Datasheet ADV7150 Datasheet (ANALOG DEVICES)

现货库存、技术资料、百科信息、热点资讯,精彩尽在鼎好!
a
FEATURES 220 MHz, 24-Bit (30-Bit Gamma Corrected) True Color Triple 10-Bit “Gamma Correcting” D/A Converters Triple 256 3 10 (256 3 30) Color Palette RAM On-Chip Clock Control Circuit Palette Priority Select Registers RS-343A/RS-170 Compatible Analog Outputs TTL Compatible Digital Inputs Standard MPU l/O Interface
10-Bit Parallel Structure
8+2 Byte Structure Programmable Pixel Port: 24-Bit, 15-Bit and
Programmable Pixel Port: 8-Bit (Pseudo)
Pixel Data Serializer
Multiplexed Pixel Input Ports; 1:1, 2:1, 4:1 +5 V CMOS Monolithic Construction 160-Lead Plastic Quad Flatpack (QFP) Thermally Enhanced to Achieve u
MODES OF OPERATION 24-Bit True Color (30-Bit Gamma Corrected)
@ 220 MHz
@ 170 MHz
@ 135 MHz
@ 110 MHz
< 1.08C/W
JC
Triple 10-Bit Video RAM-DAC
ADV7150
@ 85 MHz
8-Bit Pseudo Color 15-Bit True Color
APPLICATIONS High Resolution, True Color Graphics Professional Color Prepress Imaging
GENERAL DESCRIPTION
The ADV7150 (ADV®) is a complete analog output, Video RAM-DAC on a single CMOS monolithic chip. The part is spe­cifically designed for use in high performance, color graphics workstations. The ADV7150 integrates a number of graphic functions onto one device allowing 24-bit direct True-Color op­eration at the maximum screen update rate of 220 MHz. The ADV7150 implements 30-bit True Color in 24-bit frame buffer designs. The part also supports other modes, including 15-bit True Color and 8-bit Pseudo or Indexed Color. Either the Red, Green or Blue input pixel ports can be used for Pseudo Color.
(Continued on page 12)
ADV is a registered trademark of Analog Devices, Inc.
FUNCTIONAL BLOCK DIAGRAM
V
AA
RED (R7–R0), GREEN (G7–G0), BLUE (B7–B0) COLOR DATA
PALETTE SELECTS
(PS0, PS1)
LOADIN
LOADOUT
PRGCKOUT
SCKOUT
SCKIN
SYNC
BLANK
CLOCK CLOCK
24
A
B
C
D
P
24
X E
24
L
P O
24
R T
8
CLOCK CONTROL
CLOCK DIVIDE
SYNCHRONIZATION
÷32 ÷16, ÷8, ÷4, ÷2
ECL TO CMOS
I
&
CIRCUIT
96
8
ADV7150
MUX
4:1
MUX
4:1
ADDRESS REGISTER
8 8 8
2
MODE
REGISTER
ADDR
(A7–A0)
CE R/W C0 C1
(MR1)
REV. A
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.
256-COLOR/GAMMA
PALETTE RAM
RED
256 x 10
GREEN
256 x 10
BLUE
256 x 10
CONTROL REGISTERS
PIXEL MASK
REGISTER
TEST
REGISTERS
ID
REGISTER
MPU PORT
10 (8+2)
D9 – D0
COMMAND
REGISTERS
(CR1–CR3)
REVISION
REGISTER
10
10
10
DATA TO
PALETTES
REGISTER
10-BIT
RED DAC
10-BIT
GREEN DAC
10-BIT
BLUE DAC
30
RED
SYNC
OUTPUT
VOLTAGE
REFERENCE
CIRCUIT
COLOR REGISTERS
BLUE
REGISTER
GREEN
REGISTER
GND
IOR IOR
IOG IOG
IOB IOB
I
PLL
SYNCOUT V
REF
R
SET
COMP
© Analog Devices, Inc., 1996
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 617/329-4700 Fax: 617/326-8703
ADV7150–SPECIFICA TIONS
CL = 10 pF); IOR, IOG, IOB = GND. All specifications T
MIN
1
(V
= +5 V; V
AA
2
to T
unless otherwise noted.)
MAX
= +1.235 V; R
REF
= 280 V. IOR, IOG, IOB (RL = 37.5 V,
SET
Parameter All Versions Unit Test Conditions/Comments
STATIC PERFORMANCE
Resolution (Each DAC) 10 Bits Accuracy (Each DAC)
Integral Nonlinearity ±1 LSB max Differential Nonlinearity ±1 LSB max Guaranteed Monotonic Gray Scale Error ±5 % Gray Scale max
Coding Binary
DIGITAL INPUTS (Excluding CLOCK, CLOCK)
Input High Voltage, V Input Low Voltage, V Input Current, I Input Capacitance, C
INL
IN
IN
INH
2 V min
0.8 V max ±10 µA max VIN = 0.4 V or 2.4 V 10 pF max
CLOCK INPUTS (CLOCK, CLOCK)
Input High Voltage, V Input Low Voltage, V Input Current, I Input Capacitance, C
INL
IN
IN
INH
VAA – 1.0 V min VAA – 1.6 V max ±10 µA max VIN = 0.4 V or 2.4 V 10 pF typ
DIGITAL OUTPUT
Output High Voltage, V Output Low Voltage, V
OL
OH
2.4 V min I
0.4 V max I
SOURCE
= 3.2 mA
SINK
= 400 µA
Floating-State Leakage Current 20 µA max Floating-State Output Capacitance 20 pF typ
ANALOG OUTPUTS
Gray Scale Current Range 15/22 mA min/max Output Current
White Level Relative to Blank 17.69/20.40 mA min/max Typically 19.05 mA White Level Relative to Black 16.74/18.50 mA min/max Typically 17.62 mA Black Level Relative to Blank 0.95/1.90 mA min/max Typically 1.44 mA Blank Level on IOR, IOB 0/50 µA min/max Typically 5 µA Blank Level on IOG 6.29/8.96 mA min/max Typically 7.62 mA Sync Level on IOG 0/50 µA min/max Typically 5 µA
LSB Size 17.22 µA typ DAC-to-DAC Matching 3 % max Typically 1% Output Compliance, V Output Impedance, R
OUT
Output Capacitance, C
OC
OUT
0/+1.4 V min/V max 100 k typ 30 pF max I
OUT
= 0 mA
VOLTAGE REFERENCE
Voltage Reference Range, V Input Current, I
VREF
REF
1.14/1.26 V min/V max V +5 µA typ
= 1.235 V for Specified Performance
REF
POWER REQUIREMENTS
V
AA
3
I
AA
3
I
AA
I
AA
I
AA
I
AA
5 V nom 400 mA max 220 MHz Parts 370 mA max 170 MHz Parts 350 mA max 135 MHz Parts 330 mA max 110 MHz Parts 315 mA max 85 MHz Parts
Power Supply Rejection Ratio 0.5 %/% max Typically 0.12%/%: COMP = 0.1 µF
DYNAMIC PERFORMANCE
Clock and Data Feedthrough
Glitch Impulse 50 pV secs typ DAC-to-DAC Crosstalk
NOTES
1
±5% for all versions.
2
Temperature range (T
3
Pixel Port is continuously clocked with data corresponding to a linear ramp. TJ = 100°C.
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 to 3 volts, 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.
MIN
to T
4, 5
6
): 0°C to +70°C; TJ (Silicon Junction Temperature) 100°C.
MAX
–30 dB typ
–23 dB typ
–2–
REV. A
TIMING CHARACTERISTICS
IOR, IOG, I0B = GND. All specifications T
CLOCK CONTROL AND PIXEL PORT
MIN
to T
4
= +5 V; V
AA
3
unless otherwise noted.)
MAX
= +1.235 V; R
REF
= 280 V. IOR, IOG, IOB (RL = 37.5 V, CL = 10 pF);
SET
1
2
(V
220 MHz 170 MHz 135 MHz 110 MHz 85 MHz
Parameter Version Version Version Version Version Units
ADV7150
Conditions/Comments
f
CLOCK
t
1
t
2
t
3
t
4
f
LOADIN
220 170 135 110 85 MHz max Pixel CLOCK Rate
4.55 5.88 7.4 9.1 11.77 ns min Pixel CLOCK Cycle Time 2 2.5 3.2 4 4 ns min Pixel CLOCK High Time 2 2.5 3 4 4 ns min Pixel CLOCK Low Time 10 10 10 10 10 ns max Pixel CLOCK to LOADOUT Delay
LOADIN Clocking Rate 1:1 Multiplexing 110 110 110 110 85 MHz max 2:1 Multiplexing 110 85 67.5 55 42.5 MHz max 4:1 Multiplexing 55 42.5 33.75 27.5 21.25 MHz max
t
5
LOADIN Cycle Time 1:1 Multiplexing 9.1 9.1 9.1 9.1 9.1 ns min 2:1 Multiplexing 9.1 11.76 14.8 18.18 23.53 ns min 4:1 Multiplexing 18.18 23.53 29.63 36.36 47.1 ns min
t
6
LOADIN High Time 1:1 Multiplexing 44444ns min 2:1 Multiplexing 45689ns min 4:1 Multiplexing 8 9 12 15 18 ns min
t
7
LOADIN Low Time 1:1 Multiplexing 44444ns min 2:1 Multiplexing 45689ns min 4:1 Multiplexing 8 9 12 15 18 ns min
t
8
t
9
t
10
5
τ–t
11 6
t
PD
1:1 Multiplexing 55555CLOCKs (1 × CLOCK = t
00000ns minPixel Data Setup Time 55555ns minPixel Data Hold Time 00000ns minLOADOUT to LOADIN Delay τ–5 τ–5 τ–5 τ–5 τ–5 ns max LOADOUT to LOADIN Delay
Pipeline Delay
)
1
2:1 Multiplexing 66666CLOCKs 4:1 Multiplexing 88888CLOCKs
t
12
t
13
t
14
t
15
ANALOG OUTPUTS
10 10 10 10 10 ns max Pixel CLOCK to PRGCKOUT Delay 55555ns maxSCKIN to SCKOUT Delay 55555ns minBLANK to SCKIN Setup Time 11111ns minBLANK to SCKIN Hold Time
7
220 MHz 170 MHz 135 MHz 110 MHz 85 MHz
Parameter Version Version Version Version Version Units Conditions/Comments
t
16
t
17
t
18
t
SK
15 15 15 15 15 ns typ Analog Output Delay 11111ns typAnalog Output Rise/Fall Time 15 15 15 15 15 ns typ Analog Output Transition Time 22222ns maxAnalog Output Skew (IOR, IOG, IOB) 00000ns typ
MPU PORTS
8, 9
220 MHz 170 MHz 135 MHz 110 MHz 85 MHz
Parameter Version Version Version Version Version Units Conditions/Comments
t
19
t
20
t
21
t
22
8
t
23
9
t
24
9
t
25
33333ns minR/W, C0, C1 to CE Setup Time 10 10 10 10 10 ns min R/W, C0, C1 to CE Hold Time 45 45 45 45 45 ns min CE Low Time 25 25 25 25 25 ns min CE High Time 55555ns minCE Asserted to Databus Driven 45 45 45 45 45 ns max CE Asserted to Data Valid 20 20 20 20 20 ns max CE Disabled to Databus Three-Stated 55555ns min
t
26
t
27
REV. A
20 20 20 20 20 ns min Write Data (D0–D9) Setup Time 55555ns minWrite Data (D0–D9) Hold Time
–3–
ADV7150
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. Databus (D0–D9) loaded as shown in Figure 1. Digital output load for LOADOUT, PRGCKOUT, SCKOUT, I SYNCOUT 30 pF.
