a 96-Bit, 220 MHz True-Color Video RAM-DAC ADV7160/ADV7162 FEATURES 96-Bit Pixel Port for 1600 × 1280 × 24 Screen Resolution 220 MHz, 24-Bit (30-Bit Gamma Corrected) True-Color Triple 10-Bit “Gamma Correcting” D/A Converters 2% (max) DAC to DAC Color Matching Triple 256 × 10 (256 x 30) Color Palette RAM On-Board User Definable Cursor (64 × 64 × 2) Three Color Overlay Cursor Palette RAM Fully Programmable On-Board PLL RS-343A/RS-170 Compatible RGB Analog Outputs Tri-Level SYNC Functionality TTL Compatible Digital Inputs Standard MPU I/O Interface Programmable Pixel Port: 24-Bit, 16-Bit, 15-Bit & 8-Bit (Pseudo) Pixel Data Serializer: Multiplexed Pixel Input Ports; 2:1, 4:1, 8:1 +5 V CMOS Monolithic Construction 160-Lead Plastic Quad Flatpack (QFP): ADV7162 160-Lead “Thermally Enhanced” QFP (PQUAD): ADV7160 MODES OF OPERATION 1600 × 1200 × 30/24-Bit Resolution @ 85 Hz Screen Refresh 1600 × 1200 × 16/15-Bit Resolution @ 85 Hz Screen Refresh 1600 × 1200 × 8-Bit Resolution @ 85 Hz Screen Refresh APPLICATIONS Windows Accelerators High Resolution, True Color Graphics Professional Color Prepress Imaging Digital TV (HDTV, Digital Video) SPEED GRADES @ 220 MHz @ 170 MHz @ 140 MHz GENERAL DESCRIPTION The ADV7160/ADV7162® is a 96-bit pixel port Video RAMDAC with color enhanced triple 10-bit DACs. The device also includes a PLL and 64 × 64 hardware cursor. The ADV7160/ ADV7162 is specifically designed for use in the graphics subsystem of high performance, color graphics workstations and windows accelerators. (Continued on page 15) ADV is a registered trademark of Analog Devices, Inc. FUNCTIONAL BLOCK DIAGRAM VAA SYNCOUT TRISYNC BLANK AND SYNC LOGIC P I X E L SYNC BLANK 10 8 10 BYPASS COLOR MODE MATRIX 10 24 A 24 PIXEL DATA (P7-P0) B I N P U T 8 24 D PALETTE SELECTS (PS0, PS1) ODD/EVEN 8 3 x 256 COLOR PALETTE 8 8 8 8 24 C COLOR MODE MATRIX M U L T I P L E X E R PIXEL MASK 8 10 RED 256 x 10 8 10 BLUE 256 x 10 3 COLOR OVERLAY PALETTE 10 RED 3 x 10 2 PS FUNCTION DECODE LOGIC 10 GREEN 3 x 10 2 10 BLUE 3 x 10 64 x 64 CURSOR GENERATOR 2 CLOCK DIVIDE & SYNCHRONIZATION CIRCUITRY ÷32, ÷16, ÷8, ÷4 SCKIN CONTROL REGISTERS ADDRESS REGISTER (A10-A0) MODE REGISTER (MR1) SELECTOR 10 CLOCK CLOCK IOR 10 GREEN DAC IOG 10 ECL TO CMOS 10 BLUE DAC IOB ADV7160/ ADV7162 10 GREEN 3 x 10 CLOCK CONTROL SCKOUT RED DAC 10 RED 3 x 10 10 BLUE 3 x 10 PRGCKOUT S E L E C T O R 2 COLOR CURSOR PALETTE LOADIN LOADOUT 10 10 GREEN 256 x 10 8 CURSOR REGISTERS PIXEL MASK REGISTER TEST REGISTERS REVISION REGISTER ID REGISTER PLL REGISTERS STATUS REGISTER COMMAND REGISTERS (CR1-CR5) VOLTAGE REFERENCE CIRCUIT DATA TO PALETTES VREF RSET COMP 30 RED REGISTER GREEN REGISTER BLUE REGISTER 10 JTAG TEST ACCESS PORT MPU PORT PLL TDO 10 (8+2) PLLREF C1 R/W CE C0 D9–D0 TMS TCK TDI GND REV. 0 Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices 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 Analog Devices. © Analog Devices, Inc., 1995 One Technology Way, P.O. Box 9106, Norwood. MA 02062-9106, U.S.A. Tel: 617/329-4700 Fax: 617/326-8703 ADV7160/ADV7162–SPECIFICATIONS Parameter STATIC PERFORMANCE Resolution (Each DAC) Accuracy (Each DAC) Integral Nonlinearity Differential Nonlinearity Gray Scale Error Coding DIGITAL INPUTS Input High Voltage, VINH Input Low Voltage, VINL Input Current, IIN Input Capacitance, CIN CLOCK INPUTS (CLOCK, CLOCK) Input High Voltage, VINH Input Low Voltage, VINL Input Current, IIN Input Current, IIN (JTAG Inputs) Input Capacitance, CIN DIGITAL OUTPUTS Output High Voltage, VOH Output Low Voltage, VOL Floating-State Leakage Current Floating-State Output Capacitance ANALOG OUTPUTS Gray Scale Current Range Output Current White Level Relative to Blank White Level Relative to Black Black Level Relative to Blank Blank Level Blank Level Sync Level Tri-Sync Level Relative to Blank LSB Size DAC to DAC Matching Output Compliance, VOC Output Impedance, ROUT Output Capacitance, COUT VOLTAGE REFERENCE Voltage Reference Range, VREF Input Current, IVREF POWER REQUIREMENTS VAA IAA3 Min Typ Max Units 10 Bits ±1 ±1 ±5 LSB LSB % Gray Scale Binary Test Conditions/Comments (DAC Gain Setting = 3996) 2 0.8 ± 10 V V µA pF VAA – 1.6 ± 10 ± 50 V V µA µA pF 0.4 20 V V µA pF 22 mA 20.40 18.50 1.90 50 8.96 50 8.96 mA mA mA µA mA µA mA µA % V kΩ pF 10 VAA – 1.0 10 2.4 20 Guaranteed Monotonic VIN = 0.4 V or 2.4 V VIN = 0.4 V or 2.4 V VIN = 0.4 V or 2.4 V ISOURCE = 400 µA ISINK = 3.2 mA (DAC Gain Setting = 3996) 15 17.69 16.74 0.95 0 6.29 0 6.29 19.05 17.62 1.44 5 7.62 5 7.62 17.22 1 0 3 +1.4 30 30 1.14 1.235 5 1.26 5 V µA 0.1 V mA mA mA mA mA mA %/% –30 50 –23 dB pV secs dB 475 440 410 450 400 360 IAA3 Power Supply Rejection Ratio DYNAMIC PERFORMANCE Clock and Data Feedthrough4, 5 Glitch Impulse DAC to DAC Crosstalk6 (VAA1 = +5 V; VREF = +1.235 V; RSET = 280 Ω. IOR, IOG, IOB (RL = 37.5 Ω, CL = 10 pF). All specifications TMIN to TMAX2 unless otherwise noted.) Sync Disabled Sync Enabled IOUT = 0 mA VREF = 1.235 V for Specified Performance For 220 MHz Operation (ADV7160) For 170 MHz Operation (ADV7160) For 140 MHz Operation (ADV7160) For 220 MHz Operation (ADV7162) For 170 MHz Operation (ADV7162) For 140 MHz Operation (ADV7162) COMP = 0.1 µF NOTES 1 ± 5% for all versions. 2 Temperature range (T MIN to TMAX): 0°C to +70°C. 3 Pixel Port is continuously clocked with data corresponding to a linear ramp. T J = 100oC. 4 Clock and data feedthrough is a function of the amount of overshoot and undershoot on the digital inputs. Glitch impulse includes clock and data feedthrough. 5 TTL input values are 0 V to 3 V, with input rise/fall times ≤3 ns, measured the 10% and 90% points. Timing reference points at 50% for inputs and outputs. 6 DAC to DAC Crosstalk is measured by holding one DAC high while the other two are making low to high and high to low transitions. Specifications subject to change without notice. –2– REV. 0 ADV7160/ADV7162 2 1 (VAA = +5 V; VREF = +1.2353 V; RSET = 280 Ω. IOR, IOG, IOB (RL = 37.5 Ω, CL = 10 pF). All TIMING CHARACTERISTICS specifications TMIN to TMAX unless otherwise noted.) CLOCK CONTROL AND PIXEL PORT 4 Parameter 220 MHz Version 170 MHz Version 140 MHz Version Units Conditions/Comments fCLOCK t1 t2 t3 220 4.5 2.0 2.0 170 5.88 2.5 2.5 140 7.14 2.86 2.86 MHz max ns min ns min ns min Pixel CLOCK Rate Pixel CLOCK Cycle Time Pixel CLOCK High Time Pixel CLOCK Low Time t4 10 10 10 ns max Pixel CLOCK to LOADOUT Delay 110 55 27.5 85 42.5 21.25 70 35 17.5 MHz max MHz max MHz max 9.1 18.18 36.36 11.77 23.53 47.1 14.29 28.58 57.16 ns min ns min ns min 4 8 15 5 9 18 6 12 23 ns min ns min ns min 4 8 15 5 9 18 6 12 23 ns min ns min ns min t8 t9 0 5 0 5 0 5 ns min ns min Pixel Data Setup Time Pixel Data Hold Time t10 τ-t115 0 τ-5 0 τ-5 0 τ-5 ns min ns max LOADOUT to LOADIN Delay LOADOUT to LOADIN Delay tPD6 2:1 Multiplexing 4:1 Multiplexing 8:1 Multiplexing t12 t13 t14 t15 9 11 15 10 5 5 0 9 11 15 10 5 5 0 9 11 15 10 5 5 0 CLOCKs CLOCKs CLOCKs ns max ns max ns min ns min Parameter 220 MHz Version 170 MHz Version 140 MHz Version Units Conditions/Comments t16 t17 t18 tSK 25 1 25 2 0 25 1 25 2 0 25 1 25 2 0 ns typ ns typ ns typ ns max ns typ Analog Output Delay Analog Output Rise/Fall Time Analog Output Transition Time RGB Analog Output Skew fLOADIN 2:1 Multiplexing 4:1 Multiplexing 8:1 Multiplexing t5 2:1 Multiplexing 4:1 Multiplexing 8:1 Multiplexing t6 2:1 Multiplexing 4:1 Multiplexing 8:1 Multiplexing t7 2:1 Multiplexing 4:1 Multiplexing 8:1 Multiplexing LOADIN Clocking Rate LOADIN Cycle Time LOADIN High Time LOADIN Low Time Pipeline Delay (1 × CLOCK = t1) Pixel CLOCK to PRGCKOUT Delay SCKIN to SCKOUT Delay BLANK to SCKIN Setup Time BLANK to SCKIN Hold Time ANALOG OUTPUTS7 REV. 0 –3– ADV7160/ADV7162 MPU PORT 8,9 Parameter 220 MHz Version 170 MHz Version 140 MHz Version Units Conditions/Comments t19 t20 t21 t22 t238 t249 t259 t269 t27 t28 0 10 45 25 5 45 20 5 20 5 0 10 45 25 5 45 20 5 20 5 0 10 45 25 5 45 20 5 20 5 ns min ns min ns min ns min ns min ns max ns max ns min ns min ns min R/W, C0, C1 to CE Setup Time R/W, C0, C1 to CE Hold Time CE Low Time CE High Time CE Asserted to Data-Bus Driven CE Asserted to Data Valid CE Disabled to Data-Bus Three-Stated CE Disabled to Data Invalid Write Data (D0–D9) Setup Time Write Data (D0–D9) Hold Time NOTES General Notes 1 TTL input values are 0 to 3 volts, with input rise/fall times ≤ 3 ns, measured between the 10% and 90% points. ECL inputs (CLOCK, CLOCK) are VAA–0.8 V to VAA–1.8 V, with input rise/fall times ≤ 2 ns, measured between the 10% and 90% points. Timing reference points at 50% for inputs and outputs. Analog output load ≤ 10 pF. Data-Bus (D0–D9) loaded as shown in Figure 1. Digital output load for LOADOUT, PRGCKOUT & SCKOUT ≤ 30 pF. 2 ± 5% for all versions 3 Temperature range (T MIN to TMAX); 0°C to +70°C. Notes on PIXEL PORT 4 Pixel Port consists of the following inputs: Pixel Inputs: RED [A, B, C, D] GREEN [A, B, C, D] BLUE [A, B, C, D] Palette Selects: PS0 [A, B, C, D]; PS1[A, B, C, D] Pixel Controls: SYNC, BLANK, TRISYNC, ODD/EVEN Clock Inputs: CLOCK, CLOCK, LOADIN, SCKIN Clock Outputs: LOADOUT, PRGCKOUT, SCKOUT 5 τ is the LOADOUT Cycle Time and is a function of the Pixel CLOCK Rate and the Multiplexing Mode: 2:1 multiplexing; τ = CLOCK × 2 = 2 × t1 ns ns 4:1 multiplexing; τ = CLOCK × 4 = 4 × t1 8:1 multiplexing; τ = CLOCK × 8 = 8 × t1 ns 6 These fixed values for Pipeline Delay are valid under conditions where t 10 and τ-t11 are met. If either t 10 or τ-t11 are not met, the part will operate but the Pipeline Delay is increased. Notes on ANALOG OUTPUTS 7 Output delay measured from the 50% point of the rising edge of CLOCK to the 50% point of full-scale transition. Output rise/fall time measured between the 10% and 90% points of full-scale transition. Transition time measured from the 50% point of full scale transition to the output remaining within 2% of the final output value. (Transition time does not include clock and data feedthrough). Notes on MPU PORT 8 t23 and t24 are measured with the load circuit of Figure 1 and defined as the time required for an output to cross 0.4 V or 2.4 V. 9 t25 and t26 are derived from the measured time taken by the data outputs to change by 0.5 V when loaded with the circuit of Figure 1. The measured numbers are then extrapolated back to remove the effects of charging the 100 pF capacitor. This means that the times t 25 and t26, quoted in the Timing Characteristics are the true values for the device and as such are independent of external loading capacitances. Specifications subject to change without notice. ISINK TO OUTPUT PIN +2.1V 100pF ISOURCE Figure 1. Load Circuit for Databus Access and Relinquish Times –4– REV. 0 ADV7160/ADV7162 2 1 (VAA = +5 V; VREF = +1.235 V; 3RSET = 280 Ω. IOR, IOG, IOB (RL = 37.5 Ω, CL =10 pF). TIMING CHARACTERISTICS (Cont.) All specifications TMIN to TMAX unless otherwise noted.) JTAG PORT Parameter All Versions Units Conditions/Comments 250 ps rms 1σ 900 40 2.0 0.8 25 1.67 40 60 kHz min MHz max V max V min ns min µs max % min % max 20 15 15 15 15 15 15 0 20 5 15 MHz max ns min ns min ns max ns max ns max ns max ns min ns min ns min ns max 4 PLL PERFORMANCE Jitter PLL REFERENCE INPUT PLLREF Frequency VIH VIL PLLREF Period PLLREF Duty Cycle JTAG PERFORMANCE TCK Frequency, t29 TCK High Time, t30 TCK Low Time, t31 TDI, TMS Setup Time, t32 TDI, TMS Hold Time, t33 Digital Input to TCK Setup Time, t34 Digital Input to TCK Hold Time, t35 TCLK to TDO Drive, t36 TCLK to TDO Valid, t37 TCLK to TDO Three-State, t38 NOTES 1 TTL input values are 0 to 3 volts, with input rise/fall times ≤ 3 ns, measured between the 10% and 90% points. Timing reference points at 50% for inputs and outputs. 2 ± 5% for all versions. 3 Temperature range (T MIN to TMAX); 0°C to +70°C. 4 Jitter is measured by triggering on the output clock, delayed by 15 µs and then measuring the time period from the trigger edge to the next edge of the output clock after the delay. This measurement is repeated multiple times and the RMS value is determined. Specifications subject to change without notice. t29 t32 t31 t30 t33 TCK t34 t35 TMS, TDI DIGITAL INPUT t37 t36 TDO TDO t38 Figure 2. JTAG Timing REV. 0 –5– ADV7160/ADV7162 Timing Waveforms t2 t1 t3 CLOCK CLOCK t4 LOADOUT (2:1 MULTIPLEXING) LOADOUT (4:1 MULTIPLEXING) LOADOUT (8:1 MULTIPLEXING) Figure 3. LOADOUT vs. Pixel Clock Input (CLOCK, CLOCK) t5 t6 t7 LOADIN t9 t8 PIXEL INPUT DATA VALID DATA VALID DATA VALID DATA Figure 4. LOADIN vs. Pixel Input Data –6– REV. 0 ADV7160/ADV7162 CLOCK t10 LOADOUT LOADIN PIXEL INPUT DATA AN ... HN AN+1 ... HN+1 AN+2 ... HN+2 DIG ITA OUT L INPU T PUT PIPE TO ANA LINE LOG ANALOG OUTPUT DATA (IOR, IOG, IOB, SYNCOUT) AN–1 ... HN–1 AN ... HN AN+1 ... HN+1 AN+2 ... HN+2 tPD Figure 5. Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (8:1 Multiplex Mode) CLOCK τ τ-t11 LOADOUT LOADIN PIXEL INPUT DATA AN ... HN AN+1 ... HN+1 AN+2 ... HN+2 DIG IT OU AL IN TP UT PUT PIP TO A EL INE NAL O G ANALOG OUTPUT DATA (IOR, IOG, IOB, SYNCOUT) AN–1 ... HN–1 AN ... HN AN+1 ... HN+1 AN+2 ... HN+2 tPD Figure 6. Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (8:1 Multiplex Mode) REV. 0 –7– ADV7160/ADV7162 CLOCK t10 LOADOUT LOADIN PIXEL INPUT DATA AN ... DN AN+1 ... DN+1 AN+2 ... DN+2 DIGIT AL OUTP INPUT TO UT PIP AN ELINE ALOG ANALOG OUTPUT DATA (IOR, IOG, IOB, SYNCOUT) AN–1 ... DN–1 AN ... DN AN+1 ... DN+1 AN+2 ... DN+2 tPD Figure 7. Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (4:1 Multiplex Mode) CLOCK τ τ-t11 LOADOUT LOADIN PIXEL INPUT DATA AN ... DN AN+1 ... DN+1 AN+2 ... DN+2 DIGIT A OUT L INPUT PUT T PIPE O ANAL LINE OG ANALOG OUTPUT DATA (IOR, IOG, IOB, SYNCOUT) AN–1 ... DN–1 AN ... DN AN+1 ... DN+1 AN+2 ... DN+2 tPD Figure 8. Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (4:1 Multiplex Mode) –8– REV. 0 ADV7160/ADV7162 CLOCK t10 LOADOUT LOADIN PIXEL INPUT DATA AN ... BN AN+1 ... BN+1 AN+2 ... BN+2 DIGITAL INPUT TO ANALOG OUTPUT PIPELINE ANALOG OUTPUT DATA (IOR, IOG, IOB, SYNCOUT) AN–1 BN–1 AN BN AN+1 BN+1 AN+2 BN+2 tPD Figure 9. Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (2:1 Multiplex Mode) CLOCK τ τ-t10 LOADOUT LOADIN PIXEL INPUT DATA AN ... BN AN+1 ... BN+1 DIGITAL INPUT TO ANALOG OUTPUT DATA (IOR, IOG, IOB, SYNCOUT) AN+2 ... BN+2 ANALOG OUTPUT PIPELIN E AN–1 BN–1 AN BN AN+1 BN+1 AN+2 BN+2 tPD Figure 10. Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (2:1 Multiplex Mode) REV. 0 –9– ADV7160/ADV7162 CLOCK PRGCKOUT (CLOCK/4) PRGCKOUT (CLOCK/8) PRGCKOUT (CLOCK/16) PRGCKOUT (CLOCK/32) t12 Figure 11. Pixel Clock Input vs. Programmable Clock Output (PRGCKOUT) t14 t13 SCKIN t15 BLANKING PERIOD BLANK SCKOUT END OF SCAN LINE (N) START OF SCAN LINE (N+1) Figure 12. Video Data Shift Clock Input (SCKIN) & BLANK vs. Video Data Shift Clock Output (SCKOUT) CLOCK t18 t16 WHITE LEVEL 90% ANALOG OUTPUTS IOR IOG IOB SYNCOUT 50% FULL SCALE TRANSITION 10% t17 BLACK LEVEL NOTE: THIS DIAGRAM IS NOT TO SCALE. FOR THE PURPOSES OF CLARITY, THE ANALOG OUTPUT WAVEFORM IS MAGNIFIED IN TIME AND AMPLLITUDE W.R.T THE CLOCK WAVEFORM. SYNCOUT IS A DIGITAL VIDEO OUTPUT SIGNAL. t16 IS THE ONLY RELEVANT TIMING SPECIFICATION FOR SYNCOUT. Figure 13. Analog Output Response vs. CLOCK –10– REV. 0 ADV7160/ADV7162 t19 R/W, C0, C1 t20 VALID CONTROL DATA t21 CE t24 t22 t25 t23 D0–D9 (READ MODE) R/W = 1 t26 D0–D9 (WRITE MODE) R/W = 0 t27 t28 Figure 14. Microprocessor Port (MPU) Interface Timing ABSOLUTE MAXIMUM RATINGS 1 160-Lead QFP Configuration 120 ROW C 121 ROW D NOTES 1 Stresses above those listed under “Absolute Maximum Ratings” may cause permanent damage to the device. This is a stress rating only and functional operation of the device at these or any other conditions above those listed in the operational sections of this specification is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. 