTSC; line 17 for (M) PAL, or line 23 for (B, D, G, H, I, N, NC) PAL 10 = WSS enabled for even fields: line 283 for NTSC, line 280 for (M) PAL, or line 336 for (B, D, G, H, I, N, NC) PAL 11 = WSS enabled for both odd and even fields 00B 3-2 Sliced Teletext Enable 00 = Teletext disabled10 = Teletext system C enabled 01 = Teletext system B enabled11 = Teletext system D enabled 00B 1-0 Reserved 00B 26 FN4643.3 April 19, 2007 HMP8117 TABLE 20. SLICED VBI DATA OUTPUT REGISTER SUB ADDRESS = 0BH BIT NO. FUNCTION DESCRIPTION RESET STATE 7 Sliced Closed Caption BT.656 Output Enable If set to “1”, this bit enables output of sliced closed captioning via BT.656 ancillary data. Closed captioning must be enabled by the Sliced VBI Data Enable register 0AH. Access via the I2C interface is always available. 0B 6 Sliced WSS BT.656 Output Enable If set to “1”, this bit enables output of sliced WSS via BT.656 ancillary data. WSS must be enabled by the Sliced VBI Data Enable register 0AH. Access via the I2C interface is always available. 0B 5 Sliced Teletext BT.656 Output Enable If set to “1”, this bit enables output of sliced teletext via BT.656 ancillary data. Teletext data is not available via the I2C interface. 0B 4-1 0 Reserved 0000B RTCI BT.656 Output Enable If set to “1”, this bit enables output of RTCI data as BT.656 ancillary data. 0B TABLE 21. VBI DATA STATUS REGISTER SUB ADDRESS = 0CH BIT NO. FUNCTION DESCRIPTION RESET STATE 7 CC Odd Field Detect Status This bit is read-only. Data written to this bit is ignored. If set to “1”, Closed Captioning (CC) data is detected on the odd field. 0B 6 CC Even Field Detect Status This bit is read-only. Data written to this bit is ignored. If set to “1”, Closed Captioning (CC) data is detected on the even field. 0B 5 WSS Odd Field Detect Status This bit is read-only. Data written to this bit is ignored. If set to “1”, Wide Screen Signalling (WSS) data is detected on the odd field. 0B 4 WSS Even Field Detect Status This bit is read-only. Data written to this bit is ignored. If set to “1”, Wide Screen Signalling (WSS) data is detected on the even field. 0B 3 VBI Teletext Detect Status This bit is read-only. Data written to this bit is ignored. If set to “1”, Teletext data is detected during the vertical blanking interval. 0B 2-0 Reserved 000B 27 FN4643.3 April 19, 2007 HMP8117 TABLE 22. VIDEO STATUS REGISTER SUB ADDRESS = 0EH BIT NO. FUNCTION DESCRIPTION RESET STATE 7 Vertical Lock Status This bit is read-only. Data written to this bit is ignored. If set to “1”, the decoder is vertically locked to the input signal. 0B 6 Horizontal Lock Status This bit is read-only. Data written to this bit is ignored. If set to “1”, the decoder is horizontally locked to the input signal. 0B 5 Color Lock Status This bit is read-only. Data written to this bit is ignored. If set to “1”, the decoder is chroma locked to the input signal. 0B 4 Input Video Detect Status This bit is read-only. Data written to this bit is ignored. If set to “1”, video is detected on the input signal. 0B 3-1 0 MV Detection Status These bits are read-only. Data written to this bit is ignored. 000B = No MV present 001B = PSP present, No Stripes 010B = PSP present, 2-Line Stripes 011B = PSP present, 4-line Stripes 100B = reserved 101B = reserved 110B = No PSP present, 2-line Stripes (invalid MV scheme, may indicate false detection) 111B = No PSP present, 4-line Stripes (invalid MV scheme, may indicate false detection) This bit is read-only. Data written to this bit is ignored. If set to “1”, the decoder determined the video standard on the input signal. This bit is enabled by the Input Format Register 01H bit 4. Auto Detect Video Standard Status 000B 0B TABLE 23. INTERRUPT MASK REGISTER SUB ADDRESS = 0FH BIT NO. FUNCTION DESCRIPTION RESET STATE 7 Genlock Loss Interrupt Mask If set to “1”, an interrupt is enabled for the loss of genlock. 0B 6 Input Signal Loss Interrupt Mask If set to “1”, an interrupt is enabled for the loss of input video signal. 0B 5 Closed Caption Interrupt Mask If set to “1”, an interrupt is enabled for new data in the closed caption data registers. 0B 4 WSS Interrupt Mask If set to “1”, an interrupt is enabled for new data in the WSS data registers. 0B 3 Teletext Interrupt Mask If set to “1”, an interrupt is enabled for the detection of teletext data in the current field. 0B 2 MV Interrupt Mask If set to “1”, an interrupt is enabled for a change in the MV Detection Status Register 0EH. 0B 1 Auto Detect Video Standard Interrupt Mask If set to “1”, an interrupt is enabled for the successful auto detection of a video standard. 0B 0 Vertical Sync Interrupt Mask If set to “1”, an interrupt is enabled for the start of a new field. 0B 28 FN4643.3 April 19, 2007 HMP8117 TABLE 24. INTERRUPT STATUS REGISTER SUB ADDRESS = 10H BIT NO. FUNCTION DESCRIPTION RESET STATE 7 Genlock Loss Interrupt Status If set to “1”, this bit indicates the interrupt request was due to a loss of genlock. To clear the interrupt request, a “1” must be written to this bit. 0B 6 Input Signal Loss Interrupt Status If set to “1”, this bit indicates the interrupt request was due to a loss of input video signal. To clear the interrupt request, a “1” must be written to this bit. 0B 5 Closed Caption Interrupt Status If set to “1”, this bit indicates the interrupt request was due to new data in the closed caption data registers. To clear the interrupt request, a “1” must be written to this bit. 0B 4 WSS Interrupt Status If set to “1”, this bit indicates the interrupt request was due to new data available in the WSS data registers. To clear the interrupt request, a “1” must be written to this bit. 0B 3 Teletext Interrupt Status If set to “1”, this bit indicates the interrupt request was due to the detection of teletext data in the current field. To clear the interrupt request, a “1” must be written to this bit. 0B 2 MV Interrupt Status If set to “1”, this bit