2
±5% for all versions.
3
Temperature range (T
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; 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: 1:1 multiplexing; τ = CLOCK = t1 ns. 2:1 Multi-
plexing; τ = CLOCK × 2 = 2 × t1 ns. 4:1 Multiplexing; τ = CLOCK × 4 = 4 × t1 ns.
6
These fixed values for Pipeline Delay are valid under conditions where t10 and τ-t11 are met. If either t10 or τ-t11 are not met, the part will operate but the Pipe line De-
lay is increased by 2 additional CLOCK cycles for 2:1 Mode and is increased by 4 additional CLOCK cycles for 4:1 Mode, after calibration is performed.
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%
MIN
to T
): 0°C to +70°C; TJ (Silicon Junction Temperature) 100°C.
MAX
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).
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 is 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 number is then extrapo-
lated back to remove the effects of charging the 100 pF capacitor. This means that the time, t25, quoted in the Timing Characteristics is the true value for the device and as such is independent of external databus loading capacitances.
Specifications subject to change without notice.
I
SINK
TO OUTPUT PIN
100pF
+2.1V
PLL
and
CLOCK
CLOCK
LOADOUT
(1:1 MULTIPLEXING)
LOADOUT
(2:1 MULTIPLEXING)
LOADOUT
(4:1 MULTIPLEXING)
LOADIN
PIXEL INPUT
DATA*
I
SOURCE
Figure 1. Load Circuit for Databus Access and Relinquish Times
t
2
VALID
DATA
t
3
CLOCK
t
6
t
7
VALID
DATA
t
1
t
4
Figure 2. LOADOUT vs. Pixel Clock Input (CLOCK,
t
5
t8t
9
VALID
DATA
)
*INCLUDES PIXEL DATA (R0–R7, G0–G7, B0–B7); PALETTE SELECT INPUTS (PS0–PS1); BLANK; SYNC
Figure 3. LOADIN vs. Pixel Input Data
–4–
REV. A
CLOCK
LOADOUT
LOADIN
ADV7150
t
10
PIXEL INPUT
DATA*
ANALOG
OUTPUT
DATA
A
B
N
N
C
D
N
N
DIGITAL INPUT TO ANALOG
OUTPUT PIPELINE
IOR, IOR IOG, IOG IOB, IOB I
PLL, SYNCOUT
A
N+1 BN+1
C
N+1DN+1
A
N–1
t
PD
B
N–1CN–1DN–1
A
N+2 BN+2
C
N+2 DN+2
ANBNCNDNA
N+1BN+1CN+1DN+1
A
N+2BN+2CN+2DN+2
*INCLUDES PIXEL DATA (R0–R7, G0–G7, B0–B7); PALETTE SELECT INPUTS (PS0–PS1); BLANK; SYNC
Figure 4. Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (4:1 Multiplex Mode)
CLOCK
τ
τ– t
11
LOADOUT
LOADIN
PIXEL INPUT
DATA*
ANALOG
OUTPUT
DATA
B
N–1CN–1DN–1
A
N+1 BN+1
C
N+1DN+1
ANBNCND
IOR, IOR IOG, IOG IOB, IOB I
PLL, SYNCOUT
A
B
N
N
C
D
N
N
DIGITAL INPUT
TO ANALOG
OUTPUT
PIPELINE
A
N–1
t
PD
*INCLUDES PIXEL DATA (R0–R7, G0–G7, B0–B7); PALETTE SELECT INPUTS (PS0–PS1); BLANK; SYNC
A
N+2 BN+2
C
N+2 DN+2
A
N
N+1BN+1CN+1DN+1
A
N+2BN+2CN+2DN+2
Figure 5. Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (4:1 Multiplex Mode)
REV. A
–5–
ADV7150
LOADOUT
LOADIN
CLOCK
t
10
PIXEL INPUT
DATA*
ANALOG
OUTPUT
DATA
A
B
N
N
DIGITAL INPUT TO ANALOG
OUTPUT PIPELINE
IOR, IOR IOG, IOG IOB, IOB I
PLL, SYNCOUT
A
N+1 BN+1
A
N+2 BN+2
A
A
N-1BN-1
t
PD
B
N
A
N
N+1BN+1AN+2BN+2
*INCLUDES PIXEL DATA (R0–R7, G0–G7, B0–B7); PALETTE SELECT INPUTS (PS0–PS1); BLANK; SYNC
Figure 6. Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (2:1 Multiplex Mode)
CLOCK
τ
τ– t
11
LOADOUT
LOADIN
PIXEL INPUT
DATA*
A
B
N
N
DIGITAL INPUT TO ANALOG
OUTPUT PIPELINE
A
N+1 BN+1
A
N+2 BN+2
ANALOG
OUTPUT
DATA
IOR, IOR IOG, IOG IOB, IOB I
PLL, SYNCOUT
B
A
N-1BN-1
t
PD
A
N
A
N
N+1BN+1AN+2
B
N+2
*INCLUDES PIXEL DATA (R0–R7, G0–G7, B0–B7); PALETTE SELECT INPUTS (PS0–PS1); BLANK; SYNC
Figure 7. Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (2:1 Multiplex Mode)
–6–
REV. A
CLOCK
PRGCKOUT
(CLOCK/4)
PRGCKOUT
(CLOCK/8)
PRGCKOUT
(CLOCK/16)
PRGCKOUT
(CLOCK/32)
SCKIN
BLANK
ADV7150
t
12
Figure 8. Pixel Clock Input vs. Programmable Clock Output (PRGCKOUT)
t
13
t
15
t
14
BLANKING
PERIOD
SCKOUT
END OF SCAN LINE (N)
START OF SCAN LINE (N+1)
Figure 9. Video Data Shift Clock Input (SCKIN) & BLANK vs. Video Data Shift Clock Output (SCKOUT)
CLOCK
t
18
WHITE LEVEL
90 %
50 %
10 % BLACK LEVEL
NOTE: THIS DIAGRAM IS NOT TO SCALE. FOR THE PURPOSES OF CLARITY, THE ANALOG OUTPUT WAVEFORM IS MAGNIFIED IN TIME AND AMPLITUDE W.R.T THE CLOCK WAVEFORM. I
AND SYNCOUT ARE DIGITAL VIDEO OUTPUT SIGNALS.
PLL
t
IS THE ONLY RELEVENT OUTPUT TIMING SPECIFICATION
16
FOR I
AND SYNCOUT.
PLL
FULL-SCALE TRANSITION
ANALOG
OUTPUTS
IOR, IOR IOG, IOG IOB, IOB I
PLL, SYNCOUT
t
16
t
17
REV. A
Figure 10. Analog Output Response vs. CLOCK
–7–
ADV7150
R/W, C0, C1
(READ MODE)
D0–D9
t
19
CONTROL DATA
CE
VALID
t
23
t
20
t
21
t
22
t
24
R/W = 1
t
25
D0–D9
(WRITE MODE)
R/W = 0
t
26
t
27
Figure 11. Microprocessor Port (MPU) Interface Timing
RECOMMENDED OPERATING CONDITION
Parameter Symbol Min Typ Max Units
Power Supply V Ambient Operating Temperature T Reference Voltage V Output Load R
AA A REF L
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 ADV7150 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.
ABSOLUTE MAXIMUM RATINGS
1
VAA to GND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 V
Voltage on Any Digital Pin . . . . GND – 0.5 V to V
Ambient Operating Temperature (T Storage Temperature (T Junction Temperature (T
) . . . . . . . . . . . . . . –65°C to +150°C
S
) . . . . . . . . . . . . . . . . . . . . +150°C
J
) . . . . . –55°C to +125°C
A
+ 0.5 V
AA
Lead Temperature (Soldering, 10 secs) . . . . . . . . . . . +260°C
Vapor Phase Soldering (1 minute) . . . . . . . . . . . . . . . +220°C
Analog Outputs to GND
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.
2
. . . . . . . . . . . . . GND – 0.5 to V
AA
4.75 5.00 5.25 Volts 0 +70 °C
1.14 1.235 1.26 Volts
37.5
WARNING!
ESD SENSITIVE DEVICE
16-Lead QFP Configuration
160 121
1
PIN NO. 1 IDENTIFIER
ROW A
ROW D
ADV7150 QFP
TOP VIEW
(NOT TO SCALE)
ROW C
120
ORDERING GUIDE
1, 2, 3
Speed
220 MHz ADV7150LS220 110 MHz ADV7150LS110 170 MHz ADV7150LS170 85 MHz ADV7150LS85 135 MHz ADV7150LS135
NOTES
1
ADV7150 is packaged in a 160-pin plastic quad flatpack, QFP.
2
All devices are specified for 0°C to +70°C operation.
3
Contact sales office for latest information on package design.
–8–
40
41 80
ROW B
81
REV. A
ADV7150
ADV7150 PIN ASSIGNMENTS
Pin Pin Pin Pin Number Mnemonic Number Mnemonic Number Mnemonic Number Mnemonic
1G3 2G3 3G3 4G3 5G4 6G4 7G4 8G4 9G5 10 G5 11 G5 12 G5
A B C D A B C D A B C D
41 PS1 42 B0 43 B0 44 B0 45 B0 46 B1 47 B1 48 B1 49 B1 50 B2 5 1 B2
52 B2 13 CLOCK 53 B2 14 CLOCK 54 B3 15 LOADIN 55 B3 16 LOADOUT 56 B3 17 V 18 V
AA AA
57 B3
58 B4 19 PRGCKOUT 59 B4 20 SCKIN 60 B4 21 SCKOUT 61 B4 22 SYNCOUT 62 B5 23 GND 63 B5 24 GND 64 B5 25 GND 65 B5 26 G6 27 G6 28 G6 29 G6 30 G7 31 G7 32 G7 33 G7 34 PS0 35 PS0 36 PS0 37 PS0 38 PS1 39 PS1 40 PS1
NC = No Connect.
A B C D A B C D
A B C D A B C
66 B6
67 B6
68 B6
69 B6
70 B7
71 B7
72 B7
73 B7
74 CE 114 V
75 R/W 115 SYNC 155 G1
76 C0 116 BLANK 1 56 G1
77 C1 117 R0
78 D0 118 R0
79 D1 119 R0
80 GND 120 R0
D A B C
D A B C
D A B C
D A B C
D A B C
D A B C
D A B C
D A B C
D
81 NC 121 R1 82 D2 122 R1 83 NC 123 R1 84 GND 124 R1 85 GND 125 R2 86 GND 126 R2 87 D3 127 R2 88 D4 128 R2 89 D5 129 R3 90 V
AA
130 R3 91 D6 131 R3 92 D7 132 R3 93 D8 133 R4 94 D9 134 R4 95 GND 135 R4 96 GND 136 R4 97 GND 137 R5 98 IOB 138 R5 99 IOR 139 R5 100 IOG 140 R5 101 IOB 141 R6 102 IOG 142 R6 103 V 104 V 105 V
AA AA AA
143 R6
144 R6
145 R7 106 IOR 146 R7 107 COMP 147 R7 108 V 109 R 110 I
REF SET
PLL
148 R7
149 G0
150 G0 111 GND 151 G0 112 V 113 V
AA AA AA
A B C D
152 G0
153 G1
154 G1
157 G2
158 G2
159 G2
160 G2
A B C D A B C D A B C D A B C D A B C D A B C D A B C D
A B C D A B C D A B C D
REV. A
–9–
ADV7150
PIN FUNCTION DESCRIPTION
Mnemonic Function
RED (R0A . . . R0D–R7A . . . R7D), Pixel Port (TTL Compatible Inputs): 96 pixel select inputs, with 8 bits each for Red, 8 GREEN (G0 BLUE (B0
PS0
. . . PS0D, PS1A . . . PS1
A
LOADIN Pixel Data Load Input (TTL Compatible Input). This input latches the multiplexed
LOADOUT Pixel Data Load Output (TTL Compatible Output). This output control signal runs at
PRGCKOUT Programmable Clock Output (TTL Compatible Output). This output control signal
SCKIN Video Shift Clock Input (TTL Compatible Input). The signal on this input is internally
SCKOUT Video Shift Clock Output (TTL Compatible Output). This output is a synchronously
CLOCK,
BLANK Composite Blank (TTL Compatible Input). This video control signal drives the analog
SYNC Composite-Sync Input (TTL Compatible Input). This video control signal drives the
SYNCOUT Composite-Sync Output (TTL Compatible Output). This video output is a delayed
D0–D9 Databus (TTL Compatible Input/Output Bus). Data, including color palette values and
CE Chip Enable (TTL Compatible Input). This input must be at Logic “0,” when writing
. . . G0D–G7A. . . G7D), bits for Green and 8 bits for Blue. Each bit is multiplexed [A-D] 4:1, 2:1 or 1:1. It can
A
. . . B0D–B7A . . . B7D) be configured for 24-Bit True-Color Data, 8-Bit Pseudo-Color Data and 15-Bit True-Color
A
Data formats. Pixel Data is latched into the device on the rising edge of LOADIN.