2 Analog Output Short Circuit to any Power Supply or Common can be of an indefinite duration. ORDERING INFORMATION 1, 2, 3 220 MHz ADV7160KS2203 ADV7162KS2204 Dot Clock Speed 170 MHz ADV7160KS1703 ADV7162KS1704 81 80 ADV7160/ADV7162 QFP ROW B VAA to GND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 V Voltage on Any Digital Pin . . . . . GND – 0.5 V to VAA + 0.5 V Ambient Operating Temperature (TA) . . . . . . . . 0°C to +70°C Storage Temperature (TS) . . . . . . . . . . . . . . . –65°C to +150°C Junction Temperature (TJ) . . . . . . . . . . . . . . . . . . . . . +150°C Lead Temperature (Soldering, 10 secs) . . . . . . . . . . . . +260°C Vapor Phase Soldering (1 minute) . . . . . . . . . . . . . . . . +220°C Analog Outputs to GND2 . . . . . . . . . . . . GND – 0.5 V to VAA TOP VIEW (NOT TO SCALE) PIN NO. 1 IDENTIFIER 160 ADV7160KS1403 ADV7162KS1404 41 ROW A 140 MHz 40 1 NOTES 1 All devices are specified for 0°C to +70°C operation. 2 Contact Sales Office for latest information on package design. 3 ADV7160 is packaged in a 160-pin plastic power quad flatpack, QFP with heatsink embedded. 4 ADV7162 is packaged in a standard 160-pin plastic quad flatpack, QFP. CAUTION ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily accumulate on the human body and test equipment and can discharge without detection. Although the ADV7160/ADV7162 features proprietary ESD protection circuitry, permanent damage may occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD precautions are recommended to avoid performance degradation or loss of functionality. REV. 0 –11– WARNING! ESD SENSITIVE DEVICE ADV7160/ADV7162 ADV7160/ADV7162 PIN ASSIGNMENTS Pin No. Mnemonic Pin No. Mnemonic Pin No. Mnemonic Pin No. Mnemonic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 G2A G2B G2C G2D G3A G3B G3C G3D G4A G4B G4C G4D G5A G5B G5C G5D G6A G6B G6C G6D G7A VAA VAA GND GND VAA GND PLLREF G7B G7C G7D PS0A PS0B PS0C PS0D PS1A PS1B PS1C PS1D CLOCK 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 CLOCK SCKIN SCKOUT VAA PRGCKOUT GND LOADOUT LOADIN B0A B0B B0C B0D B1A B1B B1C B1D B2A B2B B2C B2D B3A B3B B3C B3D B4A B4B B4C B4D B5A B5B B5C B5D B6A B6B B6C B6D B7A B7B B7C B7D 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 C1 C0 R/W CE TCK TMS GND VAA TDO TDI SYNCOUT TRISYNC ODD/EVEN SYNC BLANK VREF IOB COMP RSET VAA VAA GND IOG IOR R0A R0B R0C R0D R1A R1B 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 R1C R1D R2A R2B R2C R2D R3A R3B R3C R3D R4A VAA VAA GND GND R4B R4C R4D R5A R5B R5C R5D R6A R6B R6C R6D R7A R7B R7C GND VAA R7D G0A G0B G0C G0D G1A G1B G1C G1D –12– REV. 0 ADV7160/ADV7162 PIN FUNCTION DESCRIPTION Mnemonic Function RED (R0A . . . R0B – R7A . . . R7D), GREEN (G0A . . . G0D – G7A . . . G7D), BLUE (B0A . . . B0D – B7A . . . B7D): Pixel Port (TTL Compatible Inputs): 96 pixel select inputs, with 8 bits each for Red, Green and Blue. Each bit is multiplexed [A-D] 4:1 or 2:1. It can be configured for 24-Bit True-Color Data, 8-Bit Pseudo-Color Data, 16-Bit True-Color and 15-Bit True-Color Data formats. In 8-Bit Pseudo-Color Mode, there is a special case whereby 8:1 multiplexing is also available. It will be explained in more detail later. Pixel Data is latched into the device on the rising edge of LOADIN. PS0A . . . PS0D, PS1A . . . PS1D Palette Priority Selects (TTL Compatible Inputs): The eight PS inputs provide two Bits after input multiplexing. These pixel port select inputs can be configured for three separate functions. In Overlay Mode, these inputs provide a three color overlay function. With any value other than “00” on the overlay inputs, the color displayed comes from the overlay palette instead of the main pixel inputs. For the ADV7160, in Bypass Mode, PS1 specifies for each pixel whether it should pass through the Color Matrix and Color Palette or bypass the Matrix and Palette. PS0 acts as an overlay input. (This mode is not available for the ADV7162.) Palette Select Mode is used to multiplex the RGB outputs of a number of devices. When the palette mode inputs match the PS bits in the mode register, the part operates as normal. When there is a mismatch, the RGB outputs are switched to zero, allowing the RGB outputs of another device to drive the monitor. LOADIN Pixel Data Load Input (TTL Compatible Input): This input latches the multiplexed pixel data, including PS0-PS1, BLANK, TRISYNC, SYNC and ODD/EVEN into the device. LOADOUT Pixel Data Load Output (TTL Compatible Output): This output control signal runs at a divided down frequency of the pixel clock. Its frequency is a function of the multiplex rate. It can be used to directly or indirectly drive LOADIN. fLOADOUT = fCLOCK/M where (M = 2 for 2:1 Multiplex Mode) (M = 4 for 4:1 Multiplex Mode) (M = 8 for 8:1 Multiplex Mode) PRGCKOUT Programmable Clock Output (TTL Compatible Output): This output control signal runs at a divided down frequency of the pixel Clock. Its frequency is user programmable and is determined by bits CR30 and CR31 of Command Register 3. fPRGCKOUT = fCLOCK/N where N = 4, 8, 16 & 32 SCKIN Video Shift Clock Input (TTL Compatible Input): The signal on this input is internally gated synchronously with the BLANK signal. The resultant output, SCKOUT, is a video clocking signal that is stopped during video blanking periods. It is normally driven by a divided down version of the CLOCK frequency. SCKOUT Video Shift Clock Output (TTL Compatible Output): This output is a synchronously gated version of SCKIN and BLANK. SCKOUT is a video clocking signal that is stopped during video blanking periods. CLOCK, CLOCK Clock Inputs (ECL Compatible Inputs): These differential clock inputs are designed to be driven by ECL logic levels configured for single supply (+5 V) operation. The clock rate is normally the pixel clock rate of the system. PLLREF PLL Clock Input (TTL Compatible Input): This clock input is designed to be driven by TTL logic levels. The PLL is then configured to output a specific frequency depending on the PLL Registers. See PLL section for more detail. BLANK Composite Blank (TTL Compatible Input): This video control signal drives the analog outputs to the blanking level. SYNC Composite-Sync Input (TTL Compatible Input): This video control signal drives any of the analog outputs to the SYNC level. It is only asserted during the blanking period. CR22 in Command Register 2 must be set if SYNC is to be decoded onto the IOG analog output, CR41 in Command Register 4 must be set if SYNC is to be decoded onto the IOR analog output, CR42 in Command Register 4 must be set if SYNC is to be decoded onto the IOB analog output, otherwise the SYNC input is ignored. REV. 0 –13– ADV7160/ADV7162 Mnemonic Function SYNCOUT Composite-Sync Output (TTL Compatible Output). This video output is a delayed version of SYNC. The delay corresponds to the number of pipeline stages of the device. TRISYNC Composite-Sync HDTV Control (TTL Compatible Output). This video input is enabled using Bit CR17 in Command Register 1. When TRISYNC is low, any DAC output which has Sync enabled, goes to the tri-sync level. As with the SYNC input, it should only be activated while BLANK is low. D9–D0 Data Bus (TTL Compatible Input/Output Bus). Data, including color palette values and device control information is written to and read from the device over this 10-bit, bidirectional databus. 10-bit data or 8-bit data can be used. The databus can be configured for either 10-bit parallel data or byte data (8+2) as well as standard 8-bit data. Any unused bits of the data bus should be terminated through a resistor to either the digital power plane (VCC) or GND. ODD/EVEN Odd/Even Control (TTL Compatible Input). This input indicates which field of the frame is being displayed. It is required to ensure proper operation of the ADV7160/ADV7162 cursor when interlaced display mode is selected. It is ignored when noninterlaced display mode is selected. This input should change only during the vertical blank period. It is assumed that an odd field will always follow an even field and vice versa. CE Chip Enable (TTL Compatible Input). This input must be at Logic “0” when writing to or reading from the device over the data bus (D0–D9). Internally, data is latched on the rising edge of CE. R/W Read/Write Control (TTL Compatible Input). This input determines whether data is written to or read from the device’s registers and color palette RAM. R/W and CE must be at Logic “0” to write data to the part. R/W must be at Logic “1” and CE at Logic “0” to read from the device. C0, C1 Command Controls (TTL Compatible Inputs). These inputs determine the type of read or write operation being performed on the device over the data bus, (see Interface Truth Table). Data on these inputs is latched on the falling edge of CE. IOR, IOG, IOB Red, Green & Blue Current Outputs (High Impedance Current Sources). These RGB video outputs are specified to directly drive RS-343A and RS-170 video levels into doubly terminated 75 Ω loads. VREF Voltage Reference Input (Analog Input): An external 1.235 V voltage reference is required to drive this input. An AD589 (2-terminal voltage reference) or equivalent is recommended. (Note: It is not recommended to use a resistor network to generate the voltage reference.) RSET Output Full Scale Adjust Control (Analog Input). A resistor connected between this pin and analog ground controls the absolute amplitude of the output video signal. For a value of RSET of nominally 280 Ω, with 37.5 Ω termination and using CR43 and CR44 of Command Register 4 to set the DAC Gain as shown, the required Video Standard can be achieved. CR44 0 0 1 1 CR43 0 1 0 1 Video Standard RS343A, Sync & Pedestal RS343A, Sync & No Pedestal RS343A, No Sync & No Pedestal RS170, Sync & Pedestal DAC Gain 3996 4224 4311 5592 Black to White 660 mV 17.62 mA 699 mV 18.63 mA 714 mV 19.05 mA 925 mV 24.67 mA Alternatively, RSET can be calculated by the following equation: RSET DAC Gain ×VREF Black to White Current COMP Compensation Pin. A 0.1 µF capacitor should be connected between this pin and VAA. VAA Power Supply (+5 V ± 5%). The part contains multiple power supply pins, all should be connected together to one common +5 V filtered analog power supply. GND: Analog Ground. The part contains multiple ground pins, all should be connected together to the system’s ground plane. TMS, TCK, TDI, TDO These four pins control the JTAG test access port. See Appendix 6 for more detail –14– REV. 0 ADV7160/ADV7162 (Continued from page 1) The ADV7160/ADV7162 integrates a number of graphic functions onto one device allowing 24-bit direct True-Color (30-bit Corrected-Color) operation at the maximum screen resolution of 1600 × 1280 at a refresh rate of 85 Hz. The ADV7160/ ADV7162 integrates a 256 × 30 Color Palette RAM with three high speed, 10-bit, digital-to analog converters (RGB DACs). It also contains a user-definable, X-Windows compatible, 64 × 64 × 2 cursor generator and associated RAM. An on-board Overlay Palette RAM is also included. The device’s 96-bit Programmable Pixel Port enables various data formats to be input to the part. An on-board clock and synchronization circuit controls all clocking functions for both the part and graphics subsystem. There are two video data paths through the ADV7160/ADV7162. One routes the data from the pixel port through the RAM to the DACs, the other bypasses the RAM and routes data direct from the pixel port to the DACs. Either path can be selected on a pixel by pixel basis. This allows for the overlay of an active video window on a graphics background. The on-board palette priority select inputs enable multiple palette devices to be connected together for use in multipalette and window applications. The part is controlled and programmed through the microprocessor (MPU) port. CIRCUIT DETAILS AND OPERATION OVERVIEW Digital video or pixel data is latched into the ADV7160/ADV7162 over the devices Pixel Port. This data acts as a pointer to onboard Color Palette RAM. The data at the RAM address pointed to is latched to the digital-to-analog converters (DACs) and output as an RGB analog video signal. The 30 bits of resolution, associated with the color look-up table and triple 10-bit DAC, realizes 24-bit True-Color resolution, while also allowing for the on-board implementation of linearization algorithms, such as Gamma-Correction and Monitor Callibration. This allows effective 30-bit True-Color operation. The on-chip video clock controller circuit generates all the internal clocking and some additional external clocking signals. The high accuracy, low jitter on board PLL eliminates the need for an external high speed clock generator. The PLL can be programmed to produce a pixel clock that is a multiple of the PLL reference clock. The ADV7162 is packaged in a standard plastic 160-pin quad flatpack (QFP). The ADV7160 is packaged in a plastic 160-pin power quad flatpack (PQUAD). Superior thermal distribution is achieved by the inclusion of a copper heatslug, within the standard package outline, to which the die is attached. This part is ideally suited for high performance applications where external environmental conditions are unpredictable and uncontrollable. one TTL input signal PLLREF are required to get the part operational. No additional signals or external glue logic are required to get the Pixel Port and Clock Control Circuit of the part operational. RED 8 GREEN 8 BLUE 8 For the purposes of clarity of description, the ADV7160/ADV7162 is broken down into three separate functional blocks. These are: 24 A B C 1. Pixel Port and Clock Control Circuit 24 24 24 MULTIPLEXER 24 D 2. MPU Port, Registers and Color Palette 3. Digital-to-Analog Converters and Video Outputs Figure 15. Multiplexed Color Inputs for the ADV7160/ADV7162 Pixel Port & Clock Control Circuit The Pixel Port of the ADV7160/ADV7162 is directly interfaced to the video/graphics pipeline of a computer graphics subsystem. It is connected directly or through a gate array to the video RAM of the systems Frame-Buffer (video memory). The pixel port on the device consists of: Color Data Pixel Controls Palette Selects RED, GREEN, BLUE SYNC, BLANK, TRISYNC PS0A-D, PS1A-D The associated clocking signals for the pixel port include: Clock Inputs Clock Outputs CLOCK, CLOCK, PLLREF, LOADIN, SCKIN LOADOUT, PRGCKOUT, SCKOUT These on-board clock control signals are included to simplify interfacing between the part and the frame buffer. Either two control input signals CLOCK and CLOCK (ECL Levels) or REV. 0 Pixel Port (Color Data) The ADV7160/ADV7162 has 96 color data inputs. The part has four (for 4:1 multiplexing) 24-bit wide direct color data inputs. These are user programmed to support a number of color data formats including 24-bit True-Color, 16-bit True-Color, 15-bit True-Color in 4:1 and 2:1 multiplex modes, and 8-bit Pseudo-Color (see “Multiplexing” section) in 8:1, 4:1 and 2:1 multiplex modes. Color data is latched into the parts pixel port on every rising edge of LOADIN (see Timing Waveform, Figure 4). The required frequency of LOADIN is determined by the multiplex rate, where 8:1 multiplex mode fLOADIN = fCLOCK/8 4:1 multiplex mode fLOADIN = fCLOCK/4 2:1 multiplex mode fLOADIN = fCLOCK/2 –15– ADV7160/ADV7162 Other pixel data signals latched into the device by LOADIN include SYNC, BLANK, TRISYNC and PS0A-D – PS1A-D. However, in 8:1 Mode, for 8-Bit Pseudo Color, the unused Blue Pixel Inputs are used to provide 8 extra PS inputs. The bypass mode is unavailable in this case. Internally, data is pipelined through the part by the differential pixel clock inputs, CLOCK and CLOCK or by the internal pixel clock generated by the PLL on-board. The LOADIN control signal need only have a frequency synchronous relationship to