indicates the interrupt request was due to a change in the MV Detection Status of register 0EH. To clear the interrupt request, a “1” must be written to this bit. 0B 1 Auto Detect Video Standard Interrupt Status If set to “1”, this bit indicates the interrupt request was due to the successful auto detection of a video standard. To clear the interrupt request, a “1” must be written to this bit. 0B 0 Vertical Sync Interrupt Status If set to “1”, this bit indicates the interrupt request was due to the start of a new field. To clear the interrupt request, a “1” must be written to this bit. 0B TABLE 25. RAW VBI CONTROL REGISTER SUB ADDRESS = 11H BIT NO. 7-4 FUNCTION DESCRIPTION Reserved RESET STATE 0000B 3 RAW Preamble Enable If set to “1”, enables a four byte preamble in the RAW VBI data stream. The preamble format is FFH, CNT1, CNT2 and 00H, where: CNT1: Bit 7 = even parity bar, Bit 6 = even parity[5-0], Bit 5 = 0, Bit 4 = Field (0=Odd, 1=Even), Bits 3 -0 =Linecount[8-4]. CNT2: Bit 7 = even parity bar, Bit 6 = even parity [5-0], Bits 5-4 = 00 Bits 3-0 = Linecount[3-0]. 0B 2 RAW VBI All If set to “1”, all the video lines (full field) are converted to RAW VBI data. If set to “0”, only the lines enabled by the RAW VBI LINE MASK registers are converted to RAW VBI data. 0B 1 RAW VBI Even Field If set to “1”, even field lines are converted to RAW VBI data as specified by the RAW VBI All bit and the RAW VBI LINE MASK registers. If set to “0”, the even field lines are excluded from the RAW VBI data stream. 0B 0 RAW VBI Odd Field If set to “1”, odd field lines are converted to RAW VBI data as specified by the RAW VBI All bit and the RAW VBI LINE MASK registers. If set to “0”, the odd field lines are excluded from the RAW VBI data stream. 0B 29 FN4643.3 April 19, 2007 HMP8117 TABLE 26. RAW VBI START COUNT REGISTER SUB ADDRESS = 12H BIT NO. 7-0 FUNCTION DESCRIPTION Raw VBI Start Count Specifies the start of the raw VBI data sampling window in two CLK2 period steps from the leading edge of HSYNC. RESET STATE 7AH TABLE 27. RAW VBI STOP COUNT LSB REGISTER SUB ADDRESS = 13H BIT NO. 7-0 FUNCTION DESCRIPTION Raw VBI Stop Count This 8-bit register is cascaded with Raw VBI Stop Count MSB (below) to form a 10-bit stop count LSB value. The stop count specifies the end of the raw VBI data sampling window in two CLK2 period steps from the leading edge of HSYNC. RESET STATE 4AH TABLE 28. RAW VBI STOP COUNT MSB REGISTER SUB ADDRESS = 14H BIT NO. FUNCTION DESCRIPTION 7-2 Reserved 1-0 Raw VBI Stop Count This 2-bit register is cascaded with Raw VBI Stop Count LSB (above) to form a 10-bit stop count MSB value. The stop count specifies the end of the raw VBI data sampling window in two CLK2 period steps from the leading edge of HSYNC. RESET STATE 000000B 11B TABLE 29. RAW VBI LINE MASK_7_0 REGISTER SUB ADDRESS = 15H BIT NO. 7-0 FUNCTION DESCRIPTION Raw VBI Line Mask_7_0 A “1” in each bit position enables raw VBI capture for a corresponding input video line. Refer to Table 32 below. RESET STATE FEH TABLE 30. RAW VBI LINE MASK_15_8 REGISTER SUB ADDRESS = 16H BIT NO. 7-0 FUNCTION DESCRIPTION Raw VBI Line Mask_15_8 A “1” in each bit position enables raw VBI capture for a corresponding input video line. Refer to Table 32 below. RESET STATE 1FH TABLE 31. RAW VBI LINE MASK_18_16 REGISTER SUB ADDRESS = 17H BIT NO. FUNCTION 7-3 Reserved 2-0 Raw VBI Line Mask_18_16 DESCRIPTION RESET STATE 00000B A “1” in each bit position enables raw VBI capture for a corresponding input video line. Refer to Table 32 below. 30 000B FN4643.3 April 19, 2007 HMP8117 TABLE 32. RAW VBI MASK DEFINITIONS MASK (Register = Default) MASK_18_16 (Reg. 17H) MASK_15_8 (Register 16H) MASK_7_0 (Register 15H) REGISTER BIT 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 Mask Bit 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 NTSC (Odd) Line# 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 NTSC (Even) Line# 290 289 288 287 286 285 284 283 282 281 280 279 278 277 276 275 274 273 272 PAL (Odd) Line# 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 PAL (Even) Line# 336 335 334 333 332 331 330 329 328 327 326 325 324 323 322 321 320 319 318 TABLE 33. BRIGHTNESS REGISTER SUB ADDRESS = 18H BIT NO. FUNCTION 7 Reserved 6-0 Brightness Adjust DESCRIPTION RESET STATE 0B These bits control the brightness. They may have a value of +63 (“011 1111”) to -64 (“100 0000”), with positive values increasing brightness. A value of 0 (“000 0000”) has no effect on the data. 0000000B TABLE 34. CONTRAST REGISTER SUB ADDRESS = 19H BIT NO. 7-0 FUNCTION DESCRIPTION Contrast Adjust These bits control the contrast. They may have a value of 0x (“0000 0000”) to 1.992x (“1111 1111”). A value of 1x (“1000 0000”) has no effect on the data. RESET STATE 80H TABLE 35. HUE REGISTER SUB ADDRESS = 1AH BIT NO. 7-0 FUNCTION DESCRIPTION Hue Adjust These bits control the color hue. They may have a value of +30 degrees (“0111 1111”) to -30 degrees (“1111 1111”). A value of 0 degrees (“0000 0000”) has no effect on the color data. RESET STATE 00H TABLE 36. SATURATION REGISTER SUB ADDRESS = 1BH BIT NO. 7-0 FUNCTION DESCRIPTION Saturation Adjust These bits control the color saturation. They may have a value of 0x (“0000 0000”) to 1.992x (“1111 1111”). A value of 1x (“1000 0000”) has no effect on the color data. A value of 0x (“0000 0000”) disables the color information. RESET STATE 80H TABLE 37. COLOR GAIN ADJUST REGISTER SUB ADDRESS = 1CH BIT NO. 7-0 FUNCTION DESCRIPTION Color Gain Adjust These bits control the amount of digital gain applied to the color difference (CbCr) signals. They may have a value of 0.5x (“0010 0000”) to 3.98x (“1111 1111”). A value of 1x (“0100 0000”) has no effect on the data. This register enabled by the selection of “fixed gain control” mode in the Color Processing register 06H. 31 RESET STATE 40H FN4643.3 April 19, 2007 HMP8117 TABLE 38. VIDEO GAIN ADJUST REGISTER SUB ADDRESS = 1DH BIT NO. 7-0 FUNCTION RESET STATE DESCRIPTION Video Gain Adjust This register is enabled by the selection of “fixed gain control” mode in the Analog Input Control register 05H bits 7-4. The value of this register selects a combined analog attenuation and a digital gain factor which is applied to both Luma and Chroma input channels. The gain factor is selected from nonlinear lookup table and may range in value from 0.5x (CEH) to 1.99x (33H). Refer to Table 39 below. The Register Values in italics below mark the approximate analog attenuation ladder switching points. 