D
Palette Priority Selects (TTL Compatible Inputs): These pixel port select inputs deter­mine whether or not the device’s pixel data port is selected on a pixel by pixel basis. The palette selects allow switching between multiple palette devices. The device can be preprogrammed to completely shut off the DAC analog outputs. 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. Each bit is multiplexed [A-D] 4:1, 2:1 or 1:1. PS0 and PS1 are latched into the device on the rising edge of LOADIN.
pixel data, including PS0–PS1,
BLANK and SYNC into the device.
a divided down frequency of the pixel CLOCK input. Its frequency is a function of the multiplex rate. It can be used to directly or indirectly drive LOADIN
f
LOADOUT
= f
CLOCK
/M
where M = 1 for 1:1 Multiplex Mode
where M = 2 for 2:1 Multiplex Mode where M = 4 for 4:1 Multiplex Mode.
runs at a divided down frequency of the pixel CLOCK input. Its frequency is user programmable and is determined by bits CR30 and CR31 of Command Register 3
f
PRGCKOUT
= f
CLOCK
/N
where N = 4, 8, 16 and 32.
gated synchronously with the
BLANK signal. The resultant output, SCKOUT, is a
video clocking signal that is stopped during video blanking periods.
gated version of SCKIN and
BLANK. SCKOUT, is a video clocking signal that is
stopped during video blanking periods.
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.
outputs to the blanking level.
IOG analog output to the CR22 in Command Re output, otherwise the
version of
SYNC. The delay corresponds to the number of pipeline stages of the device.
SYNC level. It is only asserted during the blanking period.
gister 2 must be set if SYNC is to be decoded onto the analog
SYNC input is ignored.
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 databus should be terminated through a resistor to either the digital power plane (V
) or GND.
CC
to or reading from the device over the databus (D0–D9). Internally, data is latched on the rising edge of CE.
–10–
REV. A
ADV7150
Mnemonic Function
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/ 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 databus (see Interface Truth Table). Data on these inputs is latched on the falling edge of
IOR, IOG; IOG, IOB; Red, Green and Blue Current Outputs (High Impedance Current Sources). These RGB
IOR;
CE.
IOB video outputs are specified to directly drive RS-343A and RS-170 video levels into dou-
bly terminated 75 loads. IOR, IOG and IOB are the complementary outputs of IOR, IOG and IOB. These out-
puts can be tied to GND if it is not required to use differential outputs.
V
REF
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.)
R
SET
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. The value of R
is derived from the full-scale output current on IOG according to the following
SET
equations:
R
(Ω) = C1 × V
SET
R
() = C2 × V
SET
/IOG (mA); SYNC on GREEN
REF
/IOG (mA); NO SYNC on GREEN.
REF
Full-Scale output currents on IOR and IOB for a particular value of R
IOR (mA)= C2 × V
REF
(V)/R
SET
()
and
IOB (mA) = C2 × V
REF
(V)/R
SET
()
where C1 = 6,050; PEDESTAL = 7.5 IRE
where C1 = 5,723; PEDESTAL = 0 IRE
and
where C2 = 4,323; PEDESTAL = 7.5 IRE
where C1 = 3,996; PEDESTAL = 0 IRE.
COMP Compensation Pin. A 0.1 µF capacitor should be connected between this pin and VAA. I
PLL
Phase Lock Loop Output Current (High Impedance Current Source). This output is
used to enable multiple ADV7150s along with ADV7151s to be synchronized together
with pixel resolution when using an external PLL. This output is triggered either from
the falling edge of
SYNC or BLANK as determined by bit CR21 of Command Register
2. When activated, it supplies a current corresponding to:
I
(mA) = 1,728 × V
PLL
When not using the I V
AA
Power Supply (+5 V ± 5%). The part contains multiple power supply pins, all should be
function, this output pin should be tied to GND.
PLL
REF
(V)/R
SET
()
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.
W and CE must
are given by:
SET
REV. A
–11–
ADV7150
A
B
C
D
MULTIPLEXER
24
24
24
24
24
RED GREEN BLUE
8
8
8
(Continued from page 1)
The device consists of three, high speed, 10-bit, video D/A con­verters (RGB), three 256 × 10 (one 256 × 30) color look-up tables, palette priority selects, a pixel input data multiplexer/ serializer and a clock generator/divider circuit. The ADV7150 is capable of 1:1, 2:1 and 4:1 multiplexing. The onboard 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. The part also contains a number of onboard test registers, associated with self diagnostic testing of the de­vice. The individual Red, Green and Blue pixel input ports al­low True-Color, image rendition. True-Color image rendition, at speeds of up to 220 MHz, is achieved through the use of the onboard data multiplexer/serializer. The pixel input port’s flex­ibility allows for direct interface to most standard frame buffer memory configurations.
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 onboard implementation of lineariza­tion algorithms, such as Gamma-Correction. This allows effec­tive 30-Bit True-Color operation.
CIRCUIT DETAILS AND OPERATION
OVERVIEW
Digital video or pixel data is latched into the ADV7150 over the devices Pixel Port. This data acts as a pointer to the onboard Color Palette RAM. The data at the RAM address pointed to is latched into the digital-to-analog converters (DACs) and output as an RGB analog video signal.
For the purposes of clarity of description, the ADV7150 is bro­ken down into three separate functional blocks. These are:
1. Pixel port and clock control circuit
2. MPU port, registers and color palette
3. Digital-to-analog converters and video outputs Table I shows the architectural and packaging differences be-
tween other devices in the ADV715x series of workstation parts. (For more details consult the relevant data sheets.)
Table I. Architectural and Packaging Differences of the ADV715x Series
Description ADV7150 ADV7152* ADV7151*
24-Bit “Gamma” True Color • 24-Bit “Standard” True Color • 8-Bit “Gamma” Pseudo Color • 8-Bit “Standard” Pseudo Color • 15-Bit True Color • 220 MHz – True Color • 220 MHz – Pseudo Color • Triple 10-Bit DACs • 4:1 Multiplexing • 2:1 Multiplexing • 1:1 Multiplexing • 160-Lead QFP • 100-Lead QFP
*See ADV7151 and ADV7150 data sheets for more information on these parts.
The on-chip video clock controller circuit generates all the inter­nal clocking and some additional external clocking signals. An external ECL oscillator source with differential outputs is all that is required to drive the CLOCK and
CLOCK inputs of the ADV7150. The part can also be driven by an external clock gen­erator chip circuit, such as the AD730.
The ADV7150 is capable of generating RGB video output sig­nals which are compatible with RS-343A and RS-170 video standards, without requiring external buffering.
Test diagnostic circuitry has been included to complement the users system level debugging.
The ADV7150 is fabricated in a +5 V CMOS process. Its monolithic CMOS construction ensures greater functionality with low power dissipation.
The ADV7150 is packaged in a plastic 160-pin power quad flat­pack (QFP). Superior thermal dissipation is achieved by inclu­sion of a copper heatslug, within the standard package outline to which the die is attached.
Pixel Port and Clock Control Circuit
The Pixel Port of the ADV7150 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 RED, GREEN, BLUE Pixel Controls
SYNC, BLANK
Palette Selects PS0–PS1 The associated clocking signals for the pixel port include: Clock Inputs CLOCK, CLOCK,
LOADIN, SCKIN
Clock Outputs LOADOUT, PRGCKOUT,
SCKOUT
These onboard clock control signals are included to simplify in­terfacing between the part and the frame buffer. Only two con­trol input signals are necessary to get the part operational, CLOCK and
CLOCK (ECL Levels). No additional signals or external glue logic are required to get the Pixel Port & Clock Control Circuit of the part operational.
Pixel Port (Color Data)
The ADV7150 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, 15-Bit True Color and 8-Bit Pseudo Color (see “Color Data Formats” section) in 4:1, 2:1 and 1:1 multiplex modes.
Figure 12. Multiplexed Color Inputs for the ADV7150
–12–
REV. A
ADV7150
CLOCK
ADV7150
CLOCK
DIVIDE BY
N (÷ N)
LOADOUT
DIVIDE BY
M (÷ M)
PRGCKOUT
LOADIN
SCKOUT
SCKIN
BLANK
LATCH
ENABLE
SYNC
TO COLOR DATA
MULTIPLEXER
ECL
TO
TTL
M IS A FUNCTION OF MULTIPLEX RATE M = 4 IN 4:1 MULTIPLEX MODE M = 2 IN 2:1 MULTIPLEX MODE M = 1 IN 1:1 MULTIPLEX MODE
N IS INDEPENDENTLY PROGRAMMABLE N= (4, 8, 16, 32)
Color data is latched into the parts pixel port on every rising edge of LOADIN (see Timing Waveform, Figure 3). The required frequency of LOADIN is determined by the multiplex rate, where:
f
LOADIN
f
LOADIN
f
LOADIN
= f = f = f
/4 4:1 Multiplex Mode
CLOCK
/2 2:1 Multiplex Mode
CLOCK
CLOCK
1:1 Multiplex Mode
Other pixel data signals latched into the device by LOADIN include
SYNC, BLANK and PS0–PS1.
Internally, data is pipelined through the part by the differential pixel clock inputs, CLOCK and
CLOCK. The LOADIN con­trol signal needs only have a frequency synchronous relationship to the pixel CLOCK (see “Pipeline Delay & Onboard Calibra­tion” section). A completely phase independent LOADIN signal can be used with the ADV7150, allowing the CLOCK to occur anywhere during the LOADIN cycle.
Alternatively, the LOADOUT signal of the ADV7150 can be used. LOADOUT can be connected either directly or indirectly to LOADIN. Its frequency is automatically set to the correct LOADIN requirement.
SYNC, BLANK
The BLANK and SYNC video control signals drive the analog outputs to the blanking 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 a Logic “1.” The
SYNC input is ignored if CR22 is set to “0.”
SYNCOUT
In some applications where it is not permissible to encode SYNC on green (IOG), SYNCOUT can be used as a separate TTL digital pendent (of the ADV7150)
SYNC output. This has the advantage over an inde-
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 set to “0” thereby pre­venting Command Register 1 is set to “1,” enabling output signal generates a TTL delay that is capable of directly driving the composite
SYNC from being encoded onto IOG. Bit CR12 of
SYNCOUT. The
SYNCOUT with correct pipeline
SYNC
signal of a computer monitor.