the pixel CLOCK (see “Pipeline Delay & On-Board Calibration” section). A completely phase independent LOADIN signal can be used with the ADV7160/ADV7162, allowing the CLOCK to occur anywhere during the LOADIN cycle. Palette Select Mode Alternatively, the LOADOUT signal of the ADV7160/ADV7162 can be used. LOADOUT can be connected either directly or indirectly to LOADIN. Its frequency is automatically set to the correct LOADIN requirement. These pixel port select inputs effectively determine whether the devices RGB analog outputs are turned-on or shut down. When the analog outputs are shut down, IOR, IOG and IOB are forced to 0 mA regardless of the state of the pixel and control data inputs. This state is determined on a pixel by pixel basis as the PS0–PS1 inputs are multiplexed in exactly the same format as the pixel port color data. These controls allow for switching between multiple palette devices. If the values of PS0 and PS1 match the values programmed into bits MR16 and MR17 of the Mode Register, then the device is selected, if there is no match the device is effectively shut down. SYNC, BLANK Bypass Mode Control (ADV7160 Only) The BLANK and SYNC video control signals drive the analog outputs to the Blank and Sync levels respectively. These signals are latched into the part on the rising edge of LOADIN. The SYNC information is encoded onto the IOG analog signal when Bit CR22 of Command Register 2 is set to “1,” the IOR analog signal when Bit CR41 of Command Register 4 is set to “1” and the IOB analog signal when Bit CR42 of Command Register 4 is set to “1.” The SYNC input is ignored if CR22, CR41 and CR42 are set to logic “0.” In this mode PS1 is used to switch between one of the color modes through the Color Palette and one of the Palette Bypass modes on a pixel by pixel basis. The color mode through the palette is selected using Bits CR27–CR24 of Command Register 2. The Bypass Color Mode is selected using Bits CR17 and CR16 of Command Register 1. PS1 then switches between the Palette Color Mode, and the Bypass Color Mode. The PS0 input continues to act as an overlay input, allowing Overlay Color 1 to be displayed. PS0 0 0 1 SYNCOUT In some applications where it is not permissible to encode SYNC on green (IOG), blue (IOB), or red (IOR), SYNCOUT can be used as a separate TTL digital SYNC output. This has the advantage over an independent (of the ADV7160/ADV7162) SYNC in that it does not necessitate knowing the absolute pipeline delay of the part. This allows complete independence between LOADIN/Pixel Data and CLOCK. The SYNC input is connected to the device as normal with Bit CR22 of Command Register 2, Bit CR41 of Command Register 4 and Bit CR42 of Command Register 4 are set to “0” thereby preventing SYNC from being encoded onto IOG, IOR and IOB. The output signal generates a TTL SYNCOUT with correct pipeline delay which is capable of directly driving the composite SYNC signal of a computer monitor. PS1 0 1 x Color Mode Palette Color Mode (CR27–CR24) Bypass Color Mode (CR17–CR16) Overlay Color 1 This mode is not available if using the ADV7162. Overlay Color Mode In this mode, the PS inputs provide control for a three color overlay. Whenever the value other than “00” is placed on the overlay inputs, the corresponding overlay color is displayed. When the overlay inputs contain “00” the color is specified by the main pixel inputs. CLOCK CONTROL CIRCUIT This input is used to generate a HDTV Sync on any of the DAC outputs. Bit CR17 of Command Register 1 is set to “1”, enabling TRISYNC. When TRISYNC is low, the analog output which has Sync enabled goes to the tri-sync level. The ADV7160/ADV7162 has an integrated Clock Control Circuit (Figure 16). This circuit is capable of both generating the ADV7160/ADV7162’s internal clocking signals as well as external graphics subsystem clocking signals. Total system synchronization can be attained by using the parts output clocking signals to drive the controlling graphics processor’s master clock as well as the video frame buffers shift clock signals. PS0A-D–PS1A-D (Palette Priority Select Inputs) CLOCK, CLOCK Inputs These multifunctional TTL compatible inputs can be configured for three separate functions. The eight PS inputs are multiplexed to provide two bits which are used to provide one of three different functions. The function is selected by Bit CR14 and Bit CR15 of Command Register 1. The Clock Control Circuit is driven by the pixel clock inputs, CLOCK and CLOCK. These inputs can be driven by a differential ECL oscillator running from a +5 V supply. TRISYNC CR15 0 0 1 1 CR14 0 1 0 1 Color Mode Palette Select Mode Bypass Mode Control (ADV7160 Only) Overlay Color Mode Ignore PS Inputs –16– REV. 0 ADV7160/ADV7162 LOADOUT(1) LOADOUT PLLREF CLOCK CLOCK PLL ECL TO TTL VIDEO FRAME BUFFER S E L E C T LOADIN LOADOUT ADV7160/ ADV7162 VIDEO FRAME BUFFER LOADOUT(2) LOADIN PIXEL DATA DIVIDE BY N (÷N) PRGCKOUT ADV7160/ ADV7162 PIXEL DATA DIVIDE BY M (÷M) LOADOUT LOADOUT LOADOUT(1) DELAY SCKOUT LATCH LOADOUT(2) LOADIN TRISYNC BLANK EN SYNC Figure 17. LOADOOUT vs Pixel Clock SCKIN Pipeline Delay and Onboard Calibration ADV7160/ ADV7162 LOADIN TO COLOR DATA MULTIPLEXER M IS A FUNCTION OF MULTIPLEX RATE M = 8 IN 8:1 MULTIPLEX MODE M = 4 IN 4:1 MULTIPLEX MODE M = 2 IN 2:1 MULTIPLEX MODE N IS INDEPENDENTLY PROGRAMMABLE N = (4, 8, 16, 32) Figure 16. Clock Control Circuit of the ADV7160/ADV7162 The ADV7160/ADV7162 has a fixed number of pipeline delays (tPD), so long as timings t10 and τ–t11 are met. However, if a fixed number of pipeline delays is not a requirement, timings t10 and τ–t11 can be ignored, a calibration cycle must be run and there is no restriction on LOADIN to LOADOUT timing. If timings t10 and τ–t11 are not met, the part will function correctly though with an increased number of pipeline delays. The ADV7160/ADV7162 has on-board calibration circuitry which synchronizes pixel data and LOADIN with the internal ADV7160/ADV7162 clocking signals. Calibration can be performed in two ways. During the device’s initialization sequence by toggling two bits of the Mode Register, MR10 followed by MR15 or by writing a “1” to Bit CR10 of Command Register 1 and a “0” to MR15 which executes a calibration on every Vertical Sync. PRGCKOUT The PRGCKOUT control signal outputs a user programmable clock frequency. It is a divided down frequency of the pixel CLOCK (see Figure 11). The rising edge of PRGCKOUT is synchronous to the rising edge of LOADOUT. CLOCK CONTROL SIGNALS LOADOUT The ADV7160/ADV7162 generates a LOADOUT control signal which runs at a divided down frequency of the pixel CLOCK. The frequency is automatically set to the programmed multiplex rate, controlled by CR37 and CR36 of Command Register 3. fLOADOUT = fCLOCK/8 fLOADOUT = fCLOCK/4 fLOADOUT = fCLOCK/2 fPRGCKOUT = fCLOCK/N where N = 4, 8, 16 or 32. One application of the PRGCKOUT is to use it as the master clock frequency of the graphics subsystems processor or controller. 8:1 multiplex mode 4:1 multiplex mode 2:1 multiplex mode The LOADOUT signal is used to directly drive the LOADIN pixel latch signal of the ADV7160/ADV7162. This is most simply achieved by tying the LOADOUT and LOADIN pins together. Alternatively, the LOADOUT signal can be used to drive the frame buffer’s shift clock signals, returning to the LOADIN input delayed with respect to LOADOUT. If it is not necessary to have a known fixed number of pipeline delays, then there is no limitation on the delay between LOADOUT and LOADIN (LOADOUT(1) and LOADOUT(2)). LOADIN and Pixel Data must conform to the setup and hold times (t8 and t9). SCKIN, SCKOUT These video memory signals are used to minimize external support chips. Figure 18 illustrates the function that is provided. An input signal applied to SCKIN is synchronously AND-ed with the video blanking signal (BLANK). The resulting signal is output on SCKOUT. Figure 12 of the Timing Waveform section shows the relationship between SCKOUT, SCKIN and BLANK. SCKOUT LATCH BLANK If however, it is required that the ADV7160/ADV7162 has a fixed number of pipeline delays (tPD) LOADOUT and LOADIN must conform to timing specifications t10 and τ–t11 as illustrated in Figures 5 to 10. SYNC ENABLE SCKIN Figure 18. SCKOUT Generation Circuit REV. 0 –17– ADV7160/ADV7162 The SCKOUT signal is essentially the video memory shift control signal. It is stopped during the screen retrace. Figure 19 shows a suggested frame buffer to ADV7160/ADV7162 interface. This is a minimum chip solution and allows the ADV7160/ADV7162 control the overall graphics system clocking and synchronization. LOADOUT LOADIN SCKIN VIDEO FRAME BUFFER ADV7160/ ADV7162 BLANK SCKOUT PLL V Register can be set from 01H to 7FH. It should not be set to 00H. If this register contains 00H, then the PLL stops. Therefore the feedback divider can be set from 12 to 519 in steps of one, or from 520 to 1038 in steps of two by setting the VSEL bit. The VSEL bit is accessed by changing bit PCR2 of the PLL Control Register. The P counter divides the output from the oscillator by 1, 2, 4 or 8 as determined by PSEL1 and PSEL0 which are set in bits PCR5 and PCR4 of the PLL Control Register. This post-scaler is useful in the generation of lower frequencies as the VCO has been optimized for high frequency operation. VCO PIXEL DATA PLLREF VCO/2 (1 + VSEL)(4(V+2) + S) (1 + RSEL)(R+2) FVCO VCO/4 FOUT VCO/8 Figure 19. ADV7160/ADV7162 Interface Using SCKIN and SCKOUT PLL The on-board PLL can be used as an alternative clock source. This eliminates the need for an external high speed clock generator such as a crystal oscillator. With the PLL, it is possible to generate an internal clock whose frequency is a multiple of the PLL reference frequency (PLLREF). Internal PLL operation is selected by setting CR56 of Command Register 5 to Logic “1.” The PLL registers can be programmed to set up the frequency required. The block diagram of the Phase Locked Loop is shown in Figure 20. The blocks consist of a phase frequency detector, a charge pump, a loop filter, a voltage controlled oscillator and a programmable divider. REFERENCE DIVIDER FPD PHASE DETECTOR CHARGE PUMP PSEL1 PSEL0 0 0 FVCO/2 FVCO/4 0 1 1 0 FVCO/8 1 1 PSEL1 PSEL0 Figure 21. PLL Transfer Function The transfer function of the PLL can be summarized by the block diagram shown in Figure 21. To optimize the performance of the on-board PLL, the following criteria should be followed: 900 kHz 300 kHz 120 MHz < PLLREF < FPD < FVCO < 40 MHz < 10 MHz < 260 MHz For FVCO > 220 MHz, VSEL should be programmed to logic “0.” VOLTAGE CONTROLLED OSCILLATOR Any lower frequency output can be achieved by using the output divider. FPD FVCO O/P DIVIDER FOUT FEEDBACK DIVIDER Figure 20. PLL Block Diagram The phase frequency detector drives the voltage controlled oscillator (VCO), to a frequency that will cause the two inputs to the phase frequency detector to be matched in frequency and phase. The corresponding output of the VCO can be calculated as: VCO = PLLREF Feedback Divider Reference Divider The Reference Divider is set by a combination of the contents of the PLL R Register and the RSEL bit. The PLL R Register has a resolution of 7 bits. It is programmed by setting the PLL R Register located at Control Register address 00CH . The PLL R Register can be set from 01H to 7FH. It should not be set to 00H. If this register contains 00H, then the PLL stops. Therefore, the Reference Divider can be set from 3 to 129 in steps of one, or from 130 to 258 in steps of two by setting the RSEL bit. The RSEL bit is accessed by changing Bit PCR1 of the PLL Control Register. The Feedback Divider is set by a combination of the contents of the PLL V Register, the VSEL bit and the S value. The S value is set up in PCR7 and PCR6 of the PLL Command Register. This S value allows a better resolution when setting the Feedback Divider value. The PLL V Register has a resolution of 7 bits. It is programmed by setting the PLL V Register located at Control Register address 00FH .The A jitter performance graph as a function of both FPD and FVCO is illustrated in Figure 22. It can be seen that jitter decreases with increasing FVCO and also that jitter decreases with increasing FPD. For each FOUT, the user should firstly maximize FVCO using the output divider and then pick PLLREF and reference divide to maximize FPD. When generating multiple output frequencies from one PLLREF value, an iterative process should be used to find the PLLREF value that gives the best trade off between jitter performance and FOUT accuracy. –18– 250 JITTER MEASURED AT 15µs FPD = 0.3MHz FPD = 0.42MHz 200 FPD = 0.57MHz RMS JITTER – ps PLLREF FOUT FVCO 150 FPD FPD FPD FPD 100 = 0.8MHz = 1.0MHz = 1.5MHz = 2.0MHz FPD = 2.7MHz FPD = 4.0MHz FPD = 5.3MHz 50 0 50 100 150 200 VCO FREQUENCY – MHz 250 300 Figure 22. PLL Jitter REV. 0 ADV7160/ADV7162 COLOR VIDEO MODES The ADV7160/ADV7162 supports a number of color video modes all at the maximum video rate. Command bits CR27–CR24 of Command Register 2 along with bit MR11 of Mode Register 1 determine the color mode. Seven color modes use the Color Palette, and three of them bypass the palette and control the DACs directly. DACs with 30-bit data, allowing the display of 15-bit GammaCorrected True-Color Images. With MR11 set to Logic “0,” the Look-Up Table is configured as a 32 location by 24 bits deep RAM (8 bits each for Red, Green and Blue) and the output of the RAM drives the DACs with 24-bit data, allowing the display of 15-bit True-Color Images. 15-BIT COLOR DATA 24-Bit True Color (CR27, CR26, CR25, CR24 = 1, 1, 1, 0) The part is set to 24-bit/30-bit “Gamma” True-Color operation with MR11 set to Logic “1” and direct 24-bit True-Color operation with MR11 set to Logic “0.” The pixel port accepts 24 bits of color data which is directly mapped to the Look-Up Table RAM. With MR11 set to Logic “1,” the Look-Up Table is configured as a 256 location by 30 bits deep RAM (10 bits each for Red, Green and Blue), the RAM is preloaded with a user determined, nonlinear function, such as a gamma correction curve and the output of the RAM drives the DACs with 30-bit data. With MR11 set to Logic “0,” the Look-Up Table is configured as a 256 location by 24 bits deep RAM (8 bits each for Red, Green and Blue), the RAM is preloaded with a linear function and the output of the RAM drives the DACs with 24bit data. 24-BIT COLOR DATA 24-BIT TO 30-BIT LOOK-UP TABLE RED 256 x 10 10 10-BIT RED DAC 10 10-BIT GREEN DAC GREEN OUT 10 10-BIT BLUE DAC BLUE OUT 8 8 8 GREEN 256 x 10 BLUE 256 x 10 ANALOG VIDEO OUTPUTS 30-BIT COLOR DATA RED OUT 15-BIT TO 24-BIT LOOK-UP TABLE RED 32 x 8 GREEN 32 x 8 BLUE 32 x 8 8 8-BIT GREEN DAC 8 8-BIT BLUE DAC RED OUT GREEN OUT BLUE OUT Figure 24. 15-Bit to 24-Bit True-Color Configuration 8-Bit Pseudo Color (CR27, CR26, CR25, CR24 = 0, 0, 0, 0 or 0, 1, 0, 0 or 1, 0, 0, 0) This mode sets the part into 8-bit Pseudo-Color operation. The pixel port accepts 8 bits of pixel data, from either the red, blue or green channel. With MR11 set to Logic “1,” a 30-bit word is indexed in the Look-Up Table RAM. The Look-Up Table is configured as a 256 location by 30 bits deep RAM (10 bits each for Red, Green and Blue). The output of the RAM drives the DACs with 30-bit data. With MR11 set to Logic “0,” a 24-Bit word is indexed in the Look-Up Table RAM. The Look-Up Table is configured as a 256 location by 24 bits deep RAM (8 bits each for Red, Green and Blue). The output of the RAM drives the DACs with 24-bit data. This mode allows for the display of 256 simultaneous colors out of a total palette of millions of addressable colors. 8-BIT TO 30-BIT LOOK-UP TABLE Figure 23. 