80H TABLE 39. VIDEO GAIN REGISTER LOOKUP TABLE Video Gain Reg. Value Video Gain Reg. Value Video Gain Reg. Value Video Gain Reg. Value Video Gain Reg. Value Video Gain Reg. Value 0.50 0.51 0.52 0.53 0.54 0.55 206/CEH 202/CAH 197/C5H 193/C1H 191/BFH 187/BBH 0.67 0.68 0.69 0.70 0.71 0.72 153/99H 151/97H 150/96H 147/93H 145/91H 143/8FH 0.84 0.85 0.86 0.87 0.88 0.89 123/7BH 121/79H 119/77H 118/76H 117/75H 115/73H 1.03 1.04 1.05 1.06 1.07 1.08 100/64H 99/63H 98/62H 97/61H 96/60H 95/5FH 1.23 1.25 1.27 1.28 1.30 1.31 83/53H 82/52H 81/51H 80/50H 79/4FH 78/4EH 1.55 1.57 1.59 1.63 1.65 1.67 66/42H 65/41H 64/40H 63/3FH 62/3EH 61/3DH 0.56 0.57 0.58 0.59 0.60 0.61 183/B7H 180/B4H 178/B2H 174/AEH 171/ABH 169/A9H 0.73 0.74 0.75 0.76 0.77 0.78 141/8DH 139/8BH 137/89H 136/88H 134/86H 132/84H 0.90 0.91 0.92 0.94 0.95 0.96 114/72H 113/71H 111/6FH 110/6EH 109/6DH 107/6BH 1.09 1.10 1.12 1.13 1.14 1.15 94/5EH 93/5DH 92/5CH 91/5BH 90/5AH 89/59H 1.33 1.34 1.37 1.38 1.40 1.42 77/4DH 76/4CH 75/4BH 74/4AH 73/49H 72/48H 1.70 1.73 1.76 1.79 1.82 1.86 60/3CH 59/3BH 58/3AH 57/39H 56/38H 55/37H 0.62 0.63 0.64 0.65 0.66 167/A7H 164/A4H 161/A1H 159/9FH 156/9CH 0.79 0.80 0.81 0.82 0.83 130/82H 128/80H 126/7EH 125/7DH 124/7CH 0.97 0.98 1.00 1.01 1.02 106/6AH 104/68H 103/67H 102/66H 101/65H 1.16 1.18 1.20 1.21 1.22 88/58H 87/57H 86/56H 85/55H 84/54H 1.44 1.46 1.48 1.51 1.52 71/47H 70/46H 69/45H 68/44H 67/43H 1.89 1.93 1.97 1.99 54/36H 53/35H 52/34H 51/33H TABLE 40. SHARPNESS ADJUST REGISTER SUB ADDRESS = 1EH BIT NO. FUNCTION 7-6 Reserved 5-0 Sharpness Adjust DESCRIPTION RESET STATE 00B Specifies the amount of high frequency gain control for luminance signals (either 2.6MHz or FSC), as determined by the Luma Processing register 08H. The gain ranges from +12dB (11 1111B) to -12dB (00 0100B). A value of 0dB (“01 0000”) has no effect on the data. 32 010000B FN4643.3 April 19, 2007 HMP8117 TABLE 41. HOST CONTROL REGISTER SUB ADDRESS = 1FH BIT NO. FUNCTION DESCRIPTION RESET STATE 7 Software Reset When this bit is set to 1, the entire device except the I2C bus is reset to a known state exactly like the RESET input going active. The software reset will initialize all register bits to their reset state. Once set this bit is self clearing. This bit is cleared on power-up by the external RESET pin. 0B 6 Power Down When this bit is set to a 1, the entire device is shut down except the I2C bus by gating off the clock. For normal decoding operations this bit should be set to a 0. 0B 5 Closed Caption Odd Field Read Status This bit is read-only. Data written to this bit is ignored. The bit is cleared when the caption data has been read out via the I2C interface or as BT.656 ancillary data. 0 = No new caption data 1 = Caption_ODD_A and Caption_ODD_B data registers contain new data. 0B 4 Closed Caption Even Field Read Status This bit is read-only. Data written to this bit is ignored. The bit is cleared when the caption data has been read out via the I2C interface or as BT.656 ancillary data. 0 = No new caption data 1 = Caption_EVEN_A and Caption_EVEN_B data registers contain new data. 0B 3 WSS Odd Field Read Status This bit is read-only. Data written to this bit is ignored. The bit is cleared when the WSS data has been read out via the I2C interface or as BT.656 ancillary data. 0 = No new WSS data 1 = WSS_ODD_A and WSS_ODD_B data registers contain new data. 0B 2 WSS Even Field Read Status This bit is read-only. Data written to this bit is ignored. The bit is cleared when the WSS data has been read out via the I2C interface or as BT.656 ancillary data. 0 = No new WSS data 1 = WSS_EVEN_A and WSS_EVEN_B data registers contain new data. 0B 1-0 Reserved 00B TABLE 42. CLOSED CAPTION_ODD_A DATA REGISTER SUB ADDRESS = 20H BIT NO. 7-0 FUNCTION DESCRIPTION Odd Field Caption Data If odd field captioning is enabled and present, this register is loaded with the first eight bits of caption data on line 18, 21, or 22. Bit 0 corresponds to the first bit of caption information. Data written to this register is ignored. RESET STATE 80H TABLE 43. CLOSED CAPTION_ODD_B DATA REGISTER SUB ADDRESS = 21H BIT NO. 15-8 FUNCTION DESCRIPTION Odd Field Caption Data If odd field captioning is enabled and present, this register is loaded with the second eight bits of caption data on line 18, 21, or 22. Data written to this register is ignored. RESET STATE 80H TABLE 44. CLOSED CAPTION_EVEN_A DATA REGISTER SUB ADDRESS = 22H BIT NO. 7-0 FUNCTION DESCRIPTION Even Field Caption Data If even field captioning is enabled and present, this register is loaded with the first eight bits of caption data on line 281, 284, or 335. Bit 0 corresponds to the first bit of caption information. Data written to this register is ignored. 33 RESET STATE 80H FN4643.3 April 19, 2007 HMP8117 TABLE 45. CLOSED CAPTION_EVEN_B DATA REGISTER SUB ADDRESS = 23H BIT NO. 15-8 FUNCTION DESCRIPTION Even Field Caption Data If even field captioning is enabled and present, this register is loaded with the second eight bits of caption data on line 281, 284, or 335. Data written to this register is ignored. RESET STATE 80H TABLE 46. WSS_ODD_A DATA REGISTER SUB ADDRESS = 24H BIT NO. 7-0 FUNCTION DESCRIPTION Odd Field WSS Data If odd field WSS is enabled and present, this register is loaded with the first eight bits of WSS information on line 17, 20, or 23. Bit 0 corresponds to the first bit of WSS information. Data written to this register is ignored. RESET STATE 00H TABLE 47. WSS_ODD_B DATA REGISTER SUB ADDRESS = 25H BIT NO. FUNCTION 15-14 Reserved 13-8 Odd Field WSS Data DESCRIPTION RESET STATE 00B If odd field WSS is enabled and present, this register is loaded with the second six bits of WSS information on line 17, 20, or 23. Data written to this register is ignored. 