PS0–PS1 (Palette Priority Select Inputs)
These pixel port select inputs determine whether or not the de­vice is selected. These controls 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 (see Appendix 4). 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.
Multiplexing
The onboard multiplexers of the ADV7150 eliminate the need for external data serializer circuits. Multiple video memory devices can be connected, in parallel, directly to the device.
VIDEO MEMORY/ FRAME BUFFER
VRAM (BANK A)
VRAM (BANK B)
VRAM (BANK C)
VRAM (BANK D)
24
33MHz
24
33MHz
24
33MHz
24
33MHz
ADV7150
MULTIPLEXER
24
132 MHz
(4 x 33 MHz)
Figure 13. Direct Interfacing of Video Memory to ADV7150
Figure 13 shows four memory banks of 33 MHz memory con­nected to the ADV7150, running in 4:1 multiplex mode, giving a resultant pixel or dot clock rate of 132 MHz. As mentioned in the previous section, the ADV7150 supports a number of color data formats in 4:1, 2:1 and 1:1 multiplex modes.
In 1:1 multiplex mode, the ADV7150 is clocked using the LOADIN signal. This means that there is no requirement for dif­ferential ECL inputs on CLOCK and
CLOCK. The pixel clock is connected directly to LOADIN. (Note: The ECL CLOCK can still be used to generate LOADOUT PRGCKOUT, etc.)
CLOCK CONTROL CIRCUIT
The ADV7150 has an integrated Clock Control Circuit (Figure
14). This circuit is capable of both generating the ADV7150’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.
Figure 14. Clock Control Circuit of the ADV7150
REV. A
–13–
ADV7150
LOADOUT
LOADIN
ADV7150
VIDEO
FRAME
BUFFER
LOADOUT
LOADIN
ADV7150
VIDEO
FRAME
BUFFER
LOADOUT(1)
LOADOUT(2)
PIXEL DATA
PIXEL
DATA
LOADIN
LOADOUT LOADOUT(1)
LOADOUT(2)
DELAY
SCKOUT
SCKIN
BLANK
LATCH
ENABLE
SYNC
LOADOUT
SCKOUT
ADV7150
VIDEO
FRAME
BUFFER
PIXEL DATA
LOADIN SCKIN
BLANK
CLOCK, CLOCK Inputs
The Clock Control Circuit is driven by the pixel clock inputs, CLOCK and ential ECL oscillator running from a +5 V supply.
Alternatively, the ADV7150 CLOCK inputs can be driven by a Programmable Clock Generator (Figure 15), such as the ICS1562. The ICS1562 is a monolithic, phase-locked-loop, clock generator chip. It is capable of synthesizing differential ECL output frequencies in a range up to 220 MHz from a single low frequency reference crystal.
LOW FREQUENCY
OSCILLATOR
V
CLOCK
Figure 15. PLL Generator Driving CLOCK, ADV7150
CLOCK CONTROL SIGNALS LOADOUT
The ADV7150 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.
f
LOADOUT
f
LOADOUT
f
LOADOUT
The LOADOUT signal is used to directly drive the LOADIN pixel latch signal of the ADV7150. 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 (t
If, however, it is required that the ADV7150 has a fixed number of pipeline delays (t form to timing specifications t ures 4 to 7.
CLOCK. These inputs can be driven by a differ-
GND
+5V
ECL ECL
CLOCK
GENERATOR
V
REF
D0-D3CSR/W
= f
CLOCK
= f
CLOCK
= f
CLOCK
and t9).
8
OUT+ OUT–
OUT V
/4 4:1 Multiplex Mode /2 2:1 Multiplex Mode
), LOADOUT and LOADIN must con-
PD
V
GND
GND
1:1 Multiplex Mode
10
V
V
CC
220
330
CC
GND
220
330
0.1 µF
V
AA
CLOCK
and τ-t11 as illustrated in Fig-
AA
+5V
CLOCK CLOCK
ADV7150
REF
GND
of the
Figure 16. LOADOUT vs. Pixel Clock Input (CLOCK,
CLOCK
PRGCKOUT
The PRGCKOUT control signal outputs a user programmable clock frequency. It is a divided down frequency of the pixel CLOCK (see Figure 8). The rising edge of PRGCKOUT is synchronous to the rising edge of LOADOUT
f
PRGCKOUT
= f
CLOCK
/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.
SCKIN, SCKOUT
These video memory signals are used to minimize external sup­port chips. Figure 17 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 9 of the Timing Waveform section shows the relationship between SCKOUT, SCKIN and
BLANK.
Figure 17. SCKOUT Generation Circuit
The SCKOUT signal is essentially the video memory shift con­trol signal. It is stopped during the screen retrace. Figure 18 shows a suggested frame buffer to ADV7150 interface. This is a minimum chip solution and allows the ADV7150 control the overall graphics system clocking and synchronization.
Figure 18. ADV7150 Interface Using SCKIN and SCKOUT
–14–
REV. A
)
8
8-BIT
RED DAC
8-BIT
BLUE DAC
RED
256 x 8
GREEN
256 x 8
BLUE
256 x 8
8
8-BIT
GREEN DAC
8
8
8
8
RED OUT
GREEN OUT
BLUE OUT
24-BIT
COLOR DATA
24-BIT TO 24-BIT
LOOK-UP TABLE
24-BIT
COLOR DATA
ANALOG VIDEO
OUTPUTS
10
10-BIT
RED DAC
10-BIT
BLUE DAC
RED
256 x 10
GREEN
256 x 10
BLUE
256 x 10
10-BIT
GREEN DAC
10
10
8
RED OUT
GREEN OUT
BLUE OUT
8-BIT PIXEL
DATA
8-BIT TO 30-BIT
LOOK-UP TABLE
30-BIT
COLOR DATA
ANALOG VIDEO
OUTPUTS
8
8-BIT
RED DAC
8-BIT
BLUE DAC
RED
256 x 8
GREEN
256 x 8
BLUE
256 x 8
8-BIT
GREEN DAC
8
8
8
RED OUT
GREEN OUT
BLUE OUT
8-BIT PIXEL
DATA
8-BIT TO 24-BIT
LOOK-UP TABLE
24-BIT
COLOR DATA
ANALOG VIDEO
OUTPUTS
Pipeline Delay and Onboard Calibration
The ADV7150 has a fixed number of pipeline delays (tPD), so long as timings t delay is not a requirement, timings t
and τ-t11 are met. However, if a fixed pipeline
10
and τ-t11 can be ignored,
10
a calibration cycle must be run and there is no restriction on LOADIN to LOADOUT timing. If timings t
and τ-t11 are not
10
met, the part will function correctly though with an increased number of pipeline delays, t
+ N CLOCKS (for 4:1 mode
PD
N = 4, for 2:1 mode N = 2, for 1:1 mode N = 0). The ADV7150 has onboard calibration circuitry which synchronizes pixel data and LOADIN with the internal ADV7150 clocking signals. Calibra­tion can be performed in two ways: during the devices 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 which executes a calibration on every Vertical Sync.
COLOR VIDEO MODES
The ADV7150 supports a number of color video modes all at the maximum video rate.
Command bits CR24–CR27 of Command Register 2 along with Bit MR11 of Mode Register 1 determine the color mode.
24-Bit “Gamma” True Color (CR25, CR26, CR27 = 1, 1, 1 and MR11 = 1)
The part is set to 24-bit/30-bit True-Color operation. The pixel port accepts 24 bits of color data which is directly mapped to 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 (10 bits each for Red, Green and Blue). The RAM is preloaded with a user determined, nonlinear function, such as a gamma correction curve.
10-BIT
RED DAC
10-BIT
GREEN DAC
10-BIT
BLUE DAC
ANALOG VIDEO
OUTPUTS
RED OUT
GREEN OUT
BLUE OUT
24-BIT
COLOR DATA
8 8 8
24-BIT TO 30-BIT LOOK-UP TABLE
RED
256 x 10
GREEN
256 x 10
BLUE
256 x 10
30-BIT
COLOR DATA
10
10
10
Figure 19. 24-Bit to 30-Bit True-Color Configuration
This mode allows for the display of full 24-bit, Gamma­Corrected True-Color Images.
24-Bit “Standard” True Color (CR25, CR26, CR27 = 1, 1, 1 and MR11 = 0)
This mode sets the part into direct 24-bit True-Color operation. The pixel port accepts 24 bits of color data which is directly mapped to 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) and essentially acts as a bypass RAM. The output of the RAM drives the DACs with 24-bit data (8 bits each for Red, Green and Blue). The RAM is preloaded with a linear function.
This mode allows for the display of full 24-bit True-Color Images.
REV. A
–15–
ADV7150
Figure 20. 24-Bit to 24-Bit Direct True-Color Configuration
8-Bit “Gamma” Pseudo Color (CR25, CR26, CR27 = X, 0, 0 or X, 1, 0 or X, 0, 1 and MR11 = 1)
This mode sets the part into 8-bit Pseudo-Color operation. The pixel port accepts 8 bits of pixel data which indexes a 30-bit word 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 (10 bits each for Red, Green and Blue).
Figure 21. 8-Bit to 30-Bit Pseudo-Color Configuration
This mode allows for the display of 256 simultaneous colors out of a total palette of millions of addressable colors.
8-Bit “Standard” Pseudo Color (CR25, CR26, CR27 = X, 0, 0 or X, 1, 0 or X, 0, 1 and MR11 = 0)
This mode sets the part into 8-bit Pseudo-Color operation. The pixel port accepts 8 bits of pixel data which indexes a 24-bit word in the Look-Up Table RAM. The Look-Up Table is con­figured as a 256 location by 24 bits deep RAM (10 bits each for Red, Green and Blue). The output of the RAM drives the DACs with 24-bit data (8 bits each for Red, Green and Blue).
Figure 22. 8-Bit to 24-Bit Pseudo-Color Configuration
This mode allows for the display of 256 simultaneous colors out of a total palette of millions of addressable colors.
ADV7150
15-Bit “Gamma” True Color (CR24, CR25, CR26, CR27 = 0, 0, 1, 1 or 1, 0, 1, 1 and MR11 = 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. The Look-Up Table is configured as a 32 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 (10 bits each for Red, Green and Blue).
15-BIT
COLOR DATA
5
5 5
15-BIT TO 30-BIT LOOK-UP TABLE
RED
32 x 10
GREEN
32 x 10
BLUE
32 x 10
30-BIT
COLOR DATA
10
10
10
ANALOG VIDEO
10-BIT
RED DAC
10-BIT
GREEN DAC
10-BIT
BLUE DAC
OUTPUTS
RED OUT
GREEN OUT
BLUE OUT
Figure 23. 15-Bit to 30-Bit True-Color Configuration
This mode allows for the display of 15-bit, Gamma-Corrected True-Color Images.
15-Bit “Standard” True Color (CR24, CR25, CR26, CR27 = 0, 0, 1, 1 or 1, 0, 1, 1 and MR11 = 0)
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. The Look-Up Table is configured as a 32 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 (8 bits each for Red, Green and Blue).