24-Bit to 30-Bit True-Color Configuration 8 16-Bit True Color (CR27, CR26, CR25, CR24 = 1, 0, 1, 1) RED 256 x 10 GREEN 256 x 10 BLUE 256 x 10 ANALOG VIDEO OUTPUTS 30-BIT COLOR DATA 10 10-BIT RED DAC 10 10-BIT GREEN DAC GREEN OUT 10 10-BIT BLUE DAC BLUE OUT RED OUT Figure 25. 8-Bit to 30-Bit Pseudo-Color Configuration PIXEL PORT MAPPING The pixel data to the ADV7160/ADV7162 is automatically mapped in the parts pixel port as determined by the pixel data mode programmed (Bits CR27–CR24 of Command Register 2). 15-Bit True Color (CR27, CR26, CR25, CR24 = 1, 1, 0, 0 or 1, 1, 0, 1) The part is set to 15-bit True-Color operation. The pixel port accepts 15 bits of color data which is mapped to the 5 LSBs of each of the red, green and blue palettes of the Look-Up Table RAM. With MR11 set to Logic “1,” the Look-Up Table is configured as a 32 location by 30 bits deep RAM (10 bits each for Red, Green and Blue) and the output of the RAM drives the REV. 0 8-BIT RED DAC 5 8-BIT PIXEL DATA The part is set to 16-bit True-Color operation. The pixel port accepts 16 bits of color data which is mapped to the 5 LSBs of each of the red and blue palettes of the Look-Up-Table RAM, and 6 LSBs of the green palette of the Look-Up-Table RAM. With MR11 set to Logic “1,” the Look-Up Table is configured as a 64 location by 30 bits deep RAM (10 bits each for Red, Green and Blue) and the output of the RAM drives the DACs with 30-Bit data, allowing the display of 16-bit GammaCorrected True-Color Images. With MR11 set to Logic “0,” the Look-Up Table is configured as a 64 location by 24 bits deep RAM (8 bits each for Red, Green and Blue); and the output of the RAM drives the DACs with 24-bit data, allowing the display of 16-bit True-Color Images. 8 5 5 ANALOG VIDEO OUTPUTS 24-BIT COLOR DATA Pixel data in the 24-bit True-Color modes is directly mapped to the 24 color inputs R7–R0, G7–G0 and B7–B0. There is one mode of operation for 16-bit True Color. Data is input to the device over the red and green color ports (R7–R0 and G7–G0) and is internally mapped to LUT Locations 0–63 according to Figure 26. (Note: Data on unused pixel inputs is ignored.) . –19– ADV7160/ADV7162 R4 R3 R2 R1 R0 G5 G4 G3 G2 G1 G0 B4 B3 B2 B1 B0 x x x x x x x x R7 R6 R5 R4 R3 R2 R1 R0 G7 G6 G5 G4 G3 G2 G1 G0 B7 B6 B5 B4 B3 B2 B1 B0 R7 0 R6 0 x R4 256 x 10 RAM (RED LUT) R3 R5 0 R4 R4 R2 R3 R3 R1 R2 R2 R1 R1 R0 R0 5 LOCATION "31" LOCATION "0" R0 10 G4 TO RED DAC G3 R3 R2 R1 R0 x x x G4 G3 G2 G1 G0 x x x B4 R6 R5 R4 R3 R2 R1 R0 G7 G6 G5 G4 G3 G2 G1 G0 B7 B3 B6 B2 B5 B4 B1 B0 x x x B3 B2 B1 B0 R5 R4 R3 R2 R1 R0 G1 G0 G0 0 x B7 x x 0 x B6 x 0 x B5 B4 x B2 x B1 x B0 B4 B3 B2 B1 B0 x 5 x LOCATION "31" LOCATION "0" DATA LATCHES FIRST 32 OR 64 LOCATIONS OF RAM x 10 x x TO BLUE DAC R4 0 R3 0 R2 0 R1 R4 R0 x R3 x R1 x R0 G4 0 G3 0 G2 0 G1 G4 G0 x G3 x G1 x G0 B4 0 B3 0 B2 0 B1 B4 B3 x B1 x LOCATION "31" LOCATION "0" 5 LOCATION "31" LOCATION "0" 5 B0 DATA INTERNALLY SHIFTED TO 5 LSBs LOCATION "31" LOCATION "0" DATA LATCHES FIRST 32 LOCATIONS OF RAM G1 G0 B3 B2 B1 B0 G4 G3 G2 G2 G1 G1 G0 G0 x 0 x 0 x x 0 B4 x B3 x B2 x B1 x B0 5 LOCATION "31" LOCATION "0" 10 TO RED DAC 256 x 10 RAM (GREEN LUT) 5 LOCATION "31" LOCATION "0" 10 TO GREEN DAC 256 x 10 RAM (BLUE LUT) 5 DATA INTERNALLY SHIFTED TO 5 LSBs LOCATION "31" LOCATION "0" DATA LATCHES FIRST 32 LOCATIONS OF RAM 10 TO BLUE DAC The part has two modes of operation for 15-bit True Color. In the first mode, data is input to the device over the red, green and blue channel (R7–R3, G7–G3 and B7–B3) and is internally mapped to Locations 0 to 31 of the Look-Up Table (LUT) according to Figure 27. 10 In the second mode, data is input to the device over just two of the color ports, red and green (R7–R0 and G7–G0) and is internally mapped to LUT Locations 0 to 31 according to Figure 30. (Note: Data on unused pixel inputs is ignored.) TO RED DAC There are three modes of operation for 8-bit Pseudo Color. Each mode maps the input pixel data differently. Data can be input into one of the three color channels, R7–R0 or G7–G0 or B7–B0. 10 In 24-bit Palette Bypass Mode, the red, blue and green color channels bypass the Pixel Mask and the Color Palette. Each 8bit color channel is mapped onto the 8 MSBs of the corresponding 10-bit DAC input. The two LSBs on each DAC are zeros. The Bypass Mode can be selected in two ways, by using CR27– CR24 of Command Register 2 or on a pixel by pixel basis using the PS inputs (ADV7160 only). TO GREEN DAC 256 x 10 RAM (BLUE LUT) B2 G2 G4 G3 256 x 10 RAM (RED LUT) Figure 28. 15-Bit True-Color Mapping using R6–R0 and G7–G0 256 x 10 RAM (GREEN LUT) G2 B0 x DATA PIN PIXEL INPUT ASSIGN- LATCHED TO DATA MENTS PIXEL PORT 5 G3 DATA PIN PIXEL INPUT ASSIGN- LATCHED TO DATA MENTS PIXEL PORT 256 x 10 RAM (RED LUT) R2 R0 0 G1 B3 R0 0 G2 B4 R1 G5 G2 x R1 G5 G4 G3 x R2 G0 G4 256 x 10 RAM (BLUE LUT) R2 0 G3 TO GREEN DAC R4 R3 G6 G4 LOCATION "0" R4 R3 G7 G5 10 0 G6 G5 LOCATION "63" 0 R5 G7 0 5 0 G1 0 256 x 10 RAM (GREEN LUT) x R6 G2 G7 Figure 26. 16-Bit True-Color Mapping using R7–R0 and G7–G0 R7 R6 G6 DATA DATA PIN PIXEL INPUT ASSIGN- LATCHED INTERNALLY SHIFTED TO DATA MENTS PIXEL TO 5 OR 6 LSBs PORT R4 R7 10 TO BLUE DAC Figure 27. 15-Bit True Color Mapping using R7–R3, G7–G3 and B7–B3 –20– In 16-bit Palette Bypass Mode, the color channels bypass the Pixel Mask and the Color Palette. The 8-bits of red pixel data and 8-bits of green pixel data are mapped onto the 5 MSBs of the red and blue DAC input and the 6 MSBs of the green DAC input as shown in Figure 29. The remaining LSBs on each DAC are zeros. The Bypass Mode can be selected in two ways, by using CR27–CR24 of Command Register 2 or on a pixel by pixel basis using the PS inputs (ADV7160 only). REV. 0 ADV7160/ADV7162 R9 R8 R7 R6 R5 G9 G8 R7 R6 R5 R4 R3 R2 R1 R0 G7 In 15-bit Palette Bypass Mode, the color channels bypass the Pixel Mask and the Color Palette. The 7 bits of red pixel data and 8 bits of green pixel data are mapped onto the 5 MSBs of the red, green and blue DAC input as shown in Figure 30. The remaining LSBs on each DAC are zeros. The Bypass Mode can be selected in two ways, by using CR27–CR24 of Command Register 2 or on a pixel by pixel basis using the PS inputs (ADV7160 only). R9 R6 R8 R5 R7 R4 R6 R3 R5 R2 0 R1 0 R0 0 G7 IOR RED DAC 0 G9 x G6 G5 B9 B8 B7 B6 B5 x x x x x x x x G7 G6 G5 G4 G3 G2 G1 G0 B7 B6 B5 B4 B3 B2 B1 B0 G7 G6 G5 G4 G3 G2 G1 G0 Multiplexing G8 The on-board multiplexers of the ADV7160/ADV7162 eliminate the need for external data serializer circuits. Multiple video memory devices can be connected, in parallel, directly to the device. Figure 31 shows four memory banks of 50 MHz memory connected to the ADV7160, running in 4:1 multiplex mode, giving a resultant pixel or dot clock rate of 200 MHz. Instead of having to provide a new pixel at the input every 5 ns, four pixels are provided together every 20 ns. The input multiplexer takes the four pixels latched in parallel, and selects them one at a time to produce a pixel stream at the pixel clock rate. In 4:1 mode, the pixels are selected in the sequence A, B, C, D, cycling continuously. In 2:1 mode, the A and B pixels are selected. The 8:1 mode is only available in 8-bit Pseudo-Color Mode. BLANK, SYNC, ODD/EVEN and TRISYNC are not multiplexed and can only change on a 1, 2, 4 or 8 pixel boundary depending on the multiplex mode. G7 G6 G5 G4 IOG 0 0 0 GREEN DAC 0 x x x x x x x x B9 B8 B7 B6 B5 IOB 0 0 0 BLUE DAC 0 0 PIXEL INPUT DATA PIN ASSIGNMENTS DATA LATCHED TO PIXEL PORT DATA LATCHED TO DAC INPUTS Figure 29. 16-Bit True-Color in Bypass Mode using R7–R0 and G7–G0 x R9 R8 R7 R6 R5 G9 G8 R7 R6 R5 R4 R3 R2 R1 R0 R9 x R8 R6 R7 R5 R6 R4 R5 R3 0 R2 0 R1 0 R0 0 On the rising edge of LOADIN, all the pixel port inputs are latched into the ADV7160/ADV7162. The LOADIN frequency must be a divided down frequency of the pixel clock frequency. This can be achieved using LOADOUT to directly drive LOADIN as LOADOUT provides the correct frequency required, or drive LOADIN after delay through some external circuitry. VIDEO MEMORY/ FRAME BUFFER 24 IOR VRAM (BANK A) 50MHz 50MHz 50MHz VRAM (BANK B) 50MHz 50MHz 50MHz RED DAC G9 G7 G6 G5 B9 B8 B7 B6 B5 x x x x x x x x G7 G6 G5 G4 G3 G2 G1 G0 B7 B6 B5 B4 B3 B2 B1 B0 G7 G6 G5 G4 G3 G2 G1 G0 G7 24 VRAM (BANK C) 50MHz 50MHz 50MHz VRAM (BANK D) 50MHz 50MHz 50MHz 200MHz (4 × 50MHz) G6 G5 0 IOG 24 0 0 0 GREEN DAC 0 x x x x x x x x Figure 31. Direct Interfacing of Video Memory to ADV7160/ADV7162 B9 B8 B7 8-Bit Pseudo Color in 8:1 Multiplexing Mode B6 B5 IOB 0 0 0 BLUE DAC DATA LATCHED TO PIXEL PORT DATA LATCHED TO DAC INPUTS Fiigure 30. 15-Bit True-Color in Bypass Mode using R6–R0 and G7–G0 REV. 0 24 MULTIPLEXER 0 PIN ASSIGNMENTS 24 G8 0 PIXEL INPUT DATA ADV7160/ADV7162 When 8:1 Multiplexing Mode is selected by setting Bit CR37 of Command Register 3 to Logic “1” and bit CR36 of Command Register 3 to Logic “0,” the ADV7160/ADV7162 goes into 8Bit Pseudo-Color Mode irrespective of the Color Mode selected by Bits CR27 to CR24 in Command Register 2. Hence LOADOUT operates at fCLOCK/8. Eight 8-bit pixels are latched in parallel by the rising edge of LOADIN. These 8-bit pixels are then selected, one at a time, to produce an 8-bit pixel stream which passes through the Pixel Mask to address the LUT. The order the eight 8-bit pixels are displayed is GA, RA, GB, RB, GC, RC, GD, RD. –21– ADV7160/ADV7162 The unused Blue pixel inputs are used, in this mode, to provide 8 extra PS inputs. These PS inputs provide 2 bits after 8:1 multiplexing. The PS inputs can be used as Overlay or Palette Select inputs. 24-BIT TO 30-BIT A G7–G0 B C D E F G H R7–R0 G7–G0 R7–R0 G7–G0 R7–R0 G7–G0 R7–R0 A B C D E F G PS1–PS0 B1–B0 PS1–PS0 B1–B0 PS1–PS0 B1–B0 PS1–PS0 H B1–B0 8 8 8 8-BIT COLOR DATA P I X E L 8 8 8 RED 256 x 10 I N P U T 8 8 GREEN 256 x 10 8 M U L T I P L E X E R 8 8 8 8 8 8 8 ANALOG VIDEO OUTPUTS 30-BIT COLOR DATA LOOK-UP-TABLE BLUE 256 x 10 10 10-BIT RED DAC 10 10-BIT GREEN DAC GREEN OUT 10 10-BIT BLUE DAC BLUE OUT RED OUT 2 8 PS Figure 32. 8-Bit Pseudo Color in 8:1 Multiplexing Mode MICROPROCESSOR (MPU PORT) The ADV7160/ADV7162 supports a standard MPU Interface. All the functions of the part are controlled via this MPU port. Direct access is gained to the Address Register, Mode Register and all the Control Registers as well as the Color Palette. The following sections describe the setup for reading and writing to all of the devices registers. MPU Interface The MPU interface (Figure 33) consists of a bidirectional, 10bit wide databus and interface control signals CE, C0, C1 and R/W. The 10-bit wide databus is user configurable as illustrated. Table I. Data-Bus Width Data-Bus Width RAM/DAC Resolution Read/Write Mode 10-Bit 10-Bit 8-Bit 8-Bit 10-Bit 8-Bit 10-Bit 8-Bit 10-Bit Parallel 8-Bit Parallel 8 + 2 Byte 8-Bit Parallel Register Mapping The ADV7160/ADV7162 contains a number of on-board registers including the Mode Register (MR17–MR10), Address Register (A10–A0) and many Control Registers as well as Color Palette Registers. These registers control the entire operation of the part. Figure 34 shows the internal register configuration. Control lines C1 and C0 determine which register the MPU is accessing. C1 and C0 also determine whether the Address Register is pointing to the color registers and Look-Up Table RAM or the control registers. If C1, C0 = 1, 0 the MPU has access to whatever control register is pointed to by the Address Register (A10–A0). If C1, C0 = 0, 1 the MPU has access to the LookUp Table RAM (Color Palette) or the Overlay Palette through the associated color registers. The CE input latches data to or from the part. The R/W control input determines between read or write accesses. The truth tables show all modes of access to the various registers and color palette for both the 8-bit wide databus configuration and 10-bit wide data bus configuration. It should be noted that after power-up, the devices MPU port is automatically set to 10-bit wide operation (see Power-On Reset section). CONTROL REGISTERS STATUS REGISTER PIXEL MASK REGISTER ADDRESS REGISTER MODE REGISTER ADDR (A10-A0) (MRI) C1 C0 0 0 C1 C0 1 1 CURSOR REGISTERS COMMAND REGISTERS (CR1–CR5) DATA TO PALETTES TEST REGISTERS ID REGISTER C1 C0 1 0 30 REVISION REGISTER RED REGISTER PLL REGISTERS COLOR REGISTERS GREEN REGISTER BLUE REGISTER C1 C0 0 1 MPU PORT 10 (8+2) CE R/W C0 C1 D9–D0 Figure 33. MPU Port and Register Configuration –22– REV. 0 ADV7160/ADV7162 C1 1 ADDRESS REGISTER (A10–A0) C0 0 CONTROL REGISTERS 7FFH – 400H CURSOR IMAGE 3FFH – 305H RESERVED 304H CURSOR COLOR 1 303H CURSOR COLOR 2 302H – 205H RESERVED 204H CURSOR CONTROL REG 203H CURSOR Y-HI REG 202H CURSOR Y-LO REG 201H CURSOR X-HI REG 200H CURSOR X-LO REG 1FFH – 016H RESERVED 015H – 014H TEST REGISTER 013H SIGNATURE MISC REG 012H SIGNATURE BLUE REG 011H SIGNATURE GREEN REG 010H SIGNATURE RED REG 00FH PLL V REG 00EH TEST REG 00DH COMMAND REG 5 00CH PLL R REG 00BH REVISION REG 001H 00AH STATUS REG 009H PLL COMMAND REG 008H COMMAND REG 4 007H COMMAND REG 3 006H COMMAND REG 2 005H COMMAND REG 1 004H PIXEL MASK REG 003H ID REG 002H – 000H TEST REGISTERS MODE REGISTER (MR17–MR10) C1 ADDRESS REGISTER (A10–A0) C1 1 0 C0 1 C0 0 C1 0 RED REGISTER (R9-R0) C0 1 GREEN REGISTER (G9-G0) POINTS TO LOCATION CORRESPONDING TO ADDRESS REGISTER (A10–A0) ADDRESS REGISTER (A10–A0) COLOR PALETTE 7FFH – 104H RESERVED 103H – 101H OVERLAY COLOR 1–3 (3 x 30) 100H RESERVED 0FFH – 000H LOOK-UP TABLE RAM (256 x 30) Figure 34. Internal Register Configuration and Address Decoding REV. 0 BLUE REGISTER (B9-B0) –23– ADDRESS REG = ADDRESS REG +1 ADV7160/ADV7162 blue are concatenated into a single 30-bit/24-bit word and written to the RAM location as specified in the address register (A10–A0). Power-On Reset On power-up, the ADV7160/ADV7162 executes a power-on reset operation. This initializes the pixel port such that the pixel sequence ABCD starts at A. The Mode Register (MR17–MR10), Command Register 2 (CR27–CR20), Command Register 3 (CR37–CR30) have all bits set to a Logic “1” and Address Register, Command Register 1 (CR17–CR10), Command Register 4 (CR47–CR40) and Command Register 5 (CR57–CR50) have all bits set to a Logic “0.” The address register then automatically increments to point to the next RAM location and a similar red, green and blue palette write sequence is performed. The address register resets to 000H following a blue write cycle to color palette RAM location 0FFH. The three color overlay palette is located in address space above the main color palette. To access the Overlay Palette, the Address Register must first be written with address 101H. From then on, the colors are accessed in the same way as the main Color Palette, with the Address Register incrementing after each blue access. The output clocking signals are also set during this reset period. PRGCKOUT = CLOCK/32 LOADOUT = CLOCK/4: The power-on reset is activated when VAA goes from 0 V to 5 V This reset is active for 1 µs. The ADV7160/ADV7162 should not be accessed during this reset period. The pixel clock should be applied at power-up. Data is read from the Color Palette by firstly writing to the address register of the color palette location to be read. The MPU performs three successive read cycles from each of the red, green and blue locations (10-bit or 8-bit) of the RAM. Figures 35 to 38 illustrate read operations for a 10-bit databus using the DACs in 8-bit and 10-bit mode and read operations for an 8-bit databus using the DACs in 8-bit and 10-bit mode. An internal pointer moves from red to green to blue after each read is completed. This pointer is reset to red after a blue read or whenever the address register is written. The address register then automatically increments to point to the next RAM location and a similar red, green and blue palette read sequence is performed. The address register resets to 000H following a blue read cycle of color palette RAM location 0FFH. Similarly for the Overlay Palette, the Address Register must first be written with address 101H. From then on, the colors are read in the same way as the main Color Palette, with the Address Register incrementing after each blue access. Color Palette Accesses The Color Palette consists of 256 RAM locations, each location containing 30 bits of color information. Data is written to the color palette by firstly writing to the address register of the color palette location to be modified. The MPU performs three successive write cycles for each of the red, green and blue registers (10-bit or 8-bit). Figures 35 to 38 illustrate write operations for a 10-bit databus using the DACs in 8-bit and 10-bit mode and write operations for an 8-bit databus using the DACs in 8-bit and 10-bit mode. An internal pointer moves from red to green to blue after each write is completed. This pointer is reset to red after a blue write or whenever the address register is written. During the blue write cycle, the three bytes of red, green and First Write Operation Palette R9 Databus R8 D7 R7 R6 D6 D5 R5 D4 R4 D3 R3 D2 R2 D1 R1 R0 D0 Second Write Operation Palette R9 Databus R8 D7 R7 R6 D6 D5 R5 D4 R4 D3 R3 D2 R2 D1 R1 R0 R9 Databus R8 D7 R7 R6 D6 D5 R5 D4 R4 D3 R3 D2 R2 D1 R1 R0 D0 Second Read Operation Palette Databus R9 R8 R7 R6 R5 R4 R3 x x x x x x R2 D1 C1 C0 Palette Write 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 . . Write to Address Register (Lo- Byte) Write to Address Register (Hi- Byte) Write Red Data (R9–R2) Write Red Data (R1–R0) Write Green Data (G9–G2) Write Green Data (G1–G0) Write Blue Data (B9–B2) Write Blue Data (B1–B0) Write Red Data (R9–R2) R/W C1 C0 Palette Read 0 0 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 . . Write to Address Register (Lo- Byte) Write to Address Register (Hi- Byte) Read Red Data (R9–R2) Read Red Data (R1–R0) Read Green Data (G9–G2) Read Green Data (G1–G0) Read Blue Data (B9–B2) Read Blue Data (B1–B0) Read Red Data (R9–R2) D0 First Read Operation Palette R/W R1 R0 D0 Figure 35. 