000000B TABLE 48. WSS_CRC_ODD DATA REGISTER SUB ADDRESS = 26H BIT NO. FUNCTION 7-6 Reserved 5-0 Odd Field WSS CRC Data DESCRIPTION RESET STATE 00B If odd field WSS is enabled and present during NTSC operation, this register is loaded with the six bits of CRC information on line 20. It is always a “000000” during PAL operation. Data written to this register is ignored. 000000B TABLE 49. WSS_EVEN_A DATA REGISTER SUB ADDRESS = 27H BIT NO. 7-0 FUNCTION DESCRIPTION Even Field WSS Data If even field WSS is enabled and present, this register is loaded with the first eight bits of WSS information on line 280, 283, or 336. Bit 0 corresponds to the first bit of WSS information. Data written to this register is ignored. RESET STATE 00H TABLE 50. WSS_EVEN_B DATA REGISTER SUB ADDRESS = 28H BIT NO. FUNCTION 15-14 Reserved 13-8 Even Field WSS Data DESCRIPTION RESET STATE 00B If even field WSS is enabled and present, this register is loaded with the second six bits of WSS information on line 280, 283, or 336. Data written to this register is ignored. 34 000000B FN4643.3 April 19, 2007 HMP8117 TABLE 51. WSS_CRC_EVEN DATA REGISTER SUB ADDRESS = 29H BIT NO. FUNCTION 7-6 Reserved 5-0 Even Field WSS CRC Data DESCRIPTION RESET STATE 00B If even field WSS is enabled and present during NTSC operation, this register is loaded with the six bits of CRC information on line 283. It is always a “000000” during PAL operation. Data written to this register is ignored. 000000B TABLE 52. START H_BLANK LSB REGISTER SUB ADDRESS = 30H BIT NO. 7-0 FUNCTION DESCRIPTION Assert BLANK Output Signal This 8-bit register is cascaded with Start H_BLANK High Register to form a 10-bit start horizontal blank REGISTER. It specifies the horizontal count (in 1x clock cycles) at which to assert BLANK each scan line. Bit 0 is always a “0”, so the start of horizontal blanking may only be done with two pixel resolution. The leading edge of HSYNC is count 000H. RESET STATE 4AH TABLE 53. START H_BLANK MSB REGISTER SUB ADDRESS = 31H BIT NO. 15-10 9-8 FUNCTION DESCRIPTION Reserved RESET STATE 000000B Assert BLANK Output Signal This 2-bit register is cascaded with Start H_BLANK Low Register to form a 10-bit start horizontal blank register. It specifies the horizontal count (in 1x clock cycles) at which to assert BLANK each scan line. The leading edge of HSYNC is count 000H. 11B TABLE 54. END H_BLANK REGISTER SUB ADDRESS = 32H BIT NO. 7-0 FUNCTION DESCRIPTION Negate BLANK Output Signal This 8-bit register specifies the horizontal count (in 1x clock cycles) to negate BLANK each scan line. For proper operation, bit 0 must always be set to “0”; therefore, the end of horizontal blanking may only set with two pixel resolution. If bit 0 is set to “1”, the chroma/luma output data may be swapped. The leading edge of HSYNC is count 000H. RESET STATE 7AH TABLE 55. START V_BLANK LSB REGISTER SUB ADDRESS = 33H BIT NO. 7-0 FUNCTION DESCRIPTION Assert BLANK Output Signal This 8-bit register is cascaded with Start V_BLANK High Register to form a 9-bit start vertical blank register. It specifies the line number to assert BLANK each field. For NTSC operation, it occurs on line (n + 5) on odd fields and line (n + 268) on even fields. For PAL operation, it occurs on line (n + 5) on odd fields and line (n + 318) on even fields. RESET STATE 02H TABLE 56. START V_BLANK MSB REGISTER SUB ADDRESS = 34H BIT NO. 15-9 8 FUNCTION DESCRIPTION Reserved RESET STATE 0000000B Assert BLANK Output Signal This 1-bit register is cascaded with Start V_BLANK Low Register to form a 9-bit start vertical blank register. 35 1B FN4643.3 April 19, 2007 HMP8117 TABLE 57. END V_BLANK REGISTER SUB ADDRESS = 35H BIT NO. 7-0 FUNCTION DESCRIPTION Negate BLANK Output Signal This 8-bit register specifies the line number to negate BLANK each field. For NTSC operation, it occurs on line (n + 5) on odd fields and line (n + 268) on even fields. For PAL operation, it occurs on line (n + 5) on odd fields and line (n + 318) on even fields. RESET STATE 12H TABLE 58. END HSYNC REGISTER SUB ADDRESS = 36H BIT NO. 7-0 FUNCTION DESCRIPTION Negate HSYNC Output Signal This 8-bit register specifies the horizontal count at which to negate HSYNC each scan line. Values may range from 0 (0000 0000) to 510 (1111 1111) CLK2 cycles. The leading edge of HSYNC is count 00H. RESET STATE 30H TABLE 59. HSYNC DETECT WINDOW REGISTER SUB ADDRESS = 37H BIT NO. 7-0 FUNCTION DESCRIPTION Horizontal Sync Detect Window This 8-bit register specifies the width of the timing window (in 1x clock samples) for the digital PLL to accept horizontal sync pulses in each line. The window is centered about where the horizontal sync pulse should be located. If the horizontal sync pulse falls inside the window, the digital PLL maintains normal lock timing. If the horizontal sync pulse falls outside this window, the digital PLL will to enter the horizontal lock acquisition mode based on the current setting for bits 3-2 of register 04H. Recommend changing this register to 90H following reset in order to widen the window for poorly timed input video sources. RESET STATE 20H (Use 90H) TABLE 60. MV CONTROL SUB ADDRESS = 41H BIT NO. FUNCTION DESCRIPTION RESET STATE 7 MV Stripe Detection Set to “1” to enable the detection and bypass of the MV Color Striping component. If this bit is and Bypass Enable not enabled and the MV Color Striping component exists on the input signal, artifacts will be clearly visible as horizontal streaks in the output data. This bit must be enabled for the MV Detection Status of register 0EH to be updated. 