8-BIT
RED DAC
8-BIT
GREEN DAC
8-BIT
BLUE DAC
ANALOG VIDEO
OUTPUTS
RED OUT
GREEN OUT
BLUE OUT
15-BIT
COLOR DATA
5 5 5
15-BIT TO 24-BIT
LOOK-UP TABLE
RED
32 x 8
GREEN
32 x 8
BLUE
32 x 8
24-BIT
COLOR DATA
8
8
8
Figure 24. 15-Bit to 24-Bit True-Color Configuration
R7
R4 R3 R2 R1 R0
G4 G3 G2 G1 G0
B4 B3 B2 B1 B0
PIXEL INPUT DATA
x x x
x x x
x x x
R6 R5 R4 R3 R2 R1 R0
G7 G6 G5 G4 G3 G2 G1 G0
B7 B6 B5 B4 B3 B2 B1
B0
PIN ASSIGN­MENTS
R4 R3 R2 R1 R0
x x x
G4 G3 G2 G1 G0
x x x
B4 B3 B2 B1 B0
x x x
DATA LATCHED TO PIXEL PORT
0 0
0 R4 R3
5
R2 R1 R0
0
0
0
G4 G3
5
G2 G1 G0
0
0
0
B4
B3
5
B2 B1 B0
DATA INTERNALLY SHIFTED TO 5 LSBS
256 x 10
RAM
(RED LUT)
LOCATION "31"
LOCATION "0"
256 x 10
RAM
(GREEN LUT)
LOCATION "31"
LOCATION "0"
256 x 10
RAM
(BLUE LUT)
LOCATION "31"
LOCATION "0"
DATA LATCHES FIRST 32 LOCATIONS OF RAM
10
TO RED DAC
10
TO GREEN DAC
10
TO BLUE DAC
Figure 25. 15-Bit True-Color Mapping Using R3–R7, G3–G7 and B3–B7
This mode allows for the display of 15-bit True-Color Images.
PIXEL PORT MAPPING
The pixel data to the ADV7150 is automatically mapped in the parts pixel port as determined by the pixel data mode pro­grammed (Bits CR24–CR27 of Command Register 2).
Pixel data in the 24-bit True-Color modes is directly mapped to the 24 color inputs R0–R7, G0–G7 and B0–B7.
There are three modes of operation for 8-bit Pseudo Color. Each mode maps the input pixel data differently. Data can be input one of the three color channels, R0–R7 or G0–G7 or B0–B7.
–16–
REV. A
ADV7150
R7 R6 R5 R4 R3 R2 R1 R0
G7 G6 G5 G4 G3 G2 G1 G0
B7 B6 B5 B4 B3 B2 B1
B0
PIN ASSIGN­MENTS
R4 R3 R2 R1 R0 G4 G3 G2
x G1 G0 G4 B4 B3 B2
B1
x x x x x x x x
DATA LATCHED TO PIXEL PORT
0 0
0 R4 R3
5
R2 R1 R0
0
0
0
G4 G3
5
G2
G1 G0
0
0
0 B4 B3
5
B2 B1 B0
DATA INTERNALLY SHIFTED TO 5 LSBS
256 x 10
RAM
(RED LUT)
LOCATION "31"
LOCATION "0"
256 x 10
RAM
(GREEN LUT)
LOCATION "31"
LOCATION "0"
256 x 10
RAM
(BLUE LUT)
LOCATION "31"
LOCATION "0"
DATA LATCHES FIRST 32 LOCATIONS OF RAM
10
TO RED DAC
10
TO GREEN DAC
10
TO BLUE DAC
R4 R3 R2 R1 R0 G4 G3 G2
x G1 G0 B4 B3 B2 B1
B0
x
x
x
x
x
x
x
x
PIXEL INPUT DATA
Figure 26. 15-Bit True-Color Mapping Using R0–R7 and G0–G6
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 (R3–R7, G3–G7 and B3–B7) and is internally mapped to locations 0 to 31 of the Look-Up Table (LUT) ac­cording to Figure 25. In the second mode, data is input to the device over just two of the color ports, red and green (R0–R7 and G0–G6) and is internally mapped to LUT locations 0 to 31 according to Figure 26. (Note: Data on unused pixel inputs is ignored.)
MICROPROCESSOR (MPU) PORT
The ADV7150 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 27) consists of a bidirectional, 10-bit wide databus and interface control signals and R/
W. The 10-bit wide databus is user configurable as
illustrated.
Table II. Databus Width Table
Databus RAM/DAC Read/Write Width Resolution Mode
10-Bit 10-Bit 10-Bit Parallel 10-Bit 8-Bit 8-Bit Parallel 8-Bit 10-Bit 8+2 Byte 8-Bit 8-Bit 8-Bit Parallel
Register Mapping
The ADV7150 contains a number of onboard registers includ­ing the Mode Register (MR17–MR10), Address Register (A7– A0) and nine Control Registers as well as Red (R9–R0), Green (G9–G0) and Blue (B9–B0) Color Registers. These registers control the entire operation of the part. Figure 28 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 (A7–A0). If C1, C0 = 0, 1 the MPU has access to the Look-Up Table RAM (Color Palette) through the associated color regis­ters. The
The R/
CE input latches data to or from the part.
W control input determines between read or write ac-
cesses. The Truth Tables III and IV show all modes of access to the various registers and color palette for both the 8-bit wide databus configuration and 10-bit wide databus 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).
Color Palette Accesses
Data is written to the color palette by first 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). 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 blue are concatenated into a single 30-bit/24-bit word and written to the RAM location as specified in the ad­dress register (A7–A0). 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 00H following a blue write cycle to color pal­ette RAM location FFH.
CE, C0, C1
REV. A
–17–
ADV7150
ADDRESS
REGISTER
ADDR
(A7–A0)
CE
CONTROL REGISTERS
R/W
MODE
REGISTER
(MR1)
C0 C1
PIXEL MASK
REGISTER
TEST
REGISTERS
ID
REGISTER
MPU PORT
10 (8+2)
D9 – D0
COMMAND
REGISTERS
(CR1–CR3)
REVISION
REGISTER
DATA TO
PALETTES
REGISTER
Figure 27. MPU Port and Register Configuration
RED
30
COLOR REGISTERS
GREEN
REGISTER
BLUE
REGISTER
Data is read from the color palette by first 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. 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 00H following a blue read cycle of color palette RAM location FFH.
MODE REGISTER
(MR17–MR10)
C1 = 1 C0 = 0
ADDRESS
REGISTER
(A15–A0)
00H
01H
02H
03H 04H 05H 06H 07H 08H 09H 0AH 0BH
* THIS REGISTER IS READ ONLY. A READ CYCLE WILL RETURN ZEROS "00".
CONTROL
REGISTERS
PIXEL TEST REGISTER R G B DAC TEST REGISTER R G B
SYNC, BLANK & I TEST REGISTER
ID REGISTER (READ ONLY) PIXEL MASK REGISTER COMMAND REGISTER 1 COMMAND REGISTER 2 COMMAND REGISTER 3 RESERVED* (READ ONLY) RESERVED* (READ ONLY) RESERVED* (READ ONLY) REVISION REGISTER
PLL
ADDRESS REGISTER
(A7–A0)
Register Accesses
The MPU can write to or read from all of the ADV7150s regis­ters. C0 and C1 determine whether the Mode Register or Ad­dress Register is being accessed. Access to these registers is direct. The Control Registers are accessed indirectly. The Address Register must point to the desired Control Register. Figure 28 along with the 8-bit and 10-bit Interface Truth Tables illustrate the structure and protocol for device communication over the MPU port.
C1 = 1 C0 = 1
C1 = 0 C0 = 0
C1 = 0 C0 = 1
RED
REGISTER
(R9–R0)
POINTS TO LOCATION CORRESPONDING TO ADDRESS REG (A7–A0)
GREEN
REGISTER
(G9–G0)
LOOK-UP TABLE RAM
(256 x 30)
ADDRESS REG = ADDRESS REG + 1
BLUE
REGISTER
(B9–B0)
Figure 28. Internal Register Configuration and Address Decoding
–18–
REV. A
ADV7150
Table III. Interface Truth Table (10-Bit Databus Mode)
R/W C1 C0 Databus (D9–D0) Operation Result
0 1 1 DB7–DB0 Write to Mode Register DB7–DB0 MR17–MR10 0 0 0 DB7–DB0 Write to Address Register DB7–DB0 A7–A0 0 1 0 DB7–DB0 Write to Control Registers DB7–DB0 Control Register
(Particular Control Register Determined by Address Register)
0 0 1 DB9–DB0 Write to RED Register DB9–DB0 R9–R0 0 0 1 DB9–DB0 Write to GREEN Register DB9–DB0 G9–G0 0 0 1 DB9–DB0 Write to BLUE Register DB9–DB0 B9–B0
Write RGB Data to RAM Location Pointed to by Address Register (A7–A0) Address Register = Address Register + 1
1 1 1 DB7–DB0 Read Mode Register MR17–MR10 DB7–DB0 1 0 0 DB7–DB0 Read Address Register A7–A0 DB7–DB0 1 1 0 DB7–DB0 Read Control Registers Register Data DB7–DB0
(Particular Control Register Determined by Address Register)
1 0 1 DB9–DB0 Read RED RAM Location R9–R0 DB9–DB0 1 0 1 DB9–DB0 Read GREEN RAM Location G9–G0 DB9–DB0 1 0 1 DB9–DB0 Read BLUE RAM Location B9–B0 DB9–DB0
(RAM Location Pointed to by Address Register(A7–A0))
Address Register = Address Register + 1
DB = Data Bit.
Table IV. Interface Truth Table (8-Bit Databus Mode)*
R/W C1 C0 Databus (D7–D0) Operation Result
0 1 1 DB7–DB0 Write to Mode Register DB7–DB0 MR17–MR10 0 0 0 DB7–DB0 Write to Address Register DB7–DB0 A7–A0 0 1 0 DB7–DB0 Write to Control Registers DB7–DB0 Control Registers
(Particular Control Register Determined by Address Register (A7–A0))
0 0 1 DB9–DB2 Write to RED Register DB9–DB2 R9–R2 0 0 1 DB1–DB0 Write to RED Register DB1–DB0 R1–R0 0 0 1 DB9–DB2 Write to GREEN Register DB9–DB2 G9–G2 0 0 1 DB1–DB0 Write to GREEN Register DB1–DB0 G1–G0 0 0 1 DB9–DB2 Write to BLUE Register DB9–DB2 B9–B2 0 0 1 DB1–DB0 Write to BLUE Register DB1–DB0 B1–B0
Write RGB Data to RAM Location Pointed to by Address Register (A7-A0) Address Register = Address Register + 1
1 1 1 DB7–DB0 Read Mode Register MR17–MR10 DB7–DB0 1 0 0 DB7–DB0 Read Address Register A7–A0 DB7–DB0 1 1 0 DB7–DB0 Read Control Registers Register Data DB7–DB0
(Particular Control Register Determined by Address Register)
1 0 1 DB9–DB2 Read RED RAM Location R9–R2 DB9–DB2 1 0 1 DB1–DB0 Read RED RAM Location R1–R0 DB1–DB0 1 0 1 DB9–DB2 Read GREEN RAM Location G9–G2 DB9–DB2 1 0 1 DB1–DB0 Read GREEN RAM Location G1–G0 DB1–DB0 1 0 1 DB9–DB2 Read BLUE RAM Location B9–B2 DB9–DB2 1 0 1 DB1–DB0 Read BLUE RAM Location B1–B0 DB1–DB0
(RAM Location Pointed to by Address Register (A7–A0))
Address Register = Address Register + 1
*Writing or reading 10-bit data (DB9–DB0) over an 8-bit databus (D7–D0) requires two write or two read cycles.
:DB9–DB2 is mapped to D7–D0 on the first cycle. :DB1–DB0 is mapped to D1–D0 on the second cycle.
DB = Data Bit.
REV. A
–19–
ADV7150
Power-On Reset
On power-up of the ADV7150 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) and Command Register 3 (CR37–CR30) have all bits set to a Logic “1.” Command Regis­ter 1 (CR17–CR10) has all bits set to a Logic “0.”
The output clocking signals are also set during this reset period.
PRGCKOUT = CLOCK/32 LOADOUT = CLOCK/4
The power-on reset is activated when V
goes from 0 V to
AA
5 V. This reset is active for 1 µs. The ADV7150 should not be accessed during this reset period. The pixel clock should be applied at power-up.