8-Bit Data Bus Using 10-Bit DACs –24– REV. 0 ADV7160/ADV7162 Register Accesses registers is direct. The Control Registers are accessed indirectly. The Address Register must point to the desired Control Register. Figure 33 and Figures 35 to 38 illustrate the structure and protocol for device communication over the MPU port. The MPU can write to or read from all of the ADV7160/ ADV7162’s registers. C0 and C1 determine whether the Mode Register or Address Register is being accessed. Access to these Write Operation Palette R9 R8 D7 Databus R7 R6 D6 D5 R5 D4 R4 D3 R3 D2 R2 D1 R1 D0 R/W C1 C0 Palette Write 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 . . Write to Address Register (Lo-Byte) Write to Address Register (Hi-Byte) Write Red Data (R9–R0) Write Green Data (G9–G0) Write Blue Data (B9–B0) Write Red Data (R9–R0) R/W C1 C0 Palette Read 0 0 1 1 1 1 0 0 0 0 0 0 0 0 1 1 1 1 . . Write to Address Register (Lo-Byte) Write to Address Register (Hi-Byte) Read Red Data (R9–R0) Read Green Data (G9–G0) Read Blue Data (B9–B0) Read Red Data (R9–R0) R0 0 0 Read Operation Palette R9 Databus R8 D7 R7 D6 R6 D5 R5 D4 R4 D3 R3 R2 D2 D1 R1 R0 D0 Figure 36. 8-Bit Databus Using 8-Bit DACs Write Operation Palette Databus R9 D9 R8 D8 R7 D7 R6 D6 R5 D5 R4 D4 R3 D3 R2 D2 R1 D1 R/W C1 C0 Palette Write 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 . . Write to Address Register (Lo-Byte) Write to Address Register (Hi-Byte) Write Red Data (R9–R0) Write Green Data (G9–G0) Write Blue Data (B9–B0) Write Red Data (R9–R0) R/W C1 C0 Palette Read 0 0 1 1 1 1 0 0 0 0 0 0 0 0 1 1 1 1 . . Write to Address Register (Lo-Byte) Write to Address Register (Hi-Byte) Read Red Data (R9–R0) Read Green Data (G9–G0) Read Blue Data (B9–B0) Read Red Data (R9–R0) R0 D0 Read Operation Palette Databus R9 D9 R8 D8 R7 D7 R6 D6 R5 D5 R4 D4 R3 D3 R2 D2 R1 D1 R0 D0 Figure 37. 10-Bit Databus Using 10-Bit DACs Write Operation Palette Databus R9 D9 R8 D8 R7 D7 0 Databus D6 R5 D5 R4 R3 R2 R1 R0 D4 D3 D2 D1 D0 R2 R1 R0 0 0 R/W C1 C0 Palette Write 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 . . Write to Address Register (Lo-Byte) Write to Address Register (Hi-Byte) Write Red Data (R9–R0) Write Green Data (G9–G0) Write Blue Data (B9–B0) Write Red Data (R9–R0) R/W C1 C0 Palette Read 0 0 1 1 1 1 0 0 0 0 0 0 0 0 1 1 1 1 . . Write to Address Registe (Lo-Byte) Write to Address Register (Hi-Byte) Read Red Data (R9–R0) Read Green Data (G9–G0) Read Blue Data (B9–B0) Read Red Data (R9–R0) Read Operation Palette 0 R6 R9 D9 R8 D8 R7 D7 R6 D6 R5 D5 R4 D4 R3 D3 D2 D1 D0 Figure 38. 10-Bit Databus Using 8-Bit DACS REV. 0 –25– ADV7160/ADV7162 REGISTER PROGRAMMING RAM-DAC Resolution Control (MR11) The following section describes each register, including Address Register, Mode Register and each of the Control Registers in terms of its configuration. When this is programmed with a “1,” the RAM is 30 bits deep (10 bits each for red, green and blue), and each of the three DACs is configured for 10-bit resolution. When MR11 is programmed with a “0,” the RAM is 24 bits deep (8 bits each for red, green and blue), and the DACs are configured for 8-bit resolution. The two LSBs of the 10-bit DACs are pulled down to zero in 8-bit RAM-DAC mode. Address Register (A10–A0) As illustrated in the previous tables, the C1 and C0 control inputs, in conjunction with this address register specify which control register, or color palette location is accessed by the MPU port. The Address Register is 11 bits wide and can be read from as well as written to. To access the Address Register, two consecutive MPU accesses with C1 and C0 set to Logic “0” are required. The first one accesses the low byte; and when a second access of the same type is performed, i.e., two consecutive reads or two consecutive writes, the high byte is accessed. If the type of access is changed, or if an access to a different register is inserted between the first and the second, then the second access will access the low byte again. When writing to or reading from the color palette on a sequential basis, only the start address needs to be written. After a red, green and blue write sequence, the address register is automatically incremented. Mode Register (MR1) The mode register is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register (MR19 and MR18 are both reserved). It is denoted as MR17–MR10 for simplification purposes. Figure 39 shows the various operations under the control of the mode register. This register can be read from as well written to. In read mode, if MR19 and MR18 are read back, they are both returned as zeros. MODE REGISTER BIT DESCRIPTION Reset Control (MR10) MR18 This bit determines the width of the MPU port. It is configured as either a 10-bit wide (D9–D0) or 8-bit wide (D7–D0) bus. Ten-bit data can be written to the device when configured 8-bit wide mode. The 8 MSBs are first written on D7–D0, then the two LSBs are written over D1–D0. Bits D9–D8 are zeros in 8bit mode. Operational Mode Control (MR14–MR13) When MR14 and MR13 are “0” the part operates in normal mode. Calibrate LOADIN (MR15) This bit automatically calibrates the on-board LOADIN/ LOADOUT synchronization circuit. A “0” to “1” transition initiates calibration. This bit is set to “0” in normal operation. See “Pipeline Delay & Calibration” section. This bit must run this cycle during the initialization sequence. Palette Select Match Bits Control (MR17–MR16) These bits allow multiple palette devices to work together. When bits PS1 and PS0 match MR17 and MR16 respectively, the device is selected. If these bits do not match, the device is not selected and the analog video outputs drive 0 mA. See “Palette Priority Select Inputs” section. CONTROL REGISTERS This bit is used to reset the pixel port sampling sequence. This ensures that the pixel sequence ABCD starts at A. It is reset by writing a “1” followed by a “0” followed by a “1.” This bit must run this cycle during the initialization sequence. MR19 MPU Data Bus Width (MR12) MR17 MR16 A large bank of registers plus the 64 × 64 cursor image can be accessed through the Control Register. Access is made first by writing the Address Register with the appropriate address to point to the particular Control Register (see Figure 34), and MR15 PCR4 MR14 MR13 PALETTE SELECT MATCH BITS CONTROL MR16 PS0 MR17 PS1 MR15 MR11 MR10 MPU DATA BUS WIDTH RESERVED* CALIBRATE LOADIN MR12 MR12 0 1 8-BIT (D7–D0) 10-BIT (D9–D0) RAM-DAC RESOLUTION CONTROL MR11 *THESE BITS ARE READ-ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00." OPERATIONAL MODE CONTROL 0 1 8-BIT 10-BIT MR14 MR13 0 0 1 1 0 1 0 1 NORMAL OPERATION RESERVED RESERVED RESERVED RESET CONTROL MR10 Figure 39. Mode Register 1 (MR1) (MR19–MR10) –26– REV. 0 ADV7160/ADV7162 then performing an MPU access to the Control Register. When accessing Control Registers in the range 200H to 204H, and when accessing the cursor image, the Address Register autoincrements after each register access. On accessing the last cursor image location at address 7FFH, the address register reverts to address 000H. The Address Register also auto-increments after a blue access, when accessing color registers in the address range 303H to 304H. ID Register (Address Reg (A10–A0) = 003H) This is an 8-bit wide “Identification” read-only register. For the ADV7160 it will always return the hexadecimal value 76H. For the ADV7162 it will always return the hexadecimal value 79H. Pixel Mask Register (Address Reg (A10–A0) = 004H) The contents of the pixel mask register are individually bit-wise logically ANDed with the Red, Green and Blue pixel input stream of data. It is an 8-bit read/write register with D0 corresponding to R0, G0 and B0. For normal operation, this register is set with FFH. COMMAND REGISTER 1 (CR1) (Address Reg (A10–A0) = 005H) This register contains a number of control bits as shown in the diagram. CR1 is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register (CR19 to CR18 reserved). Figure 40 shows the various operations under the control of CR1. This register can be read from as well written to. In write mode zero should be written to CR12. In read mode, CR19 and CR18 are returned as zeros. COMMAND REGISTER 1-BIT DESCRIPTION Calibration Control (CR10) This bit automatically calibrates the on-board LOADIN/ LOADOUT synchronization circuit on every vertical Sync. MR15 of Mode Register MR1 must be set to “0.” CR19 CR18 CR17 CR16 RESERVED* CR15 Hi-Byte Control (CR13) This bit enables access to the Hi Byte of the Address Register. When CR13 is set to Logic “0”, the part is compatible to the ADV7150. To access the hi-byte of the address register, this bit is set to Logic “1.” PS Function Control (CR15–CR14) These bits control the functions of the PS inputs. They are used to enable the Overlay Mode, Bypass Mode or the Palette Select Mode. In Palette Select Mode (CR15 and CR14 = “0”), these inputs are used to multiplex the RGB outputs of a number of devices. On a pixel by pixel basis, PS1 and PS0 are compared against the PS match bits, MR17 and MR16. If they match, then the part behaves normally. If they don’t match, then the analog output currents are switched to zero for that clock cycle, thus allowing another device, whose PS match bits match during this time, to drive the monitor. In Bypass Mode (CR15 = “0,” CR14 = “1”), PS1 is used to switch between one of the color modes through the Color Palette and one of the Palette Bypass Modes, on a pixel by pixel basis. The color mode through the palette is selected using CR17 and CR16. It is illegal to program CR27 to CR24 to select one of the bypass modes when using the PS bits to select a bypass mode at the pixel rate. This switching on a pixel by pixel basis is only allowed when using an ADV7160 device. Therefore, for the ADV7162, this mode (CR15 = “0,” CR14 = “1”), is reserved and should not be used. In Overlay Mode (CR15 = “1,” CR14 = “0”), the PS inputs provide control for a three color overlay. Whenever a value other than “00” is placed on the overlay inputs, the corresponding overlay color is displayed. When the overlay inputs contain “00,” the color is specified by the pixel inputs. When CR15 and CR14 = “1,” the PS inputs are completely ignored. There is no overlay, no bypass switching and the RGB outputs are enabled. Bypass Color Mode Control (CR17–CR16) These bits control the mode during bypass switching. There are three different modes: 24-bit Bypass, 16-bit Bypass or 15-bit Bypass Mode. CR14 CR13 PS FUNCTION CONTROL RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00." 0 0 1 1 0 1 0 1 PALETTE SELECTS BYPASS MODE** OVERLAYS IGNORE PS INPUTS **THIS MODE IS ONLY AVAILABLE ON THE ADV7160. IT IS RESERVED ON THE ADV7162. BYPASS COLOR MODE ADDRESS REGISTER HI-BYTE CONTROL CR17 CR16 0 1 0 1 CR11 CR12 (0) 15-BIT BYPASS 16-BIT BYPASS 24-BIT BYPASS RESERVED THIS BIT SHOULD BE SET TO ZERO TEST MODE CONTROL CALIBRATION CONTROL CR11 0 DISABLE 1 ENABLE TEST MODE CR13 0 NO ACCESS TO HI-BYTE (ADV7150 COMPATIBLE) 1 ACCESS TO HI-BYTE Figure 40. Command Register 1 (CR1) (CR19–CR10) REV. 0 CR10 CR15 CR14 *THESE BITS ARE READ-ONLY 0 0 1 1 CR12 –27– CR10 0 DISABLE 1 CALIBRATES ON EVERY VERTICAL SYNC (MR15=0) ADV7160/ADV7162 CR29 CR28 CR27 CR26 CR25 CR24 CR23 CR27 CR26 CR25 CR24 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 1 1 0 1 0 1 0 1 1 1 1 0 0 1 1 0 1 1 1 CR20 CR22 TRUE COLOR/PSEUDO COLOR MODE CONTROL 1 1 CR21 SYNC RECOGNITION CONTROL (GREEN) RESERVED* 1 1 CR22 0 1 COLOR MODE 8-BIT PSEUDO COLOR ON R7–R0 8-BIT PSEUDO COLOR ON G7–G0 8-BIT PSEUDO COLOR ON B7–0 16-BIT BYPASS MODE USING R7–R0, G7–G0 15-BIT BYPASS MODE USING R6–R0, G7–G0 16-BIT TRUE COLOR ON R7–R0, G7–G0 15-BIT TRUE COLOR ON R7–R3, G7–G3, B7–B3 15-BIT TRUE COLOR ON R6–R0, G7–G0 24-BIT TRUE COLOR 24-BIT BYPASS MODE 0 IGNORE 1 DECODE CR21–CR20 (00) PEDESTAL ENABLE CONTROL CR23 0 0 IRE 1 7.5 IRE ZERO SHOULD BE WRITTEN TO THESE BITS *THESE BITS ARE READ-ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00." Figure 41. Command Register 2 (CR2) (CR29–CR20) COMMAND REGISTER 2 (CR2) (Address Reg (A10–A0) = 006H) COMMAND REGISTER 3 (CR3) (Address Reg (A10–A0) = 007H) This register contains a number of control bits as shown in the diagram. CR2 is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register (CR29 and CR28 are both reserved). This register contains a number of control bits as shown in the diagram. CR3 is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register (CR39 and CR38 are both reserved). Figure 41 shows the various operations under the control of CR2. This register can be read from as well written to. In write mode zero should be written to CR21 and CR20. In read mode, CR29 and CR28 are returned as zeros. Figure 42 shows the various operations under the control of CR3. This register can be read from as well written to. In write mode zero should be written to CR35. In read mode, CR39 and CR38 are returned as zeros. COMMAND REGISTER 2-BIT DESCRIPTION SYNC Recognition Control on Green (CR22) COMMAND REGISTER 3 BIT DESCRIPTION PRGCKOUT Frequency Control (CR31–CR30) This bit specifies whether the video SYNC Input is to be encoded onto the IOG analog output or ignored. These bits specify the output frequency of the PRGCKOUT output. PRGCKOUT is a divided down version of the pixel CLOCK. Pedestal Enable Control (CR23) BLANK Pipeline Delay Control (CR34–CR32) This bit specifies whether a 0 IRE or a 7.5 IRE blanking pedestal is to be generated on the video outputs. True-Color/Bypass/Pseudo-Color Mode Control (CR27–CR24) These 4 bits specify the various color modes. These include a 24-bit true-color and bypass mode, one 16-bit true-color and bypass mode, two 15-bit true-color modes, one 15-bit bypass mode and three 8-bit pseudo color modes. These bits specify the additional pipeline delay that can be added to the BLANK function, relative to the overall device pipeline delay (tPD). As the BLANK control normally enters the Video DAC from a shorter pipeline than the video pixel data, this control is useful in de-skewing the pipeline differential. Pixel Multiplex Control (CR37–CR36) These bits specify the device’s multiplex mode. It therefore also determines the frequency of the LOADOUT signal. LOADOUT is a divided down version of the pixel CLOCK. –28– REV. 0 ADV7160/ADV7162 CR39 CR38 CR37 CR36 CR35 CR34 CR33 CR32 CR35 (0) *THESE BITS ARE READ-ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00." CR34 CR33 CR32 0 0 0 ZERO SHOULD BE WRITTEN TO THIS BIT 0 0 1 ......... ......... 