0B 6 MV PSP Detection Enable Set to “1” to enable detection of the MV Pseudo Sync Pulse (PSP) component. If the MV PSP component exists on the input signal, this bit must be enabled for the MV Detection Status of register 0EH to be updated. 0B 5-3 MV PSP Detection Count Defines the number of extra sync pulses required before declaring the Pseudo Sync Pulse (PSP) component in the MV Detection Status of register 0EH. The PSP component must also be present for the number of fields defined in bits 2-0 below. 100B 2-0 MV Detection Field Count Defines the minimum number of fields that an MV component must be present for in order to change the MV Detection Status of register 0EH. Add 2 to bits 2-0 to obtain the minimum field count. Ex: The default of 110B is actually 6 + 2 = 8 fields. 110B TABLE 61. RESERVED SUB ADDRESS = 42H BIT NO. 7-0 FUNCTION DESCRIPTION Reserved Set bits 5-4 to 11B for optimum performance. 36 RESET STATE 00H (Use 30H) FN4643.3 April 19, 2007 HMP8117 TABLE 62. PROGRAMMABLE FRACTIONAL GAIN SUB ADDRESS = 50H BIT NO. 7-6 5 4-0 FUNCTION DESCRIPTION Reserved Set to 00B for proper operation. Select PFG Enable Set to “1” to enable the recommended PFG value in bits 4-0 below. PFG Programmable Fractional Gain (PFG). When enabled by bit 5, changes the loop gain (response time) of the AGC logic. Slower values provide some noise immunity to input signals with poor sync/back-porch characteristics. Recommend using the slowest PFG value of 00001B for optimum performance. (Thus the recommended 8-bit register value = 21H). RESET STATE 00B 0B (Use 1B) 01100B (Use 01H) The 5-bit PFG value has a fractional representation as: 20. 2-1 2-2 2-3 2-4 Sample PFG values: 00000B: Gain = 0.0000 (freezes AGC at current value) 00001B: Gain = 0.0625 (slowest AGC response time -- recommended PFG) 01100B: Gain = 0.7500 (default) 10000B: Gain = 1.0000 11111B: Gain = 1.9375 (fastest AGC response time -- not recommended) TABLE 63. MV STRIPE GATE SUB ADDRESS = 51H BIT NO. FUNCTION DESCRIPTION 7-6 Reserved Set to 00B for proper operation. 5-0 MV Stripe Gate Defines the start of the gate for MV Color Stripe detection in 4xFSC counts. The gate should start prior to the chroma burst. Default value of 010100B (14H) is valid for NTSC. Recommend 100000B (20H) for PAL. RESET STATE 00B 010100B TABLE 64. RESERVED SUB ADDRESS = 52H BIT NO. 7-0 FUNCTION DESCRIPTION Reserved Set bit 5 to “1” for optimum performance. RESET STATE 02H (Use 22H) TABLE 65. AGC HYSTERESIS SUB ADDRESS = 53H BIT NO. FUNCTION DESCRIPTION 7-4 3-0 Defines the amount of hysteresis in the AGC logic. Larger hysteresis values stabilize the AGC with poor quality input signals. For example: 0000B = No Hysteresis 1000B = Default Hysteresis 1111B = Maximum Hysteresis (Recommended hysteresis value) Reserved Set to 0000B for proper operation. RESET STATE 1000B (Use F0H) 0000B TABLE 66. DEVICE REVISION SUB ADDRESS = 7FH BIT NO. FUNCTION 7-0 Device Revision DESCRIPTION This 8-bit register specifies the device revision number. Data written to this read-only register is ignored. The production baseline revision number is 01H. 37 RESET STATE 01H FN4643.3 April 19, 2007 HMP8117 Pinout GND VCC REF_CAP NC LCAP VCC NC NC GND HSYNC VSYNC GND VCC FIELD DVALID BLANK 80 LEAD PQFP TOP VIEW 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 AGND VAA AGND NC CVBS3(Y) CVBS2 CVBS1 YIN YOUT AGND AGND VAA NC VAA AGND AGND A/D_TEST NC C NC AGND AGND AGND AGND AGND 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 P15 P14 GND VBIVALID P13 VCC P12 P11 P10 P9 P8 GND VCC P7 P6 P5 P4 P3 GND P2 INTREQ P1 P0 SCL GND VCC SA RSET CCAP NC VCC NC GND RESET GND GND VCC CLK2 GND SDA 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Pin Descriptions PIN NAME PIN NUMBER CVBS1, CVBS2, CVBS3(Y) 7, 6, 5 YOUT 9 I/O I O YIN 8 C 19 A/D TEST 17 RSET 28 REF_CAP 78 PASSIVE DESCRIPTION Composite Video Inputs. CVBS3(Y) is the Luminance (Y) signal in S-Video mode. These inputs must each be terminated by a 75Ω resistor to AGND and AC-coupled by a 1.0µF 75Ω Term, capacitor as shown in the Reference Schematic. These components should be as close to 1μF AC-coupled this pin as possible for best performance. If not used, this pin should be connected to AGND through a 0.1μF capacitor. External AntiAlias Filter I Analog output of the video multiplexer. A external low-pass anti-alias filter between the YOUT and YIN pins, as shown in the Reference Schematic. The filter components should be as close as possible to the YOUT and YIN pins for best performance. Analog input to the ADC. See YOUT description above. Chrominance (C) S-Video input. This input must each be terminated by a 75Ω resistor to AGND and AC-coupled by a 1.0μF capacitor as shown in the Reference Schematic. These components, and the corresponding anti-alias low-pass filter, should be as close to this pin as possible for best performance. If not used, this pin should be connected to AGND through a 0.1μF capacitor. I 75Ω Term, 1μF AC-coupled and External AntiAlias Filter O none O 12.1kΩ to AGND A 12.1kΩ resistor must be connected between this pin and AGND. This resistor should be as close to this pin as possible for best performance. The function of this pin has changed from the HMP8112A/15 GAIN_CNTRL input. Do not use capacitor decoupling for this output pin. O 1.0µF to AGND Voltage reference capacitor. A 1µF ceramic capacitor must be connected between this pin and AGND. This capacitor should be as close to this pin as possible for best performance. 