MR19
RESERVED*
* THESE BITS ARE READ-ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00."
MR18 MR15 MR14 MR13 MR12
PALETTE SELECT
MATCH BITS CONTROL
MR16 PS0
MR17 PS1
MR17 MR16
CALIBRATE
LOADIN
MR15
OPERATIONAL MODE CONTROL
MR14 MR13
0 0 RESERVED 0 1 NORMAL OPERATION 1 0 RESERVED 1 1 RESERVED
REGISTER PROGRAMMING
The following section describes each register, including Address Register, Mode Register and each of the nine Control Registers in terms of its configuration.
Address Register (A7–A0)
As illustrated in the previous tables, the C0 and C1 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 8-bits wide and can be read from as well as written to. 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.
MR11 MR10
MPU DATA BUS WIDTH
MR12
0 8-BIT (D7–D0) 1 10-BIT (D9–D0)
RAM-DAC
RESOLUTION CONTROL
MR11
0 8-BIT 1 10-BIT
RESET CONTROL
MR10
Mode Register 1 (MR1) (MR19–MR10)
MODE REGISTER MR1 (MR19–MR10)
The mode register is a 10-bit wide register. However for pro­gramming purposes, it may be considered as an 8-bit wide regis­ter (MR18 and MR19 are both reserved). It is denoted as MR17–MR10 for simplification purposes.
The diagram 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 MR18 and MR19 are read back, they are both returned as zeros.
Mode Register (MR17–MR10) Bit Description Reset Control (MR10)
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 be run through this cycle during the initialization sequence.
RAM-DAC Resolution Control (MR11)
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
–20–
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.
MPU Databus Width (MR12)
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. 10-bit data can be written to the device when configured in 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 8-bit mode.
Operational Mode Control (MR14–MR13)
When MR14 is “0” and MR13 is “1,” the part operates in normal mode.
Calibrate LOADIN (MR15)
This bit automatically calibrates the onboard LOADIN/ LOADOUT synchronization circuit. A “0” to “1” transition initiates calibration. This bit is set to “0” in normal operation. See “Pipeline Delay and Calibration” section. This bit must be run through this cycle during the initialization sequence.
REV. A
ADV7150
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
The ADV7150 has 9 control registers. To access each register, two write operations must be performed. The first write to the address register specifies which of the 9 registers is to be ac­cessed. The second access determines the value written to that particular control register.
Pixel Test Register (Address Reg (A7–A0) = 00H)
This register is used when the device is in test/diagnostic mode. It is a 24-bit (8 bits each for RED, GREEN and BLUE) wide read-only register which allows the MPU to read data on the pixel port, see “Test Diagnostic” section.
DAC Test Register (Address Reg (A7–A0) = 01H)
This register is used when the device is in test/diagnostic mode. It is a 30-bit (10 bits each for RED, GREEN and BLUE) wide read-only register which allows MPU access to the DAC port, see “Test Diagnostic” section.
SYNC, BLANK and I
Test Register
PLL
(Address Reg (A7–A0) = 02H)
This register is used when the device is in test/diagnostic mode. It is a 3-bit wide (3 LSBs) read/write register which allows MPU access to these particular pixel control bits, see “Test Diagnos­tic” section.
ID Register (Address Reg (A7–A0) = 03H)
This is an 8-bit wide “Identification” read-only register. For the ADV7150 it will always return the hexadecimal value 8EH.
Pixel Mask Register (Address Reg (A7–A0) = 04H)
The contents of the pixel mask register are individually bit-wise logically AND-ed 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 (A7–A0) = 05H)
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 (CR18 to CR19 are reserved).
The diagram below shows the various operations under the con­trol of CR1. This register can be read from as well as written to. In write mode, “0” should be written to CR11 and CR13 to CR17. In read mode, CR11 and CR13 to CR19 are returned as zeros.
COMMAND REGISTER 1-BIT DESCRIPTION Calibration Control (CR10)
This bit automatically calibrates the onboard LOADIN/ LOADOUT synchronization circuit. MR15 of Mode Register MR1 must be set to “0.”
SYNCOUT Control (CR12)
This bit specified whether the video SYNCOUT signal is to be enabled. On power up a “0” is written to the bit and “
SYNCOUT” is set three-state.
CR18CR19
RESERVED*
*THESE BITS ARE READ–ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00."
CR17 CR16 CR15 CR14 CR12 CR11 CR10CR13
CR17-CR13
(00000)
THESE BITS SHOULD BE SET TO ZERO
SYNCOUT CONTROL
CR12
0 DISABLE 1 ENABLE SYNCOUT
Command Register 1 (CR1) (CR19–CR10)
CR11
(0)
THIS BIT SHOULD BE SET TO ZERO
CALIBRATION CONTROL
CR10
0
DISABLE
1
CALIBRATES ON EVERY VERTICAL SYNC (MR15=0)
REV. A
–21–
ADV7150
COMMAND REGISTER 2 (CR2) (Address Reg (A7–A0) = 06H)
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 (CR28 and CR29 are both reserved).
The diagram shows the various operations under the control of CR2. This register can be read from as well written to. In read mode, CR28 and CR29 are both returned as zeros.
COMMAND REGISTER 2-BIT DESCRIPTION R7 Trigger Polarity Control (CR20)
This bit is used when the device is in test/diagnostic mode. It determines whether the pixel data is latched into the test regis­ters in the rising or falling edge of R7. (See “Test Diagnostics” section.)
I
Trigger Control (CR21)
PLL
This bit specifies whether the I
output is triggered from
PLL
BLANK or SYNC.
SYNC Recognition Control (CR22)
This bit specifies whether the video SYNC input is to be encoded onto the IOG analog output or ignored.
Pedestal Enable Control (CR23)
This bit specifies whether a 0 IRE or a 7.5 IRE blanking pedes­tal is to be generated on the video outputs.
True-Color/Pseudo-Color Mode Control (CR27–CR24)
These 4 bits specify the various color modes. These include a 24-bit true-color mode, two 15-bit true-color modes and three 8-bit pseudo color modes.
CR29 CR28
RESERVED*
*THESE BITS ARE READ­ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00."
TRUE COLOR/PSEUDO-COLOR MODE CONTROL
CR27 CR26 CR24CR25
0
0
1 1100
1101
00
1000 000
CR26 CR25 CR24CR27
PEDESTAL ENABLE
CONTROL
CR23
0  0 IRE 1  7.5 IRE
MODE
8-BIT PSEUDO COLOR ON R7–R0 8-BIT PSEUDO COLOR ON G7–G0 8-BIT PSEUDO COLOR ON B7–B0 15-BIT TRUE COLOR ON R7–R3, G7–G3, B7–B3 15-BIT TRUE COLOR ON R7–R0, G6–G0
0111
24-BIT TRUE COLOR R7–R0, G7–G0, B7–B0
SYNC RECOGNITION CONTROL
CR22
0  IGNORE 1  DECODE
I
TRIGGER CONTROL
PLL
CR21
0
1
Command Register 2 (CR2) (CR29–CR20)
CR21CR23 CR22
SYNC
BLANK
R7 TRIGGER POLARITY CONTROL
CR20
0
1
CR20
–22–
REV. A
ADV7150
COMMAND REGISTER 3 (CR3) (Address Reg (A7–A0) = 07H)
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 (CR38 and CR39 are both reserved).
The diagram shows the various operations under the control of CR3. This register can be read from as well written to. In read mode, CR38 and CR39 are both returned as zeros.
COMMAND REGISTER 3-BIT DESCRIPTION PRGCKOUT Frequency Control (CR31–CR30)
These bits specify the output frequency of the PRGCKOUT output. PRGCKOUT is a divided down version of the pixel CLOCK.
CR39 CR38 CR37 CR36 CR35 CR34 CR32 CR31 CR30CR33
RESERVED*
*THESE BITS ARE READ­ONLY RESERVED BITS. A READ CYCLE WILL RETURN ZEROS "00."
EXTRA BLANK PIPELINE DELAY CONTROL (ADDS TO PIXEL PIPELINE DELAY; tPD)
CR35 CR34 CR33 CR32
0
0
0
0
0
0
·
·
·
·
1
1
BLANK Pipeline Delay Control (CR35–CR32)
These bits specify the additional pipeline delay that can be added to the pipeline delay (t
BLANK function, relative to the overall device
). As the BLANK control normally enters the
PD
video DAC from a shorter pipeline than the video pixel data, this control is useful in deskewing 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.
Revision Register (Address Reg (A7–A0) = 0BH)
This register is a read only register containing the revision of silicon.
0 0 1
·
·
1
t
0 1 0
·
·
1
PD
t
+ 1 x LOADOUT
PD
t
+ 2 x LOADOUT
PD
t
+ 15 x LOADOUT
PD
·
·
PIXEL MULTIPLEX CONTROL
CR37 CR36
1:1 MUXING: LOADOUT = CLOCK ÷ 1
0
0 0 1 1
2:1 MUXING LOADOUT = CLOCK ÷ 2
1
RESERVED
0
4:1 MUXING :LOADOUT = CLOCK÷ 4
1
Command Register 3 (CR3) (CR39–CR30)
PRGCKOUT FREQUENCY
CONTROL
CR31 CR30
0 1 0 1
CLOCK ÷ 4
CLOCK ÷ 8
CLOCK ÷ 16 CLOCK ÷ 32
0 0 1
1
REV. A
–23–
ADV7150
DIGITAL-TO-ANALOG CONVERTERS (DACS) AND VIDEO OUTPUTS
The ADV7150 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). Other analog signals on the part include I complementary video outputs
IOR, IOG, IOB. These comple­mentary outputs can be used to drive differentially terminated video loads, they will have equal but opposite output levels to IOR, IOG and IOB when loaded with a resistive load similar to IOR, IOG and IOB.
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 = “l”) or IOR, IOG, IOB (bit = “0”). (Normally IOR, IOG, IOB = 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 load (doubly terminated 75 ).
IOR, IOG, IOB
DACs
ZS = 75
(SOURCE
TERMINATION)
ZO = 75
(CABLE)
Figure 29. DAC Output Termination (Doubly Terminated 75
Load)
PLL
and V
as well as
REF
ZL = 75 (MONITOR)
Reference Input and R
SET
An external 1.23 V voltage reference is required to drive the analog outputs of the ADV7150. The reference voltage is con­nected to the V
A resistor R and ground. For specified performance, R
input.
REF
is connected between the R
SET
input of the part
SET
has a value of
SET
280 . This corresponds to the generation of RS-343A video levels (with
SYNC on IOG and Pedestal = 7.5 IRE) into a dou­bly terminated 75 load. Figure 30 illustrates the resulting video waveform, and the Video Output Truth Table shows the corresponding control input stimuli.
IOR, IOB IOG mA mA VV
1.44 0.054 9.05 0.340
00
26.67 1.0000.71419.05
0.286
7.62
00
92.5 IRE
7.5 IRE
40 IRE
GRAY SCALE
Figure 30. Composite Video Waveform ( on IOG; Pedestal = 7.5 IRE; R
= 280 Ω)
SET
SYNC
WHITE LEVEL
BLACK LEVEL
BLANK LEVEL
SYNC LEVEL
Decoded
Variations on RS-343A
Various other video output configurations can be implemented by the ADV7150, including RS-170. Values of R
for particu-
SET
lar output video formats/levels are calculated by using the equa­tions for R table shows calculated values of R
given in the “Pin Configuration” section. The
SET
for some of the most com-
SET
mon variants on the RS-343A standard. The associated wave­forms are shown in the diagrams.