1 1 BLANK PIPELINE DELAY 0 1 0 tPD tPD + 1 x LOADOUT tPD + 2 x LOADOUT 1 tPD + 7 x LOADOUT PRGCKOUT FREQUENCY CONTROL Pixel Multiplex Control CR37 CR36 0 1 0 1 CR36 EXTRA BLANK PIPELINE DELAY CONTROL (ADDS TO PIXEL PIPELINE DELAY; tPD) RESERVED* 0 0 1 1 CR37 CR31 CR30 1:1 MUXING: LOADOUT = CLOCK ÷ 1 2:1 MUXING: LOADOUT = CLOCK ÷ 2 8:1 MUXING: LOADOUT = CLOCK ÷ 8 (PSEUDO COLOR ONLY) 4:1 MUXING: LOADOUT = CLOCK ÷ 4 0 0 1 1 CLOCK ÷ 4 CLOCK ÷ 8 CLOCK ÷ 16 CLOCK ÷ 32 0 1 0 1 Figure 42. Command Register 3 (CR3) (CR39–CR30) SYNC Recognition Control on Red (CR41) COMMAND REGISTER 4 (CR4) (ADDRESS REG (A10–A0) = 008H) This bit specifies whether the video SYNC Input is to be encoded onto the IOR analog output or ignored. This register contains a number of control bits as shown in the diagram. CR4 is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register (CR49 and CR48 are both reserved). SYNC Recognition Control on Blue (CR42) This bit specifies whether the video SYNC Input is to be encoded onto the IOB analog output or ignored. Figure 43 shows the various operations under the control of CR4. This register can be read from as well written to. In read mode, CR49 and CR48 are both returned as zeros. Gain Control (CR44–CR43) These bits specifies the amount of gain on the DAC depending on the standard required. See “DAC and Video Outputs” section for more detail. For gain settings that have no pedestal, the pedestal is automatically disabled independently of CR23. COMMAND REGISTER 4-BIT DESCRIPTION HDTV SYNC Enable (CR40) Signature Clock Control (CR45) This bit specifies whether the video TRISYNC Input is to be encoded, enabling the DAC outputs to generate a Tri-Level Sync. CR49 CR48 CR47 CR46 This bit enables or disables the clock to the signature analyzer. CR45 CR44 CR43 CR42 CR41 CR40 Reserved* RESERVED* DAC GAIN *THESE BITS ARE READ-ONLY SIGNATURE RESET RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00." 0 0 1 1 CR46 0 1 ENABLE DISABLE 0 DISABLE 1 ENABLE 0 1 0 1 CR41 3996 4224 4311 5592 0 IGNORE 1 DECODE SYNC RECOGNITION CONTROL (BLUE) SIGNATURE ACQUIRE CR47 SYNC RECOGNITION CONTROL (RED) CR44 CR43 SIGNATURE CLOCK CONTROL CR45 CR42 0 IGNORE 1 DECODE CR40 0 DISABLE CLOCK 0 DISABLE TRI-SYNC 1 ENABLE CLOCK 1 ENABLE TRI-SYNC Figure 43. Command Register 4 (CR4) (CRF49–CR40) REV. 0 HDTV SYNC CONTROL –29– ADV7160/ADV7162 Signature Reset Control (CR46) Output Divide Control (PCR5–PCR4) Taking CR46 low then high resets the signature analyzer. This is done to give a known starting point before acquiring a signature. These bits control the PLL output divider. This post-scaler is used in the generation of lower frequencies. Signature Acquire Control (CR47) PLL S Control (PCR7–PCR6) This bit should be set to Logic “1” for normal operation. See “Test Diagnostic” section for more information. These bits set up the S value in the PLL transfer function. This extra value provides extra control in setting the feedback divider value of the PLL. COMMAND REGISTER 5 (CR5) (Address Reg (A10–A0) = 00DH) Status Register (Address Reg (A10–A0) = 00AH) This register contains one control bit CR56. CR5 is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register (CR59 and CR58 are both reserved). This register is a read only 10-bit register. However SR9– SR8 are reserved bits, containing zeros and SR7–SR1 are undefined bits and should be masked in software on read back. Therefore, SR0 is the only relevant Bit in the Status Register and contains a Logic “1” if one, or more of the IOR, IOG, and IOB outputs exceed the internal voltage of the SENSE comparator circuit . It can be used to determine the presence of a CRT monitor. With some diagnostic code, the presence of loading on the individual RGB lines can be determined. The reference is generated by a voltage divider from the external voltage reference on the VREF pin. For the proper operation, the following levels should be applied to the comparator by the IOR, IOG and IOB outputs: This register can be read from as well written to. Control Bit CR56 selects either external clock or internal PLL operation. If the internal PLL is to be used, Logic “1” should be written to CR56.This should be set up immediately after power up. In write mode, zero should be written to CR57 and CR55–CR50. In read mode, CR59 and CR58 are both returned as zeros. PLL COMMAND REGISTER (PCR) (Address Reg (A10–A0) = 009H) This register contains a number of control bits as shown in the diagram. PCR is a 10-bit wide register. However, for programming purposes, it may be considered as an 8-bit wide register (PCR9 and PCR8 are both reserved). DAC Low Voltage ≤ 250 mV DAC High Voltage ≥ 450 mV Figure 44 shows the various operations under the control of PCR. This register can be read from as well written to. In write mode zero should be written to PCR3. In read mode PCR9 and PCR8 are returned as zeros. Revision Register (Address Reg (A10–A0) = 01BH) PLL Control (PCR0) PLL R Register (Address Reg (A10–A0) = 00CH) This register is a read only register containing the revision of silicon. This bit enables or disables PLL. This register is a read only 10-bit register. However, R9–R8 are reserved bits, containing zeros. Bit R7 is a read only bit. This bit should be masked in software on readback as its value may be indeterminate. Therefore, the PLL R Register may be treated as a 7-bit wide register. This register, together with the RSEL Bit in the PLL Control Register, controls the reference divider of the on-board PLL. RSEL Bit Control (PCR1) This bit enables or disables RSEL, which together with the contents of the PLL R Register affect the reference divider value of the PLL. Reference Divider = (1 + RSEL) × (R+2). VSEL Bit Control (PCR2) This bit enables or disables VSEL, which together with the contents of the PLL V Register and the PLL S value affect the feedback divider value of the PLL. Feedback Divider = (1 + VSEL) × (4(V + 2)+S). PCR9 PCR8 PCR7 PCR6 PCR5 PCR4 PCR3 PCR2 PCR1 PCR0 S VALUE RESERVED* *THESE BITS ARE READ-ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00." PCR5 (0) PCR7 PCR6 (S1 S0) 0 0 1 1 0 1 0 1 RSEL ENABLE ZERO SHOULD BE WRITTEN TO THIS BIT 0 1 2 3 FOUT 0 1 0 1 0 DISABLE 1 ENABLE VSEL ENABLE PCR5 PCR4 (PSEL1 PSEL0) 0 0 1 1 PCR1 PCR2 VCO/1 VCO/2 VCO/4 VCO/8 PLL CONTROL PCR0 0 DISABLE 1 ENABLE 0 DISABLE PLL 1 ENABLE PLL Figure 44. Command Register (PCR) (PCR9–PCR0) –30– REV. 0 ADV7160/ADV7162 PLL V Register (Address Reg (A10–A0) = 00FH) This register is a read only 10-bit register. However V9–V8 are reserved bits, containing zeros. Bit V7 is a read only bit. This bit should be masked in software on readback as its value may be indeterminate. Therefore, the PLL V Register may be treated as a 7-bit wide register. This register, together with the VSEL Bit in the PLL Control Register, controls the feedback divider of the on-board PLL. bits are stored at each address. With two bits per cursor pixel, four horizontally adjacent pixels are stored at each address. As each address location in the Cursor Image is filled, the progression is from left to right until a line is filled and top to bottom until all the lines are filled. The cursor can be displayed on both an interlaced and noninterlaced system, as controlled by CCR3 of the Cursor Control Register. On an interlaced system, only one cursor can be displayed per field. The ODD/EVEN input indicates which field of the frame is being displayed. 64 × 64 Cursor Cursor Y Coordinate Even The ADV7160/ADV7162 has a 64 × 64 cursor generator on board. Several of the control registers control the cursor. These will be described in detail. The Cursor-X and Cursor-Y registers specify the position the cursor is to be placed on the screen. The origin (0, 0) of the cursor is top left. The position of the cursor is taken relative to this point, allowing the CursorX and Cursor-Y registers to be programmed with negative numbers and thus allow the cursor to be partially or completely off the screen. The cursor can work as an X-11 or XGA cursor, controlled by Bits CCR0 and CCR1 of the Cursor Control Register. The screen X and Y coordinates are measured from the rising edge of BLANK. The first pixel after the rising edge of BLANK corresponds to the origin (0, 0). The Vertical retrace time is extracted from the composite SYNC and BLANK inputs. The start of Vertical Retrace is recognized by counting a second rising edge on SYNC while BLANK remains low. The next rising edge on BLANK is the start of line 0. Cursor X-Lo and Cursor X-Hi Register (Address Reg (A10–A0) = 200H and 201H) The Even field starts with line 0 of the cursor image on line Y of the frame. Subsequent even lines of the cursor image are displayed on subsequent lines of the Even field. On the Even field, the frame line counter starts at 0 and increments by 2 at the end of every Even field line. The Odd field starts with line 1 of the cursor image on line Y + 1 of the frame. Subsequent odd lines of the cursor image are displayed on subsequent lines of the Odd field. On the Odd field, the frame line counter starts at 1 and increments by 2 at the end of every Odd field line. Cursor Y Coordinate Odd The Even field starts with line 1 of the cursor image on line Y + 1 of the frame. Subsequent even lines of the cursor image are displayed on subsequent lines of the Even field. On the Even field, the frame line counter starts at 1 and increments by 2 at the end of every Even field line. The Odd field starts with line 0 of the cursor image on line Y of the frame. Subsequent odd lines of the cursor image are displayed on subsequent lines of the Odd field. On the Odd field, the frame line counter starts at 0 and increments by 2 at the end of every Odd field line. Cursor Control Register (Address Reg (A10–A0) = 204H) These 8-bit registers together form a 16-bit 2s complement representation of the cursor x-coordinate on the screen. The valid range for the cursor x-coordinate is ± FFFH. The negative number representation allows for part or all of the cursor to be displayed off the left-hand edge of the screen This register contains a number of control bits. CCR is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register (CCR8 and CCR9 are both reserved). In write mode zero should be written to CCR4 to CCR7. In read mode, CCR8 and CCR9 are all returned as zeros. Cursor Y-Lo and Cursor Y-Hi Register (Address Reg (A10–A0) = 202H and 203H) These 8-bit registers together form a 16-bit 2s complement representation of the cursor x-coordinate on the screen. The valid range for the cursor x-coordinate is ± FFFH. The negative number representation allows for part or all of the cursor to be displayed off the top/left of the screen. Figure 45 shows the various operations under the control of CCR. When accessing the cursor X and Y registers, the Address Register auto-increments after each access. There are no restrictions on updating the cursor coordinate registers other than they must all be written in the order X-Low, X-Hi, Y-Low, Y-Hi to update the coordinates. Only one cursor is displayed per frame, at the last X and Y coordinates written. Access to these registers is independent of the databus being configured for 8- or 10-bit operation. These bits specify which type of cursor is being used. Each cursor pixel value controls the color differently in each mode. CURSOR CONTROL REGISTER BIT DESCRIPTION CURSOR MODE CONTROL (CCR1–CCR0) Cursor Color 1 and Cursor Color 2 Register (Address Reg (A10–A0) = 304H and 303H) Each of these color registers are 30 bits wide, made up of 10 bits for Red, 10 bits for Green and 10 bits for Blue. Access to these registers behaves in the same way as access to the Color Palette with respect to the different combinations of 10/8-bit databus and 10/8-bit DAC resolution. Cursor Image (Address Reg (A10–A0) = 400H–7FFH) Table II. Bit 1 Bit 0 X-11 Cursor XGA Cursor 0 0 1 1 0 1 0 1 Transparent Transparent Color 1 Color 2 Color 1 Color 2 Transparent Bit-Wise Complement Cursor Enable Control (CCR2) This bit turns the cursor on and off. Interlace Control (CCR3) This bit determines whether the cursor is being used in interlaced or noninterlaced mode. This region contains the 64 × 64 × 2-bit Cursor Image. Eight REV. 0 –31– ADV7160/ADV7162 CCR9 CCR8 CCR7 CCR6 CCR5 CCR4 CCR3 CCR2 CCR1 CCR0 CURSOR ENABLE RESERVED* *THESE BITS ARE READ-ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00." CCR2 CCR7–CCR4 (0000) 0 DISABLE 1 ENABLE ZERO SHOULD BE WRITTEN TO THESE BITS CURSOR MODE Control CURSOR CONTROL CCR1 CCR0 CCR3 0 NONINTERLACED 1 INTERLACED 0 0 1 1 0 1 0 1 RESERVED X11 CURSOR XGA CURSOR RESERVED Figure 45. Cursor Control Register (CCR) (CCR9–CCR0) DIGITAL-TO-ANALOG CONVERTER (DACS) AND VIDEO OUTPUTS IOR, IOG, IOB ZO = 75Ω ZS = 75Ω (SOURCE TERMINATION) (CABLE) DACs The ADV7160/ADV7162 contains three high speed video DACs. The DAC outputs are represented as the three primary analog color signals IOR (red video), IOG (green video) and IOB (blue video). DACs and Analog Outputs The part contains three matched 10-bit digital-to-analog converters. The DACs are designed using an advanced, high speed, segmented architecture. The bit currents corresponding to each digital input are routed to either IOR, IOG, IOB (bit = “1”) or GND. The analog video outputs are high impedance current sources. Each of the these three RGB current outputs are specified to directly drive a 37.5 Ω load (doubly terminated 75 Ω). ZL = 75Ω (MONITOR) Figure 46. DAC Output Termination (Doubly Terminated 75 Ω Load) OUTPUT WITHOUT OUTPUT WITH SYNC ENCODED SYNC ENCODED mA V mA V 19.05 0.714 26.67 1.000 A resistor RSET is connected between the RSET input of the part and ground. For specified performance, RSET has a value of 280 Ω. This corresponds to the generation of RS-343A video levels (with SYNC on IOG and Pedestal = 7.5 IRE) into a doubly terminated 75 Ω load. In this example DAC Gain has a value of 3996 and is set using CR43 and CR44 of Command Register 4. Figure 47 illustrates the resulting video waveform and the Video Output Truth Table illustrates the corresponding control input stimuli. On the ADV7160/ADV7162 SYNC can be encoded on any of the analog signals, however in practice, SYNC is generally encoded on either the IOG output or on all of the video outputs. –32– Y 92.5 IRE G RE An external 1.23 V voltage reference is required to set up the analog outputs of the ADV7160/ADV7162. The reference voltage is connected to the VREF input. SC A LE Reference Input and R SET WHITE LEVEL 1.44 0.054 9.05 BLACK LEVEL 0.340 7.5 IRE 0 0 7.62 BLANK LEVEL 0.286 40 IRE 0 0 SYNC LEVEL Figure 47. Composite Video Waveform SYNC Decoded; Pedestal = 7.5 IRE; DAC Gain = 3996 REV. 0 ADV7160/ADV7162 Table III. Video Output Truth Table O/P with Sync Enabled (mA) O/P with Sync Disabled (mA) SYNC BLANK Description DAC Input Data WHITE LEVEL VIDEO VIDEO