38 Chroma signal A/D test pin. This pin must be left floating for proper operation. FN4643.3 April 19, 2007 HMP8117 Pin Descriptions PIN NAME PIN NUMBER LCAP 76 CCAP P0-P15 HSYNC VSYNC 42, 43, 45, 47-51, 54-58, 60, 63, 64 PASSIVE DESCRIPTION I 0.1µF to AGND Storage capacitor for Luminance signal DC restoration. The LCAP voltage offsets the sync tip to the lower reference of the A/D. A 0.1μF capacitor should be connected between this pin and AGND. This capacitor should be as close to this pin as possible for best performance. I 0.1µF to AGND Storage capacitor for Chrominance signal DC restoration. The CCAP voltage offsets the chroma signal to mid-range of the A/D. A 0.1μF capacitor should be connected between this pin and AGND. This capacitor should be as close to this pin as possible for best performance. O N/A O 10kΩ Pullup Horizontal sync output. HSYNC is asserted during the horizontal sync intervals. The polarity of HSYNC is programmable. This pin is three-stated after a RESET or software reset and should be pulled high through a 10kΩ resistor. O 10kΩ Pullup Vertical sync output. VSYNC is asserted during the vertical sync intervals. The polarity of VSYNC is programmable. This pin is three-stated after a RESET or software reset and should be pulled high through a 10kΩ resistor. O 10kΩ Pullup FIELD output. The polarity of FIELD is programmable. This pin is three-stated after a RESET or software reset and should be pulled high through a 10kΩ resistor. O 10kΩ Pullup Data valid output. DVALID is asserted during CLK2 cycles that contain valid pixel data. This pin is three-stated after a RESET or software reset and should be pulled high through a 10kΩ resistor. O 10kΩ Pullup Composite blanking output. BLANK is asserted during the horizontal and vertical blanking intervals. The polarity of BLANK is programmable. This pin is three-stated after a RESET or software reset and should be pulled high through a 10kΩ resistor. O 10kΩ Pullup Vertical Blanking Interval Valid output. VBIVALID is asserted during CLK2 cycles that contain valid VBI (Vertical Blanking Interval) data such as Closed Captioning, Teletext, and WSS data. The polarity of VBIVALID is programmable. This pin is three-stated after a RESET or software reset and should be pulled high through a 10k resistor. O 10kΩ Pullup Interrupt Request Output. This is an open-drain output and requires an external 10kΩ pullup resistor to VCC. Pixel output pins. See Table 3. These pins are three-stated after a RESET or software reset. 71 70 67 DVALID 66 VBIVALID I/O 29 FIELD BLANK (Continued) 65 61 INTREQ 44 CLK2 38 RESET 34 SA 27 I 2x pixel clock input. This clock must be a continuous, free-running clock. Refer to Table 1 for allowable CLK2 frequencies for each video standard and aspect ratio. For best performance, use termination resistor(s) to minimize pulse overshoot and reflections. I Reset control input. A logical zero for a minimum of four CLK2 cycles resets the device. RESET must be a logical one for normal operation. I 10kΩ Pullup or 0Ω Pulldown I2C slave address select input. This was formerly the WPE pin on HMP8112/15 decoders. If the SA pin is pulled low, the I2C address is 1000100xB or 88H. If the SA pin is pulled high, the address is 1000101xB or 8AH. (The ‘x’ bit is the address is the I2C read flag.) SDA 40 I/O 4kΩ Pullup I2C data input/output. This pin should be pulled high through a 4kΩ resistor. SCL 41 I 4kΩ Pullup I2C clock input. This pin should be pulled high through a 4kΩ resistor. VAA 2, 12, 14 0.1µF to AGND Analog power supply pins. All VAA pins must be connected together. I 1, 3, 10, 11, 15,16, 21, 22, 23, 24 I none VCC 26, 31,37, 52, 59, 68, 75, 79 I GND 25, 33, 35, 36, 39, 46, 53, 62, 69, 72, 80 I AGND NC Analog ground pins. All AGND pins must be connected together. Refer to Applications section for recommended grounding scheme. Digital power supply pins. All VCC pins must be connected together. Digital ground pins. All GND pins must be connected together. No Connect pins. These pins may be left floating or tied to GND. 4, 13, 18, 20, 30, 32, 73, 74, 77 39 FN4643.3 April 19, 2007 HMP8117 Applications Information Direct Interface to Video Encoders Direct interface to a video encoder will induce pixel jitter in the output video and is therefore not recommended as a primary data interface. The jitter will occur with all decoder output formats, including BT.656. However, pixel jitter may be acceptable for some applications; such as a “preview mode” prior to image capture or compression. For more detail, reference“Cycle Slipping and Real-Time Pixel Jitter” on page 7. Decoder Upgrades The following table describe the impacts to pins for upgrading from the HMP8112/A or HMP8115 to the HMP8117. TABLE 67. UPGRADING FROM HMP8112/A OR HMP8115 Pin # HMP8112/15 Pin HMP8117 Impact 28 GAIN_CNTL (Now RSET) Use single 12.1k resistor to AGND. Remove any decoupling caps. 78 DEC_T (Now REF_CAP) Change to single 1.0uF capacitor (1206-size XR7-type) to AGND. 29 CCLAMP_CAP (Now CCAP) Change to 0.1uF capacitor. 76 LCLAMP_CAP (Now LCAP) Change to 0.1uF capacitor. 9,8,19 L_OUT, L_ADIN, and C Recommend use of new anti-alias filter from Reference Schematic. 27 WPE (Now SA) Pull low for I2C address compatibility with HMP8112/A. 44 DVCC/NC (Now INTREQ) Pin actually NC on HMP8112/A. Float or use 10K pullup to VCC. 61 DGND/NC (Now VBIVALID) Pin actually NC on HMP8112/A. Float or use 10K pullup to VCC. 13 CLK2 (Now NC) Trace may be deleted to reduce reflections on CLK2 at pin 38. 