Table V. Video Output Truth Table
IOG IOR, IOB DAC
Description (mA) (mA) SYNC BLANK Input Data
WHITE LEVEL 26.67 19.05 1 1 3FFH VIDEO Video + 9.05 Video + 1.44 1 1 Data VIDEO to
BLANK Video + 1.44 Video + 1.44 0 1 Data BLACK LEVEL 9.05 1.44 1 1 000H BLACK to
BLANK 1.44 1.44 0 1 000H BLANK LEVEL 7.62 0 1 0 xxxH SYNC LEVEL 0 0 0 0 xxxH
Decoded on IOG; Pedestal = 0 IRE; R
= 265 .
SET
–24–
REV. A
ADV7150
IOR, IOB IOG mA mA VV
00
26.67 1.0000.69818.62
0.302
8.05
00
100 IRE
GRAY SCALE
43 IRE
Figure 31. Composite Video Waveform SYNC
IOR, IOB, IOG
mA V
19.05 0.714
0.054
1.44
00
92.5 IRE
GRAY SCALE
7.5 IRE
WHITE LEVEL
BLACK LEVEL
BLANK LEVEL
Figure 32. Composite Video Waveform (Pedestal = 7.5 IRE; R
IOR, IOB, IOG
mA V
19.05 0.714
= 280 Ω)
SET
WHITE LEVEL
WHITE LEVEL
BLACK/ BLANK LEVEL
SYNC LEVEL
R
(V) Video Signal
SET
265 280 No
SYNC decoded on IOG; Pedestal = 0 IRE
SYNC decoded; Pedestal = 7.5 IRE
259 No SYNC decoded; Pedestal = 0 IRE
I
Synchronization Output Control
PLL
This output synchronization signal is used in applications where it is necessary to synchronize multiple palette devices (ADV7150 + ADV7151) to subpixel resolution. Each devices I
PLL
output signal is in phase with its analog RGB output signal. If multiple devices have differing output delays, the time difference can be derived from the I to phase shift the
signals. This time difference is then used
PLL
CLOCK inputs on one or other of the devices
inputs. The I
of
signal is internally triggered by either the falling edge
PLL
SYNC or BLANK as determined by CR21 of Command
Register 2.
100 IRE
GRAY SCALE
00
Figure 33. Composite Video Waveform (Pedestal = 0 IRE; R
= 259 Ω)
SET
BLACK/ BLANK LEVEL
REV. A
–25–
ADV7150
APPENDIX 1
BOARD DESIGN AND LAYOUT CONSIDERATIONS
POWER SUPPLY DECOUPLING (0.1µF AND 0.01µF CAPACITOR FOR EACH VAA GROUP)
+5V (VAA)
0.1µF
COMP
V
AA
ADV7150
GND
0.1µF 0.01µF
ANALOG POWER PLANE
IOR
IOG
IOB
IOR
IOG
IOB
I
REF
PLL
SET
(1% METAL)
R
SET
280
75 75
V R
)
+5V (V
AA
1k
AD589 (1.2V REF)
75
COMPLIMENTARY
OUTPUTS
0.1µF
0.1µF
0.01µF
+5V (VAA)
33µF
CO-AXIAL CABLE
(75)
BNC
CONNECTORS
NOTES:
1. ALL RESISTORS ARE 1% METAL FILM
2. 0.1µF AND 0.01µF CAPACITORS ARE CERAMIC
3. ADDITIONAL DIGITALCIRCUITRY OMITTED FOR CLARITY
0.01µF
0.1µF
L1
(FERRITE BEAD)
75
75
75
MONITOR
(CRT)
0.1µF
+5V (VCC)
0.01µF
0.1µF
Recommended Analog Circuit Layout
The ADV7150 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 ADV7150 power and ground lines by shielding the digital inputs and pro­viding good decoupling. The lead length between groups of V
AA
and GND pins should by minimized so as to minimize inductive ringing.
Ground Planes
The ground plane should encompass all ADV7150 ground pins, voltage reference circuitry, power supply bypass circuitry for the ADV7150, the analog output traces, and all the digital signal traces leading up to the ADV7150. The ground plane is the graphics board’s common ground plane.
Power Planes
The ADV7150 and any associated analog circuitry should have its own power plane, referred to as the analog power plane (V
AA
).
This power plane should be connected to the regular PCB
–26–
power plane (VCC) at a single point through a ferrite bead. This bead should be located within three inches of the ADV7150.
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 ADV7150 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
For optimum performance, bypass capacitors should be installed using the shortest leads possible, consistent with reliable opera­tion, to reduce the lead inductance. Best performance is obtained with 0.1 µF ceramic capacitor decoupling. Each group of V
AA
pins on the ADV7150 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 ADV7150 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.
REV. A
ADV7150
Digital Signal Interconnect
The digital inputs to the ADV7150 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 ADV7150 should be avoided to reduce noise pickup.
Any active termination resistors for the digital inputs should be connected to the regular PCB power plane (V
), and not the
CC
analog power plane.
Analog Signal Interconnect
The ADV7150 should be located as close as possible to the out­put connectors to minimize noise pick-up 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.
APPENDIX 2
TYPICAL FRAME BUFFER INTERFACE
CLOCK
GENERATOR
CLOCK
CLOCK
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 load resistor connected to GND. These resistors should be placed as close as possible to the ADV7150 so as to minimize reflections. Normally, the differen­tial analog outputs (
IOR, IOG, IOB) are connected directly to GND. In some applications, improvements in performance are achieved by terminating these differential outputs with a resis­tive load similar in value to the video load. For a doubly termi­nated 75 load, this means that
IOR, IOG, IOB are each
terminated with 37.5 resistors.
ECL
TO
TTL
CLOCK
GRAPHICS PROCESSOR/ CONTROLLER
FRAME
BUFFER/
VIDEO
MEMORY
BLANK
SYNC
VRAM
(BANK A)
VRAM
(BANK B)
VRAM
(BANK C)
VRAM
(BANK D)
33MHz
33MHz
33MHz
33MHz
PRGCKOUT
LOADOUT
SCKOUT
BLANK
SCKIN
LOADIN
DIVIDE BY N
(÷ N)
DIVIDE BY M
LATCH
ENABLESYNC
(÷ M)
ADV7150
24
24
24
24
24
24
MULTIPLEXER
24
24
24 TO
PALETTE/RAM & DAC
REV. A
–27–
ADV7150
APPENDIX 3
10-BIT DACS AND GAMMA CORRECTION
10-Bit DACs
10-Bit RAM-DAC resolution allows for nonlinear video correc­tion, in particular Gamma Correction. The ADV7150 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 (I
) produced on a CRT by an input value IIN is given by:
CRT
I
CRT
= (IIN)
χ
where χ ranges from 2.0 to 2.8. If the individual values of χ for red, green and blue are known,
then so called “Gamma Correction” can be applied to each of the three video input signals (I
);
IN
therefore:
I
IN(corrected)
= k(IIN)
1/χ
(k = 1, normally)
Traditionally, there has been a tradeoff between implementing a nonlinear graphics function, such as gamma correction, and color dynamic range. The ADV7150 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).
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 ADV7150, all changes on the input 8-bit data are reflected in corresponding changes in the 10-bit data.
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 lookup 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.
Other applications of the 10-bit RAM-DAC include closed-loop monitor color calibration.
Gamma Correction 8 Bits vs. 10 Bits
Gamma Corrected Quantized to Quantized to
8-Bit Data (2.7) 8 Bits 10 Bits
240 0.977797 250 1001 241 0.979304 250 1002 242 0.980807 251 1004 243 0.982306 251 1005 244 0.983801 251 1007 245 0.985292 252 1008 246 0.986780 252 1010 247 0.988264 252 1011 248 0.989744 253 1013 249 0.991220 253 1015 250 0.992693 254 1016 251 0.994161 254 1018 252 0.995626 254 1019 253 0.997088 255 1021 254 0.998546 255 1022 255 1.000000 255 1023
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
DAC OUTPUT – Normalized to 1
0.20
0.10
0.00 0
GAMMA CORRECTION CURVE
LINEAR RESPONSE PRECEIVED BY THE EYE
CRT RESPONSE
32
INPUT CODE – Decimal
2241921601289664
Gamma Correction Curve (Gamma Value = 2.7)
256
–28–
REV. A
APPENDIX 4
DACs
IOR, IOG, IOB
ZO = 75
ZS = 75
(SOURCE
TERMINATION)
ZL = 75 (MONITOR)
(CABLE)
DACs
IOR, IOG, IOB
(DEVICE: 2)
(DEVICE: 1)
MULTIPLE PALETTE APPLICATIONS
ADV7150
Palette Priority Select Inputs
The palette priority selection inputs allow up to four separate palette devices to be used in a single system to drive a single monitor with subpixel resolution. The IOR, IOG and IOB ana­log video output signals of each device are connected together, as shown. Signal inputs (PS0, PS1) determine on a pixel by pixel basis which palette device drives the monitor. This allows for implementation of multiple windows applications with each device acting as an independent palette. During initialization, each device is assigned two match bits, MR16 (PS0) and MR17 (PS1) in Mode Register MR1. PS0 and PS1 inputs will select one of the preprogrammed devices at any instant when PS0, PS1 matches MR16, MR17, respectively. PS0 and PS1 are multi­plexed similar to the pixel data, thus allowing for subpixel resolu­tion. The diagrams show an example of one ADV7150 operating in
ADV7150
R0–R7 G0–G7 B0–B7
PS0, PS1
256 x 30
RAM
PALETTE SELECT BITS
MR16 MR17
0 0
ANALOG
O/P
conjunction with three ADV7151’s (Pseudo-Color RAM-DACs). Each displayed window on the monitor is driven by one of the four devices, as determined on a pixel basis by PS0, PS1. Each device’s analog output signals are connected together as shown.
Note: Only one palette device is selected at any particular instant. The analog output levels of the unselected devices will be 0 mA.
Other applications for the palette priority function using a mini­mum of two devices (one ADV7150 and one ADV7151) include:
Cursor Overlay on 24-Bit Graphics Active Live Video Overlay (from Frame Grabber) Text/Character Generation and Overlay
Multiple Devices Termination for a Single Monitor
REV. A
P0–P7
ADV7151 (1)
256 x 30
PALETTE
PALETTE SELECT BITS
MR16 MR17
0 1
ADV7151 (2)
256 x 30
PALETTE
PALETTE SELECT BITS
MR16 MR17
1 0
ADV7151 (3)
256 x 30
PALETTE
PALETTE SELECT BITS
MR16 MR17
1 1
Multiple Devices Driving a Multiwindow Application
RGB
ANALOG
VIDEO
RGB
ANALOG
VIDEO
RGB
ANALOG
VIDEO
VIDEO TO MONITOR
–29–
MONITOR
TRUE-COLOR BACKGROUND
WINDOW 1
(Pseudo-Color)
PS0=0: PS1=1
WINDOW 2
(Pseudo-Color)
PS0=1: PS1=0
WINDOW 3
(Pseudo-Color)
PS0=1: PS1=1
ADV7150
APPENDIX 5
INITIALIZATION AND PROGRAMMING
ADV7150 Initialization
After power has been supplied, the ADV7150 must be initial­ized. The Mode Register and Control Registers must be set. The values written to the various registers will be determined by the desired operating mode of the part, i.e., True Color/Pseudo Color, 2:1 Muxing/2:1 Muxing, etc.