to BLANK BLACK LEVEL BLACK to BLANK BLANK LEVEL SYNC LEVEL 26.67 Video + 9.05 Video + 1.44 9.05 1.44 7.62 0 19.05 Video + 1.44 Video + 1.44 1.44 1.44 0 0 1 1 0 1 0 1 0 1 1 1 1 1 0 0 3FFH Data Data 000H 000H xxxH xxxH Variations on RS-343A OUTPUT WITHOUT SYNC ENCODED mA V 19.05 0.714 WHITE LEVEL Video Signal 4224 4311 5592 RS343A, SYNC decoded on output; Pedestal = 0 IRE RS343A, No SYNC decoded; Pedestal = 0 IRE RS170, SYNC decoded; Pedestal = 7.5 IRE 100 IRE 0 GR Gain EY SC AL E Various other video output configurations can be implemented by the ADV7160/ADV7162, including RS-170. The table shows calculated values of DAC Gain for some of the most common variants on the RS-343A standard. The associated waveforms are shown in the diagrams. BLANK/BLACK LEVEL 0 OUTPUT WITHOUT OUTPUT WITH SYNC ENCODED SYNC ENCODED mA V mA V 18.62 0.698 26.67 1.000 Figure 50. Composite Video Waveform Pedestal = 0 IRE; DAC Gain = 4311 WHITE LEVEL Output Currents EY GR 100 IRE SC AL E The various output currents are set by VREF, RSET and the DAC gain. By programming the Command Register Bits and choosing the correct DAC Gain value, video waveforms conforming to the common variations on RS-170 and RS-343A, as well as HDTV standards may be generated. The currents generated can be summarized as: 0 0 8.05 BLANK/ BLACK LEVEL 0.302 IOUT = IDAC + IBLANK + ISYNC + ITRISYNC 43 IRE 0 SYNC LEVEL 0 IDAC (mA) = I BLANK (mA) = 0.0817 × IDAC Figure 48. Composite Video Waveform SYNC Decoded; Pedestal = 0 IRE; DAC Gain = 4224 I SYNC (mA) = 0.4322 × IDAC I TRISYNC (mA) = 0.4322 × IDAC OUTPUT WITH OUTPUT WITHOUT SYNC ENCODED SYNC ENCODED V 26.67 1.00 mA V WHITE LEVEL 37.33 1.400 R G 92.5 IRE EY SC A LE mA TRISYNC LEVEL 21.24 0.800 2.00 0.075 12.67 0.475 7.5 IRE 0 0 10.67 BLACK LEVEL BLANK LEVEL 0.400 40 IRE 0 0 SYNC LEVEL Figure 49. Composite Video Waveform SYNC and TRISYNC decoded; Pedestal = 7.5 IRE; DAC Gain = 5592 REV. 0 DAC GAIN × V REF RSET –33– ADV7160/ADV7162 APPENDIX 1 BOARD DESIGN AND LAYOUT CONSIDERATIONS The ADV7160/ADV7162 is a highly integrated circuit containing both precision analog and high speed digital circuitry. It has been designed to minimize interference effects on the integrity of the analog circuitry by the high speed digital circuitry. It is imperative that these same design and layout techniques be applied to the system level design such that high speed, accurate performance is achieved. The “Recommended Analog Circuit Layout” shows the analog interface between the device and monitor. The layout should be optimized for lowest noise on the ADV7160/ADV7162 power and ground lines by shielding the digital inputs and providing good decoupling. The lead length between groups of VAA and GND pins should by minimized so as to minimize inductive ringing. Ground Planes The ground plane should encompass all ADV7160/ADV7162 ground pins, voltage reference circuitry, power supply bypass circuitry for the ADV7160/ADV7162, the analog output traces, and all the digital signal traces leading up to the ADV7160/ ADV7162. The ground plane is the graphics board's common ground plane. Power Planes The ADV7160/ADV7162 and any associated analog circuitry should have it’s own power plane, referred to as the analog power plane (VAA). This power plane should be connected to the regular PCB power plane (VCC) at a single point through a ferrite bead. This bead should be located within three inches of the ADV7160/ADV7162. The PCB power plane should provide power to all digital logic on the PC board, and the analog power plane should provide power to all ADV7160/ADV7162 power pins and voltage reference circuitry. Plane-to-plane noise coupling can be reduced by ensuring that portions of the regular PCB power and ground planes do not overlay portions of the analog power plane, unless they can be arranged such that the plane-to-plane noise is common mode. Supply Decoupling operation, to reduce the lead inductance. Best performance is obtained with 0.1 µF ceramic capacitor decoupling. Each group of VAA pins on the ADV7160/ADV7162 must have at least one 0.1 µF decoupling capacitor to GND. These capacitors should be placed as close as possible to the device. It is important to note that while the ADV7160/ADV7162 contains circuitry to reject power supply noise, this rejection decreases with frequency. If a high frequency switching power supply is used, the designer should pay close attention to reducing power supply noise and consider using a three terminal voltage regulator for supplying power to the analog power plane. Digital Signal Interconnect The digital inputs to the ADV7160/ADV7162 should be isolated as much as possible from the analog outputs and other analog circuitry. Also, these input signals should not overlay the analog power plane. Due to the high clock rates involved, long clock lines to the ADV7160/ADV7162 should be avoided to reduce noise pickup. Any active termination resistors for the digital inputs should be connected to the regular PCB power plane (VCC), and not the analog power plane. Analog Signal Interconnect The ADV7160/ADV7162 should be located as close as possible to the output connectors to minimize noise pickup and reflections due to impedance mismatch. The video output signals should overlay the ground plane, and not the analog power plane, to maximize the high frequency power supply rejection. Digital Inputs, especially Pixel Data Inputs and clocking signals (CLOCK, LOADOUT, LOADIN, etc.) should never overlay any of the analog signal circuitry and should be kept as far away as possible. For best performance, the analog outputs (IOR, IOG, IOB) should each have a 75 Ω load resistor connected to GND. These resistors should be placed as close as possible to the ADV7160/ADV7162 so as to minimize reflections. For optimum performance, bypass capacitors should be installed using the shortest leads possible, consistent with reliable –34– REV. 0 ADV7160/ADV7162 POWER SUPPLY DECOUPLING (0.1µF CAPACITOR FOR EACH VAA GROUP) 0.1µF 0.01µF 0.1µF 0.01µF 0.1µF 0.01µF +5V (VAA) ANALOG POWER PLANE +5V (VAA) L1 (FERRITE BEAD) 0.01µF +5V (VCC) 0.1µF 33µF +5V (VAA) VAA 0.1µF COMP VREF RSET 1kΩ (1% METAL) RSET 0.1µF AD589 (1.2V REF) MONITOR (CRT) COAXIAL CABLE (75Ω) 280Ω ADV7160 IOR 75Ω IOG 75Ω IOB 75Ω 75Ω 75Ω 75Ω BNC CONNECTORS GND NOTES: 1. ALL RESISTERS ARE 1% METAL FILM 2. 0.1µF AND 0.01µF CAPACITORS ARE CERAMIC 3. ADDITIONAL DIGITAL CIRCUITRY OMITTED FOR CLARITY Recommended Analog Circuit Layout REV. 0 0.1µF –35– ADV7160/ADV7162 APPENDIX 2 TYPICAL FRAME BUFFER INTERFACE PLLREF PLL CLOCK ECL TO TTL CLOCK PRGCKOUT S E L E C T DIVIDE BY M (÷M) DIVIDE BY N (÷N) LOADOUT CLOCK GRAPHICS PROCESSOR/ CONTROLLER SCKOUT LATCH BLANK BLANK SYNC / TRISYNC ENABLE SYNC / TRISYNC SCKIN VRAM (BANK A) VRAM (BANK B) ADV7160/ ADV7162 LOADIN FRAME BUFFER/ VIDEO MEMORY 24 24 24 24 24 24 24 24 50 MHZ 50 MHZ VRAM (BANK C) 50 MHZ VRAM (BANK D) 50 MHZ 24 –36– MULTIPLEXER TO PALETTE/RAM & DAC REV. 0 ADV7160/ADV7162 APPENDIX 3 10-BIT DACs AND GAMMA CORRECTION GAMMA CORRECTION 8 Bits vs. 10 Bits 10-Bit DACs 10-Bit RAM-DAC resolution allows for nonlinear video correction, in particular Gamma Correction. The ADV7160/ADV7162 allows for an increase in color resolution from 24-bit to 30-bit effective color without the necessity of a 30-bit deep frame buffer. In true-color mode, for example, the part effectively operates as a 24-bit to 30-bit color look-up table. Up to now we have assumed that there exists a linear relationship between the actual RGB values input to a monitor and the intensity produced on the screen. This, however, is not the case. Half scale digital input (1000 0000) might correspond to only 20% output intensity on the CRT (Cathode Ray Tube). The intensity (ICRT) produced on a CRT by an input value IIN is given by: ICRT = (IIN)c where c ranges from 2.0 to 2.8. If the individual values of c for red, green and blue are known, then so called “Gamma Correction” can be applied to each of the three video input signals (IIN); therefore: IIN(corrected) = k(IIN)1/c Traditionally, there has been a trade-off between implementing a nonlinear graphics function, such as gamma correction, and color dynamic range. The ADV7160/ADV7162 overcomes this by increasing the individual color resolution of each of the red, green and blue primary colors from 8 bits per color channel to 10 bits per channel (24 bits to 30 bits). 8-Bit Data Gamma Corrected (2.7) Quantized to 8 Bits Quantized to 10 Bits 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 0.977797 0.979304 0.980807 0.982306 0.983801 0.985292 0.986780 0.988264 0.989744 0.991220 0.992693 0.994161 0.995626 0.997088 0.998546 1.000000 250 250 251 251 251 252 252 252 253 253 254 254 254 255 255 255 1001 1002 1004 1005 1007 1008 1010 1011 1013 1015 1016 1018 1019 1021 1022 1023 1.00 The table highlights the loss of resolution when 8-bit data is gamma-corrected to a value of 2.7 and quantized in a traditional 8-bit system. Note that there is no change in the 8-bit quantized data for linear changes in the input data over much of the transfer function. On the other hand, when quantized to 10 bits via the 10-bit RAMs and 10-bit DACs of the ADV7160/ ADV7162, all changes on the input 8-bit data are reflected in corresponding changes in the 10-bit data. VE UR 0.80 NC E TIO C 0.70 E RR MA 0.60 CO D BY E TH VE M GA E CI 0.50 E RE S ON 0.40 SP AR 0.30 E NS RE O P ES NE LI 0.20 EY T R CR 0.10 0.00 The graph shows a typical gamma curve corresponding to a gamma value of 2.7. This is programmed to the red, green and blue RAMs of the color look-up table instead of the more traditional linear function. Different curves corresponding to any particular gamma value can be independently programmed to each of the red, green and blue RAMs. 0 32 64 96 128 160 192 INPUT CODE – DECIMAL 224 256 Gamma Correction Curve (Gamma Value = 2.7) Other applications of the 10-bit RAM-DAC include closed-loop monitor color calibration. REV. 0 DAC OUTPUT – NORMALISED TO 1) 0.90 –37– ADV7160/ADV7162 APPENDIX 4 INITIALIZATION AND PROGRAMMING ADV7160/ADV7162 INITIALIZATION After power has been supplied, the ADV7160/ADV7162 must be initialized. The Mode Register and Control Registers must then be set up. The values written to the various registers will be determined by the desired operating mode of the part, i.e., True-Color/ Pseudo-Color, 4:1 Muxing/2:1 Muxing, PLL on/off, Bypass Mode on/off etc. . . . The following section gives a recommended initialization of the ADV7160/ADV7162 and an example of the ADV7162 operating in a specific mode. ADV7160/ADV7162 Initialization C1 C0 R/W Comment Write (xx000xx1)* to Mode Register (MR1) Write (xx000xx0)* to Mode Register (MR1) Write (xx000xx1)* to Mode Register (MR1) 1 1 1 1 1 1 0 0 0 Resets ADV7160/62 Write 05H to Address Register (A7–A0) Write (xxxx100x)* to Command Register 1 (CR1) 0 0 0 0 0 0 Address Reg points to Command Register 1 (CR1) Address Reg points to CR1 for high byte access Write 06H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write (xxxxxx00)* to Command Reg 2 (CR2) 0 0 1 0 0 0 0 0 0 Address Reg points to Command Register 2 (CR2) Setup CR2 as required Write 07H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write (xx0xxxxx)* to Command Reg 3 (CR3) 0 0 1 0 0 0 0 0 0 Address Reg points to Command Register 3 (CR3) Setup CR3 as required Write 08H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write (xxxxxxxx)* to Command Reg 4 (CR4) 0 0 1 0 0 0 0 0 0 Address Reg points to Command Register 4 (CR4) Setup CR4 as required Write 0DH to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write (0x000000)* to Command Reg 5 (CR5) 0 0 1 0 0 0 0 0 0 Address Reg points to Command Register 5 (CR5) Setup CR 5 as required Write 04H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write (xxxxxxxx)* to Pixel Mask Register 0 0 1 0 0 0 0 0 0 Address Reg points to Pixel Mask Register Set up Pixel Mask as required Write 04H to Address Register (A7–A0) Write 02H to Address Register (A10–A8) Write (xxxxxxxx)* to Cursor Control Register 0 0 1 0 0 0 0 0 0 Necessary only if CCR to be used Address Reg points to Cursor Control Register (CCR) Set up CCR as required Write 0FH to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write (xxxxxxxx)* to PLL V Register 0 0 1 0 0 0 0 0 0 Necessary only if PLL to be used Address Reg points to PLL V Register Set up V as required Write 0CH to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write (xxxx0xxx)* to PLL R Register 0 0 1 0 0 0 0 0 0 Necessary only if PLL to be used Address Reg points to PLL R Register Set up R as required Write 09H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write (xxxxxxxx)* to PLLCommand Register 0 0 1 0 0 0 0 0 0 Necessary only if PLL to be used Address Reg points to PLL Command Register (PCR) Set up PCR as required Write (xx0xxxxx)* to Mode Register (MR1) Write (xx1xxxxx)* to Mode Register (MR1) Write (xx0xxxxx)* to Mode Register (MR1) 1 1 1 1 1 1 0 0 0 Necessary only if manual claibration is required Toggles MR15 *x represents either a 0 or 1 value that the bit should be set to, depending on the desired operating mode of the ADV7160/ADV7162. –38– REV. 0 ADV7160/ADV7162 Example Color Mode: 24-Bit Gamma Corrected True Color (30-Bits) through Color Palette Multiplexing: 2:1, Databus: 10-Bit, RAM-DAC Resolution: 10-Bit, SYNC: on Green, Pedestal: 0 IRE, Calibration: Every Vertical Sync, Internal PLL: 220 MHz (Reference = 15 MHz) Register Initialization C1 C0 R/W Comment Write 07H to Mode Register (MR1) Write 06H to Mode Register (MR1) Write 07H to Mode Register (MR1) 1 1 1 1 1 1 0 0 0 Resets ADV7162* 10-Bit Data Bus, 10-Bit DAC Resolution Write 05H to Address Register (A7–A0) Write 09H to Command Register 1 (CR1) 0 0 0 0 0 0 Address Reg points to Command Register 1 (CR1) High byte access, Calibrate every Vertical Sync Write 06H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write E4H to Command Reg 2 (CR2) 0 0 1 0 0 0 0 0 0 Address Reg points to Command Register 2 (CR2) 24-Bit True Color, 0 IRE, Sync on Green Write 07H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write 40H to Command Reg 3 (CR3) 0 0 1 0 0 0 0 0 0 Address Reg points to Command Register 3 (CR3) 2:1 Muxing, PRGCKOUT = CLOCK ÷ 4 Write 08H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write 00H to Command Reg 4 (CR4) 0 0 1 0 0 0 0 0 0 Address Reg points to Command Register 4 (CR4) DAC GAIN = 3996 Write 0DH to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write 40H to Command Reg 5 (CR5) 0 0 1 0 0 0 0 0 0 Address Reg points to Command Register 5 (CR5) Internal PLL to be used Write 04H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write FFH to Pixel Mask Register 0 0 1 0 0 0 0 0 0 Address Reg points to Pixel Mask Register Set up Pixel Mask Write 0FH to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write 09H to PLL V Register 0 0 1 0 0 0 0 0 0 Address Reg points to PLL V Register Set up V value Write 0CH to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write 01H to PLL R Register 0 0 1 0 0 0 0 0 0 Address Reg points to PLL R Register Set up R value Write 09H to Address Register (A7–A0) Write 00H to Address Register (A10–A8) Write 06H to PLL Command Register 0 0 1 0 0 0 0 0 0 Address Reg points to PLL Command Register (PCR) Set up PCR as required Color Palette RAM Initialization C1 C0 R/W Comment Write Write Write Write Write Write Write Write 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0 0 0 0 0 Points to Color Palette RAM (Initializes Palette RAM (to a Linear Ramp** ( ( ( ( . . . ( Write FFH (red data) to RAM location (FFH) Write FFH (green data) to RAM location (FFH) . 0 0 . 1 1 . 