30, 32, DEC_L, DGND, Pins no longer used (NC). Capacitors 73, 77 DGND, AGC_CAP used at these pins may be removed. Typical Programming Sequence The following pseudo code provides a typical programming sequence to initialize the HMP8117 using the default 16-bit YCbCr output data format. SetReg 0x1F = 0x80 // Soft Reset SetReg 0x37 = 0x90 // Wider HSYNC Detect Window SetReg 0x42 = 0x30 // Recommended Value SetReg 0x50 = 0x21 // Slower PFG SetReg 0x52 = 0x22 // Recommended Value SetReg 0x53 = 0xF0 // Large AGC Hysteresis SetReg 0x03 = 0xC0 // Enable Data/Timing Outputs PCB Layout 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 40 implement the lowest possible noise on the power and ground planes by providing excellent decoupling. The optimum layout places the HMP8117 as close as possible to the power supply connector and the video input connector. Place external components as close as possible to the appropriate pin using short, wide traces. ANALOG POWER PLANE The analog power plane (VAA) is recommended to be separate from the common board digital power plane (VCC) with a gap between the two planes of at least 1/8 inch. The VAA plane should be connected to the VCC 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 switching currents, improving the performance of HMP8117. If a separate linear regulator is used to provide power to the analog power plane, the power-up sequence should be designed to ensure latch up will not occur. A separate linear regulator is recommended if the power supply noise on the VAA pins exceeds 200mV. ANALOG GROUND PLANE A separate analog ground (AGND) plane is recommended with a single point connection to the digital ground (GND) plane using a ferrite bead as mentioned above. POWER SUPPLY DECOUPLING Decouple each VAA and VCC pin to the appropriate ground plane using a 0.1μF ceramic chip capacitor. Bulk decouple the power planes using a 1.0μF ceramic chip capacitor located at each corner of the device. (One capacitor placed at the top left corner for VAA and three capacitors placed at the other corners for VCC.) A single 47μF decoupling capacitor for the analog power plane may also be used to control excessive low-frequency power supply ripple. See Figure 20, HMP8117 Reference Schematics. ANALOG SIGNALS Traces containing digital signals should not be routed over, under, or adjacent to the analog output traces to minimize cross-talk. If this is not possible, coupling can be minimized by routing the digital signals at a 90 degree angle to the analog signals. The analog traces should also not overlay the VCC power plane to maximize high-frequency power supply rejection. Evaluation Board HMPVIDEVAL/ISA The HMPVIDEVAL/ISA board provides a complete video frame-grabber platform to evaluate all modes of the video decoder and encoder. The ISA style PC add-in board supports a complete Windows 95 software application to easily operate all features of the evaluation platform. FN4643.3 April 19, 2007 HMP8117 Related Application Notes Application Notes are also available on the Intersil Multimedia web site at http://www.intersil.com/mmedia. AN9644: Composite Video Separation Techniques AN9716: Wide Screen Signalling AN9717: YCbCr to RGB Considerations AN9728: BT.656 Video Interface for ICs AN9806: Advantages of the HMP8117 Videolyzer Operation EXTERNAL 75Ω VIDEO SOURCES 75 VID1 C1 1.0μF 7 VID2 C2 1.0μF 6 Y_IN C3 1.0μF 5 75 75 R3 75 R2 75 CVBS1 P15 P14 P13 P12 P11 P10 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 CVBS2 CVBS3(Y) R1 75 LUMA ANTI-ALIAS FILTER 9 L1 R4 8.2μH 324 YIN HMP8117 L2 R7 C6 324 1.0μF C7 22pF 19 8.2μH C9 0.1μF C C8 47pF 76 LCAP 29 CCAP C10 0.1μF 78 C11 1.0μF 28 R9 4K 71 HSYNC VSYNC 70 VSYNC FIELD 67 FIELD DVALID 66 DVALID BLANK 65 BLANK VBIVALID 61 VBIVALID SDA 40 SCL 41 SA 27 RESET 34 REF_CAP RSET VCC L3 R11 38 BULK DECOUPLING RESET CLK2 SERIES TERMINATION NEAR CLK2 SOURCE 100 DIGITAL - VCC DECOUPLING C18 C19 C20 C21 C22 C23 C24 C25 C26 1.0μF 0.1μF 0.1μF 0.1μF 0.1μF 0.1μF 0.1μF 0.1μF 0.1μF BOT LEFT PIN 26 PIN 31 PIN 37 PIN 52 PIN 59 PIN 68 PIN 72 PIN 79 NEAR FERRITE BOT RIGHT 47μF TOP LEFT 1.0μF C27 1.0μF PIN 14 C17 C15 0.1μF PIN 12 GND SCL JP1: ADDRESS SELECT JUMPER TOP RIGHT C14 0.1μF PIN 1 FERRITE BEAD SDA 1.0μF C13 0.1μF L4 INTREQ C16 C12 FERRITE BEAD AGND HSYNC CLK2 VAA R10 4K INTREQ 44 R8 12.1K ANALOG - VAA DECOUPLING VCC VCC VCC RP1 10K CHROMA ANTI-ALIAS FILTER C_IN R6 75 YOUT R5 4.02K C5 47pF C4 22pF FROM 75Ω SOURCE 8 P[15:0] 64 63 60 58 57 56 55 54 51 50 49 48 47 45 43 42 FIGURE 20. HMP8117 REFERENCE SCHEMATICS 41 FN4643.3 April 19, 2007 HMP8117 Absolute Maximum Ratings Thermal Information Digital Supply Voltage (VCC to GND) . . . . . . . . . . . . . . . . . . . . 7.0V Analog Supply Voltage (VAA to GND) . . . . . . . . . . . . . . . . . . . 7.0V Digital Input Voltages . . . . . . . . . . . . . . . GND - 0.5V to VCC + 0.5V ESD Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Class 1 Thermal Resistance (Typical, See Note 41) θJA (°C/W) MQFP Package . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Maximum Power Dissipation HMP8117CN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1.78W Maximum Storage Temperature Range . . . . . . . . . .-65°C to +150°C Maximum Junction Temperatures . . . . . . . . . . . . . . . . . . . . . +150°C Maximum Lead Temperature (Soldering 10s) . . . . . . . . . . . +300°C Operating Temperature Range HMP8117CN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0°C to +70°C CAUTION: Stresses above those listed in “Absolute Maximum Ratings” may cause permanent damage to the device. This is a stress only rating and operation of the device at these or any other conditions above those indicated in the operational sections of this specification is not implied. NOTE: 41. θJA is measured with the component mounted on an evaluation PC board in free air. Dissipation rating assumes device is mounted with all leads soldered to printed circuit board. Electrical Specifications VCC = VAA = 5.0V, TA = +25°C PARAMETER SYMBOL TEST CONDITION MIN TYP MAX UNITS 4.75 5 5.25 V 279 305 mA POWER SUPPLY CHARACTERISTICS Power Supply Voltage Range VCC , VAA Total Power Supply Current ITOT Digital Power Supply Current ICC Analog Power Supply Current Total Power Dissipation (Note 42) CLK2 = 29.5MHz, VCC = VAA = 5.25V Outputs Not Loaded 132 mA IAA 147 mA PTOT 1.46 1.60 W DC CHARACTERISTICS: DIGITAL I/O (EXCEPT CLK2 and I2C INTERFACE) Input Logic High Voltage VIH VCC = Max Input Logic Low Voltage VIL VCC = Min Output Logic High Voltage VOH IOH = -4mA, VCC = Max Output Logic Low Voltage VOL IOL = 4mA, VCC = Min Input Leakage Current IIH, IIL Input/Output Capacitance CIN , COUT Three-State Output Current Leakage VCC = Max, Input = 0V or 5V 2.0 0.8 2.4 V V - 0.4 