0 0 ( ( ( Write FFH (blue data) to RAM location (FFH) 0 1 0 ( RAM Initialization Complete 00H to Address Register (A7–A0) 00H to Address Register (A10–A8) 00H (red data) to RAM location (00H) 00H (green data) to RAM location (00H) 00H (blue data) to RAM location (00H) 01H (red data) to RAM location (01H) 01H (green data) to RAM location (01H) 01H (blue data) to RAM location (01H) **These command lines reset the ADV7162. The pipelines for each of the Red, Green & Blue pixel inputs are synchronously reset to the Multiplexer's “A” input. Mode Register bit MR10 is written by a “1” followed by “0” followed by “1.” **This sequence of instructions would, of course, normally be coded using some form of loop instruction. REV. 0 –39– ADV7160/ADV7162 APPENDIX 5 SIGNATURE ANALYZER SIGNATURE REGISTER I/P S19 B0 B1 B2 B3 B4 B5 B6 B7 B8 B9 G0 G1 G2 G3 G4 G5 G6 G7 G8 G9 R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 '0' '0' SIGNATURE CELL S0 S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28 S29 S30 S31 S32 S32 Signature Register The ADV7160/ADV7162 contains onboard circuitry that enables both device and system level test diagnostics. The ADV7160/ ADV7162 has a signature analyzer in the pixel datapath, just before the DAC decoders. The signature analyzer consists of a 33-bit linear feedback shift register. The 30-bit pixel value is fed as a parallel input into the analyzer. The signature analyzer only accumulates a signature during active display time when BLANK is high. Bit CR45 to CR47 of Command Register 4 control the signature analyzer. When CR45 of Command Register 4 is set to Logic “1,” the clock to the signature analyzer is enabled. Toggling CR46 low and then high resets the signature analyzer. This is done to give a known starting point before acquiring a signature. CR47 of Command Register 4 controls the feedback inputs to the analyzer. When CR47 of Command Register 4 is a Logic “0,” the feedback is disabled and on each clock cycle, the 30-bit pixel value is latched directly into the analyzer. To acquire a signature as the analyzer is clocked, CR47 of Command Register 4 is set to Logic “1.” To acquire a signature the following procedure must be followed: CR42 S19 S0 CR42 B0 S1 3. CR45 of Command Register 4 is set to Logic “0” during the following vertical retrace and the acquired signature is read. At least 20 clock cycles should be allowed for the final pixels of the frame to travel down the pipeline of the ADV7160/ ADV7162 before the signature clock is disabled. The signature analyzer is read from control registers 010H to 013H. These are read only 10-bit registers. The access to these registers depends whether the part is in 8-bit or 10-bit data bus mode and operates in the same way as accessing the color palette. Address Register CONTROL (A10–A0) REGISTERS CONTENTS 0013H Signature Misc Register 0 0 0012H Signature Blue Register 0011H Signature Green Register S20 S19 S18 S17 S16 S15 S14 S13 S12 S11 0010H Signature Red Register S10 S9 0 0 0 0 0 S32 S31 S0 S8 S7 S6 S5 S4 S3 S2 S1 S30 S29 S28 S27 S26 S25 S24 S23 S22 S21 1. CR45 and CR47 of Command Register 4 are set to Logic “1” during vertical retrace and CR46 of Command Register 4 is toggled to reset the analyzer. 2. A signature is acquired during the following active screen. –40– REV. 0 ADV7160/ADV7162 APPENDIX 6 JTAG TEST PORT (IEEE1149.1) TDI CE R/W C0 C1 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 THREE-STATE CONTROL LOADIN SCKIN SCKOUT CLOCK CLOCK LOADOUT PRGCKOUT PS0A PS0B PS0C PS0D PS1A PS1B PS1C PS1D R0A R0B R1A G1A B1A R1B G1B B1B R1C G1C B1C R1D G1D B1D R2A R2B G2A G2B B2A B2B R2C G2C R2D G2D B2C B2D R3A G3A G3B B3A R3B R3C G3C B3C R3D R4A G3D G4A B3D R4B R4C G4B B4B G4C B4C R4D G4D B4D R5A G5A B5A R5B R5C G5B B5B B5C R5D R6A G5D G6A R6B R6C R6D G6B G6C G6D R7A R7B G7A G7B B7A R7C R7D G7C B7C G7D B7D G0A G0B B0A B0B B0C B0D TRISYNC ODD/EVEN SYNC BLANK SYNCOUT G0D R0D B4A G5C G0C R0C B3B B5D B6A B6B B6C B6D B7B TDO JTAG Boundary Scan Chain JTAG Test Port The Private1 instruction is for internal use in production test only. JTAG Test Port is a 4-pin interface consisting of: TCK: Test Clock TMS: Test Mode Select TDI: Test Data Input TDO: Test Data Output The IDCode is a 32-bit number which can be scanned out through TDO. Its contents are defined below: To put the ADV7160/ADV7162 into the required mode, the Instruction Register must be loaded. INSTRUCTION EXTEST SAMPLE/PRELOAD IDCODE PRIVATE 1 BYPASS BYPASS BYPASS BYPASS INSTRUCTION REGISTER CODE 000 001 010 011 100 101 110 111 The ADV7160 implementation has the mandatory instructions: Bypass, Sample/Preload and Extest, and the optional instruction: IDCode. There is also one private instruction: Private1. REV. 0 VERSION (4 BITS) PART NUMBER (16 BITS) MANUFACTURER ID (11 BITS) LSB ADV7160 1H 2776H 0E5H 1H ADV7162 1H 2779H 0E5H 1H The Boundary Scan Chain is a fundamental feature of the JTAG Test Port. It allows all the digital input and output pins on the part to be connected into a shift register between the TDI and TDO pins. The digital pins can be sampled, or controlled over the JTAG port to carry out testing. The is no boundary scan cell on the PLLREF pin. The Three-State Control cell controls the three-state status of the microport databus. There are 131 cells in total on the Boundary Scan Chain. –41– ADV7160/ADV7162 APPENDIX 7 THERMAL AND ENVIRONMENTAL CONSIDERATIONS The ADV7160/ADV7162 is a very highly integrated monolithic silicon device. This high level of integration, in such a small package, inevitably leads to consideration of thermal and environmental conditions which the ADV7160/ADV7162 must operate in. Reliability of the device is enhanced by keeping it as cool as possible. In order to avoid destructive damage to the device, the absolute maximum junction temperature of 150°C must never be exceeded. Certain applications, depending on ambient temperature and pixel data rates may require forced air cooling or external heatsinks. The following data is intended as a guide in evaluating the operating conditions of a particular application so that optimum device and system performance is achieved. It should be noted that information on package characteristics published herein may not be the most up to date at the time of reading this. Advances in package compounds and manufacture will inevitably lead to improvements in the thermal data. Please contact your local sales office for the most up-to-date information. Power Dissipation The diagrams show graphs of power dissipation in watts versus pixel clock frequency for the ADV7160 and ADV7162. When using the ADV7162 in Bypass Mode, the Pixel Mask Register should be programmed to 00H to reduce power further. 2.25 VAA = +5V VREF = +1.2V POWER DISSIPATION – Watts 2.00 TA = +25°C 1.75 ADV7160 ADV7162 1.50 1.25 1.00 Heatsinks The maximum silicon junction temperature should be limited to 100°C. Temperatures greater than this will reduce long-term device reliability. To ensure that the silicon junction temperature stays within prescribed limits, the addition of an external heatsink may be necessary. Heatsinks will reduce θJA as shown in the Thermal Characteristics vs. Airflow table. Table A. Thermal Characteristics vs. Airflow–ADV7160* Air Velocity (Linear Feet/min 0 50 (Still Air) 100 200 θJA °C/W No Heatsink EG&G D10100-28 Heatsink Thermalloy 2290 Heatsink 25.5 23 19 21 18 15 19 16 12 23 20 17 *These figures do not include thermal conduction through the package leads into the PCB. Thermal conduction through the leads can provide up to 10oC/W reduction in θJA. Table B. Thermal Characteristics vs. Airflow–ADV7162* Air Velocity (Linear Feet/min) 0 50 (Still Air) 100 200 θJA °C/W No Heatsink EG&G D10850-40 Heatsink EG&G D10851-36 Heatsink 37 28 32 30 22 19 28 19 14 32 24 24 *These figures do not include thermal conduction through the package leads into the PCB. Thermal conduction through the leads can provide up to 5oC/W reduction in θJA. Thermal Model The junction temperature of the device in a specific application is given by: 0.75 0.50 60 100 140 180 220 PIXEL CLOCK FREQUENCY – MHz 260 Note: The "Worst Case On-Screen Pattern" corresponds to full-scale transition on each pixel value for every CLOCK edge (00H, FFH, 00H, ...). The "Typical On-Screen Pattern" corresponds to linear changes in tne pixel input (i.e., a Black to White Ramp). In general, color images tend to approximate this characteristic. Typical Power Dissipation vs. Pixel Rate Package Characteristics The tables of thermal characteristics show typical information for the ADV7160 (160-Lead Plastic Power QFP) and AD7162 (160-Lead Plastic QFP) using various values of Airflow. Junction-to-Case (θJC) Thermal Resistance for this particular part is: θJC (AD7160) = 0.4°C/W θJC (AD7162) = 6.7°C/W (Note: θJC is independent of airflow.) TJ = TA + PD (θJC + θCA) (1) TJ = TA + PD (θJA) (2) or where: TJ = Junction Temperature of Silicon (°C) TA = Ambient Temperature (°C) PD = Power Dissipation (W) θJC = Junction to Case Thermal Resistance (°C/W) θCA = Case to Ambient Thermal Resistance (°C/W) θJA = Junction to Ambient Thermal Resistance (°C/W) Package Enhancements for ADV7160 The standard PQFP package has been enhanced to a PowerQuad2 package. This supports an improved thermal performance compared to standard PQFP. In this case, the die is attached to a heat slug so that the power that is dissipated can be conducted to the external surface of the package. This provides a highly efficient path for the transfer of heat to the package surface. The package configuration also provides and efficient thermal path from the ADV7160 to the Printed Circuit Board. –42– REV. 0 ADV7160/ADV7162 PAGE INDEX Topic Page FEATURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 GENERAL DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . 1 & 15 ADV7160/ADV7162 BLOCK DIAGRAM . . . . . . . . . . . . . . . . . 1 ADV7160/ADV7162 SPECIFICATIONS . . . . . . . . . . . . . . . . . 2 ADV7160/ADV7162 TIMING CHARACTERISTICS . . . . . 3-5 TIMING WAVEFORMS . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-11 ORDERING GUIDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 ABSOLUTE MAXIMUM RATINGS . . . . . . . . . . . . . . . . . . . 11 PIN CONFIGURATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 PIN FUNCTION DESCRIPTION . . . . . . . . . . . . . . . . . . 13-14 CIRCUIT DETAILS AND OPERATION . . . . . . . . . . . . . . . . 15 PIXEL PORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-16 CLOCK CONTROL CIRCUIT . . . . . . . . . . . . . . . . . . . . . 16-17 CLOCK CONTROL SIGNALS . . . . . . . . . . . . . . . . . . . . . 17-18 PLL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 COLOR VIDEO MODES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 PIXEL PORT MAPPING . . . . . . . . . . . . . . . . . . . . . . . . . 19-21 MULTIPLEXING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-22 MPU PORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 INTERNAL REGISTER CONFIGURATION . . . . . . . . . . . . 23 COLOR PALETTE ACCESS . . . . . . . . . . . . . . . . . . . . . . 24-25 ON-CHIP REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Address Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Mode Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Pixel Mask Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Command Register 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Command Register 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Command Register 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28-29 Command Register 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29-30 Command Register 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 PLL Command Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 PLL R Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 PLL V Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Status Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Revision Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 CURSOR DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . 31-32 Cursor X-Low and X-High Register . . . . . . . . . . . . . . . . . . . 31 Cursor Y-Low & Y-High Register . . . . . . . . . . . . . . . . . . . . . 31 Cursor Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Cursor Y Coordinate Even . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Cursor Y Coordinate Odd . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Cursor Control Register . . . . . . . . . . . . . . . . . . . . . . . . . 31–32 DACS & VIDEO OUTPUTS . . . . . . . . . . . . . . . . . . . . . . . 32-33 APPENDIX 1 Board Design and Layout Considerations . . . . . . . . . . . . 34-35 APPENDIX 2 Typical Frame Buffer Interface . . . . . . . . . . . . . . . . . . . . . . . 36 APPENDIX 3 10-Bit DACs and Gamma Correction . . . . . . . . . . . . . . . . . . 37 APPENDIX 4 Initialization and Programming . . . . . . . . . . . . . . . . . . . . 38-39 APPENDIX 5 Signature Analyzer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 APPENDIX 6 JTAG Test Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 APPENDIX 7 Thermal and Environmental Considerations . . . . . . . . . . . . . 42 INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 OUTLINE DIMENSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 REV. 0 FIGURE INDEX Figure Title 1 Load Circuit for Data-Bus Access &Relinquish Times 2 JTAG Port Timing 3 LOADOUT vs. Pixel Clock Input 4 LOADIN vs. Pixel Input Data 5 Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (8:1 Mode) 6 Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (8:1 Mode) 7 Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (4:1 Mode) 8 Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (4:1 Mode) 9 Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (2:1 Mode) 10 Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (2:1 Mode) 11 Pixel Clock Input vs. Programmable Clock Output 12 SCKIN vs. SCKOUT 13 Analog Output Response vs, Pixel Clock 14 MPU Timing 15 Multiplexed Color Inputs 16 Clock Control Circuit 17 LOADOUT vs. Pixel Clock 18 SCKOUT Generation Circuit 19 Interface Using SCKIN and SCKOUT 20 PLL Block Diagram 21 PLL Transfer Function 22 PLL Jitter 23 24-Bit to 30-Bit True Color Configuration 24 15-Bit to 24-Bit True Color Configuration 25 8-Bit to 30-Bit Pseudo Color Configuration 26 16-Bit Tue Color Mapping Using R7–R0 and G7–G0 27 15-Bit True Color Mapping Using R7–R3, G7–G3 and B7–B3 28 15-Bit True Color Mapping Using R6–R0 and G7–G0 29 16-Bit True Color (Bypass) Using R7–R0 and G7–G0 30 15-Bit True Color (Bypass) Using R6–R0 and G7–G0 31 Direct Interfacing of Video Memory 32 8-Bit Pseudo Color in 8:1 Multiplexing Mode 33 MPU Port and Register Configuration 34 Internal Register Configuration and Address Decoding 35 8-Bit Databus Using 10-Bit DACs 36 8-Bit Databus Using 8-Bit DACs 37 10-Bit Databus Using 10-Bit DACs 38 10-Bit Databus Using 8-Bit DACs 39 Mode Register 1 40 Command Register 1 41 Command Register 2 42 Command Register 3 43 Command Register 4 44 PLL Command Register 45 Cursor Control Register 46 DAC Output Termination 47 Composite Video Waveform, SYNC decoded; Pedestal = 7.5 IRE; DAC Gain = 3996 48 Composite Video Waveform, SYNC decoded; Pedestal = 0 IRE; DAC Gain = 4224 49 Composite Video Waveform, SYNC & TRISYNC decoded; Pedestal = 7.5 IRE; DAC Gain = 5592 50 Composite Video Waveform, Pedestal = 0 IRE; DAC Gain = 4311 –43– ADV7160/ADV7162 OUTLINE DIMENSIONS Dimensions shown in inches and (mm). 1.239 (31.45) SQ 1.219 (30.95) 1.107 (28.10) SQ 1.100 (27.90) 0.160 (4.07) MAX 0.037 (0.95) 0.026 (0.65) C2013–6–4/95 S-160 160-Lead Plastic Quad Flatpack 6°±4° 120 121 81 80 4°±4° MAX TOP VIEW (PINS DOWN) SEATING PLANE 10° 0.004 (0.10) MAX PIN 1 160 41 40 1 0.070 (1.77) 0.062 (1.57) 0.070 (1.77) 0.062 (1.57) 0.026 (0.65) MIN 0.014 (0.35) 0.011 (0.27) PRINTED IN U.S.A. 0.145 (3.67) 0.125 (3.17) –44– REV. 0