V -10 ± 10 μA - pF 10 μA 8 f = 1MHz (Note 42) All Measurements Referenced to Ground, TA = +25°C IOZ V -10 DC CHARACTERISTICS: CLK2 DIGITAL INPUT Input Logic High Voltage VIH VCC = Max Input Logic Low Voltage VIL VCC = Min Input Leakage Current IIH VCC = Max Input = 0V or VCC IIL Input Capacitance CIN CLK2 = 1MHz (Note 42) All Measurements Referenced to Ground, TA = +25°C Input Logic High Voltage VIH VCC = Max Input Logic Low Voltage VIL VCC = Min 0.7xVCC V -10 0.3xVCC V 10 μA μA -450 - 8 - pF DC CHARACTERISTICS: I2C INTERFACE Output Logic High Voltage VOH IOH = -1mA, VCC = Max Output Logic Low Voltage VOL IOL = 3mA, VCC = Min 42 0.7xVCC V 0.3xVCC 3.0 0 V V 0.4 V FN4643.3 April 19, 2007 HMP8117 Electrical Specifications VCC = VAA = 5.0V, TA = +25°C (Continued) PARAMETER SYMBOL Input Leakage Current IIH, IIL Input/Output Capacitance CIN, COUT TEST CONDITION MIN TYP VCC = Max, Input = 0V or 5V MAX UNITS 10 μA 8 SCL = 400kHz, (Note 42) All Measurements Referenced to GND, TA = +25°C pF AC CHARACTERISTICS: DIGITAL I/O (EXCEPT I2C INTERFACE) CLK2 Frequency CLK2 Waveform Symmetry (Note 42) 24.54 29.5 MHz 40 60 % CLK2 Pulse Width High tPWH 13 ns CLK2 Pulse Width Low tPWL 13 ns 10 ns ns Data and Control Setup Time tSU Data and Control Hold Time tHD 0 tDVLD 0 23 ns 1 12 ns 400 kHz CLK2 to Output Delay Data and Control Rise/Fall Time tr , tf (Note 43) (Note 42) AC CHARACTERISTICS: I2C INTERFACE SCL Clock Frequency fSCL 0 SCL Pulse Width Low tLOW 1.3 μs SCL Pulse Width High tHIGH 0.6 μs Data Hold Time tHD:DATA 0 ns Data Setup Time tSU:DATA 100 ns SDA, SCL Rise Time tR SDA, SCL Fall Time tF (Note 42) 300 ns 300 ns ANALOG INPUT PERFORMANCE Composite Video Input Amplitude (Sync Tip to White Level) Input Termination of 75Ω and 1.0μF AC-Coupled 0.5 1.0 2.0 VP-P Luminance (Y) Video Input Amplitude (Sync Tip to White Level) Input Termination of 75Ω and 1.0μF AC-Coupled 0.5 1.0 2.0 VP-P Chrominance (C) Video Input Amplitude (Burst Amplitude) Input Termination of 75Ω and 1.0μF AC-Coupled, (Note 42) 0.143 0.286 0.6 VP-P Video Input Impedance RAIN Video Input Bandwidth BW ADC Input Range (Note 42) 1VP-P Sine Wave Input to -3dBc Reduction, (Note 42) 200 kΩ 5 MHz AIN FULL SCALE 1 VP-P AIN OFFSET 1.5 V ±2 LSB ± 0.35 LSB 2 % 1 Deg. 2 Deg. 2 % NTC-7 Composite (Note 42) 2 % Pedestal Input (Note 42) 50 dB ADC Integral Nonlinearity INL ADC Differential Nonlinearity DNL Best Fit Linearity VIDEO PERFORMANCE Differential Gain DG Differential Phase DP Hue Accuracy Modulated Ramp (Note 42) 75% Color Bars (Note 42) Color Saturation Accuracy Luminance Nonlinearity SNR SNRL WEIGHTED 43 FN4643.3 April 19, 2007 HMP8117 Electrical Specifications VCC = VAA = 5.0V, TA = +25°C (Continued) PARAMETER SYMBOL TEST CONDITION MIN TYP 2 3 MAX UNITS GENLOCK PERFORMANCE Horizontal Locking Time tLOCK Long-Term horizontal Sync Lock Range Time from Initial Lock Acquisition to an Error of 1 Pixel. (Note 42) Range over specified pixel jitter is maintained. Assumes line time changes by amount indicated slowly between over one field. (Note 42) Number of Missing Horizontal Syncs Before Lost Lock Declared HSYNC LOST Number of Missing Vertical Syncs Before Lost Lock Declared VSYNC LOST Programmable via register 04H (Note 42) Long-Term Color Subcarrier Lock Range Range over color subcarrier locking time and accuracy specifications are maintained. Subcarrier frequency changes by amount indicated slowly over 24 hours. (Note 42) Vertical Sample Alignment (Notes 42, 44) ±- Fields 5 % 1 or 12 1 or 12 1 or 12 HSYNCs 1 or 3 1 or 3 1 or 3 VSYNCs ±200 ±400 Hz 1/8 Pixel 10 ns NOTES: 42. Guaranteed by design or characterization. 43. Test performed with CL = 40pF, IOL = 4mA, IOH = -4mA. Input reference level is 1.5V for all inputs. VIH = 3.0V, VIL = 0V. 44. Since the HMP8117 does not generate the sample clock, any clock jitter present on the CLK2 input will directly translate to pixel jitter on the output data. The Vertical Sample Alignment parameter specifies the spatial pixel alignment from one scan line to the next using a stable CLK2 source. 44 FN4643.3 April 19, 2007 HMP8117 Metric Plastic Quad Flatpack Packages (MQFP) D Q80.14x20 (JEDEC MS-022GB-1 ISSUE B) 80 LEAD METRIC PLASTIC QUAD FLATPACK PACKAGE D1 -D- INCHES SYMBOL -A- -B- E E1 e PIN 1 SEATING A PLANE -H- 0.076 0.003 12o-16o -C- 0.40 0.016 MIN 0.20 M 0.008 C A-B S 0o MIN D S b L b1 MIN MAX NOTES A - A1 0.010 0.134 - 3.40 - - 0.25 - - A2 0.098 0.114 2.50 2.90 - b 0.012 0.018 0.30 0.45 6 b1 0.012 0.016 0.30 0.40 - D 0.908 0.918 23.08 23.32 3 D1 0.782 0.792 19.88 20.12 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 - N 80 80 7 e 0.032 BSC 0.80 BSC - ND 24 24 - NE 16 16 Rev. 1 4/99 NOTES: 1. Controlling dimension: MILLIMETER. Converted inch dimensions are not necessarily exact. 5. Dimensions D1 and E1 do not include mold protrusion. Allowable protrusion is 0.25mm (0.010 inch) per side. BASE METAL WITH PLATING 3. Dimensions D and E to be determined at seating plane -C- . 4. Dimensions D1 and E1 to be determined at datum plane -H- . 0.13/0.17 0.005/0.007 12o-16o MAX 2. All dimensions and tolerances per ANSI Y14.5M-1982. A2 A1 0o-7o MILLIMETERS MIN 0.13/0.23 0.005/0.009 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. All Intersil U.S. products are manufactured, assembled and tested utilizing ISO9000 quality systems. Intersil Corporation’s quality certifications can be viewed at www.intersil.com/design/quality Intersil products are sold by description only. Intersil Corporation reserves the right to make changes in circuit design, software and/or specifications at any time without notice. Accordingly, the reader is cautioned to verify that data sheets are current before placing orders. Information furnished 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 patents 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 subsidiaries. For information regarding Intersil Corporation and its products, see www.intersil.com 45