Example 1
Color Mode 24-Bit True Color Multiplexing 2:1 Databus 8-Bit RAM-DAC Resolution 8-Bit SYNC Enabled on IOG Pedestal 7.5 IRE
Register Initialization C1 C0 R/
Write 09H to Mode Register (MR1) 1 1 0 Resets to Normal Operation, 8-Bit Bus/RAM-DAC Write 08H to Mode Register (MR1) 1 1 0 *(Initializes Pipelining Write 09H to Mode Register (MR1) 1 1 0 *( “ Write 29H to Mode Register (MR1) 1 1 0 *(Calibrates LOADOUT/LOADIN Timing Write 09H to Mode Register (MR1) 1 1 0 *( “ Write 04H to Address Register (A7–A0) 0 0 0 Address Reg Points to Pixel Mask Register Write FFH to Pixel Mask Register 1 0 0 Sets the Pixel Mask to All “1s” Write 05H to Address Register (A7–A0) 0 0 0 Address Reg Points to Command Register 1 (CR1) Write 00H to Command Reg 1 (CR1) 1 0 0 Write 06H to Address Register (A7–A0) 0 0 0 Address Reg Points to Command Register 2 (CR2) Write ECH to Command Reg 2 (CR2) 1 0 0 Sets 24-Bit Color, 7.5 IRE, Write 07H to Address Register (A7–A0) 0 0 0 Address Reg Points to Command Register 3 (CR3) Write C0H to Command Reg 3 (CR3) 1 0 0 Sets 2:1 Multiplexing, PRGCKOUT = CLOCK/4
Color Palette RAM Initialization C1 C0 R/W Comment
Write 00H to Address Register (A7–A0) 0 0 0 Points to Color Palette RAM Write 00H (Red Data) to RAM Location (00H) 0 1 0 (Initializes Palette RAM Write 00H (Green Data) to RAM Location (00H) 0 1 0 ( to a Linear Ramp** Write 00H (Blue Data) to RAM Location (00H) 0 1 0 ( Write 01H (Red Data) to RAM Location (01H) 0 1 0 ( Write 01H (Green Data) to RAM Location (01H) 0 1 0 ( Write 01H (Blue Data) to RAM Location (01H) 0 1 0 (
••• • ••(
••• • ••( Write FFH (Red Data) to RAM Location (FFH) 0 1 0 ( Write FFH (Green Data) to RAM Location (FFH) 0 1 0 ( Write FFH (Blue Data) to RAM Location (FFH) 0 1 0 (RAM Initialization Complete
The following section gives examples of initialization of the ADV7150 operating in various modes.
W Comment
SYNC on Green (IOG)
**These four command lines reset the ADV7150. The pipelines for each of the Red, Creen and 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.” LOADIN/LOADOUT timing is internally synchronized by writing a “0” followed by a “1” followed by a “0” to Mode Register MR15.
**This sequence of instructions would, of course, normally be coded using some form of loop instruction.
–30–
REV. A
ADV7150
Example 2
Color Mode 24-Bit Gamma Corrected True Color (30 Bits) Multiplexing 2:1 Databus 10 Bit RAM-DAC Resolution 10 Bit SYNC Ignored Pedestal 0 IRE Calibration Every Vertical Sync
Register Initialization C1 C0 R/W Comment
Write 0FH to Mode Register (MR1) 1 1 0 Resets to Normal Operation, 10-Bit Bus/RAM-DAC Write 0EH to Mode Register (MR1) 1 1 0 *(Initializes Pipelining Write 0FH to Mode Register (MR1) 1 1 0 *( “ Write 2FH to Mode Register (MR1) 1 1 0 *(Calibrates LOADOUT/LOADIN Timing Write 0FH to Mode Register (MR1) 1 1 0 *( “ Write 04H to Address Register (A7–A0) 0 0 0 Address Reg Points to Pixel Mask Register Write FFH to Pixel Mask Register 1 0 0 Sets the Pixel Mask to All “1s” Write 05H to Address Register (A7–A0) 0 0 0 Address Reg Points to Command Register 1 (CR1) Write 01H to Command Reg 1 (CR1) 0 0 0 Calibrates Every Vertical Sync Write 06H to Address Register (A7–A0) 0 0 0 Address Reg Points to Command Register 2 (CR2) Write E0H to Command Reg 2 (CR2) 1 0 0 Sets 24-Bit Color, 0 IRE, No Write 07H to Address Register (A7–A0) 0 0 0 Address Reg Points to Command Register 3 (CR3) Write 41H to Command Reg 3 (CR3) 1 0 0 Sets 2:1 Multiplexing, PRGCKOUT =
Color Palette RAM Initialization C1 C0 R/
W Comment
Write 00H to Address Register (A7–A0) 0 0 0 Points to Color Palette RAM Write 000H (Red Data) to RAM Location (00H) 0 1 0 (Initializes Palette RAM Write 000H (Green Data) to RAM Location (00H) 0 1 0 ( to a “Gamma” Ramp** Write 000H (Blue Data) to RAM Location (00H) 0 1 0 ( Write xxxH (Red Data) to RAM Location (01H) 0 1 0 ( Write xxxH (Green Data) to RAM Location (01H) 0 1 0 ( Write xxxH (Blue Data) to RAM Location (01H) 0 1 0 (
••• • •••(
••• • •••( Write 3FFH (Red Data) to RAM Location (FFH) 0 1 0 ( Write 3FFH (Green Data) to RAM Location (FFH) 0 1 0 ( Write 3FFH (Blue Data) to RAM Location (FFH) 0 1 0 (RAM Initialization Complete
**These four command lines reset the ADV7150 The pipelines for each of the Red, Green and Blue pixel inputs are synchronously reset to the Multiplexer’s “A” in-
put. Mode Register bit MR10 is written by a “1” followed by “0” followed by “1.” LOADIN/LOADOUT timing is internally synchronized by writing a “0” followed by a “1” followed by a “0” to Mode Register MR15.
**Data for a gamma curve characteristic is obtainable in Appendix 3.
REGISTER DIAGNOSTIC TESTING
The previous examples show the register initialization sequence for the ADV7150. These show control data going to the regis­ters and palette RAM. As well as this writing function, it may also be necessary, due to system diagnostic requirements, to confirm that correct data has been transferred to each register and palette RAM location. There are two ways to incorporate register value/RAM value checking:
1. READ after each WRITE: After data is written to a particular register, it can be read back immediately. The following table shows an example with Command Registers CR2 and CR3.
C1 C0 R/W D0–D7 Comment
0 0 0 06H Select Command Register 2 (CR2) 1 0 0 E0H Sets 24-Bit True-Color 1 0 1 E0H Command Reg 2 Value Read-Back 0 0 0 07H Select Command Register 3 (CR3) 1 0 0 40H Set 2:1 Mux Mode 1 0 1 40H Command Reg 3 Value Read-Back
2. READ after all WRITEs completed: All registers and the color palette RAM are written to and set. Once this is complete, all registers are again accessed but this time in Read-Only mode. The table below shows this method for Command Registers CR2 and CR3.
C1 C0 R/W D0–D7 Comment
0 0 0 06H Select Command Register 2 (CR2) 1 0 0 E0H Sets 24-Bit True-Color 0 0 0 07H Select Command Register 3 (CR3) 1 0 0 40H Set 2:1 Mux Mode 0 0 0 06H Select CR2 1 0 1 E0H CR2 Value Read-Back 0 0 0 07H Select CR3 1 0 1 40H CR3 Value Read-Back 1 0 1 40H CR3 Value Read-Back
It is clear that this latter case requires more command lines than the previous READ after each WRITE case.
SYNC
CLOCK/8
REV. A
–31–
ADV7150
APPENDIX 6
TEST DIAGNOSTICS
SYNC
BLANK
PIXEL
DATA
INPUT
MUX
GRAPHICS PIPELINE GRAPHICS PIPELINE
TRIGGER
DECODE
PIXEL TEST
REGISTER
CE R/W C0 C1 D0–D9
COLOR
PALETTE
RAM
COLOR
REGISTERS
Test/Diagnostic Block Diagram
The ADV7150 contains onboard circuitry which enables both device and system level test diagnostics. The test circuitry can be used to test the frame buffer memory as well as the function­ality of the ADV7150. A number of test registers are integrated into the part which effectively allow for monitoring of the graph­ics pipeline. Pixel data is read from the graphics pipeline inde­pendent of the pixel CLOCK. The pixel data itself contains the triggering information that latches data into the test registers. This allows for system diagnostics in a continuously clocked graphics system. The test register data is then read by the micro­processor over the MPU.
Access to the test registers is as described in the “Microproces­sor (MPU) Port” section. This section also gives the address decode locations for the various test registers.
Test Trigger (R7)
The test trigger is decoded from the pixel data stream. Bit R7 of the RED channel is assigned the task of latching pixel data into the test registers. A “0” to “1” or a “1” to “0” (as determined by bit CR20 of Command Register 2) transition on R7, fills the test register with the corresponding pixel data. This effectively means that a sequence of data travels along the graphics pipe­line, with the test registers taking a sample only when there is a transition on Bit R7. The following example shows a sequence with the ADV7150 preset to sample the graphics pipeline on a low to high transition of R7.
RED GREEN BLUE
Pixel 0: 00000000 00000000 00000000
Pixel 1: 0........ ........ ........
Pixel 2: 1........ ........ ........
Pixel 3: 0........ ........ ........
.... ...
.... ...
Pixel n- l: 0........ ........ ........
Pixel n: 1........ ........ ........
Pixel n: 0........ ........ ........
In the above sequence of pixels, there is a rising edge on R7 on Pixel 2. The Red, Green and Blue data for Pixel 2, therefore, gets latched into the Pixel Test Register. Pixel 2 continues down
–32–
DACs
TRIGGER
MPU PORT
DECODE
DAC TEST
REGISTERS
SYNC
I
PLL
REGISTER
BLANK
TEST
the graphics pipeline and after a number of clocks get latched into the DAC Test Register. This data can then be read from the Pixel Test Register and the DAC Test Registers over the MPU Port. This data will remain in the Pixel Test Registers and the DAC Test Registers until the next rising edge of R7 causes new data to be latched in.
In the above example, the next rising edge of R7 occurs on the Pixel n input. Therefore the data in the Pixel Test Registers and DAC Test Registers must be read over the MPU before the Pixel n data is applied, otherwise they will be overwritten by the Pixel n data and the Pixel 2 data will be lost.
Pixel Test Register
The read-only Pixel Test Register is 24 bits wide, 8 bits each for red green and blue. It is situated directly after the Pixel Mask Register. After data is latched into this register by a transition on R7, it is read in three cycles over the MPU Port as described in the “Microprocessor (MPU) Port” section.
DAC Test Register
The DAC Test Register is latched with data some CLOCKs after the Pixel Test Register. The DAC Test Register is a 30-bit wide read-only register, corresponding to 10 bits each for red, green and blue data. It is located the Color Palette RAM. If the RAM-DAC is in 8-bit after resolution mode, the upper two bits of the red, green and blue data will be zero. After data is latched into the DAC Test Register by a transition on R7, it is read in three or six cycles over the MPU Port as described in the “Microprocessor (MPU) Port” section.
SYNC, BLANK and I
Test Register
PLL
This is an 8-bit wide register but with only three effective bits. The three lower bits correspond to
SYNC, BLANK and I
PLL
respectively. The upper bits should be masked in software. This register is at the same position in the graphics pipeline as the DAC Test Register. When pixel data is latched into the DAC Test Register, the corresponding status of I
is latched into this register. It is read over the MPU Port as
PLL
SYNC, BLANK and
described in the “Microprocessor (MPU) Port” section. (Note: If
BLANK is low, the corresponding pixel data to the
DAC Test Register will be all “0s.”)
REV. A
APPENDIX 7
THERMAL AND ENVIRONMENTAL CONSIDERATIONS
ADV7150
The ADV7150 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 in which the ADV7150 must operate. Reliability of the device is significantly 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 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 pub-
lished 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 diagram shows graphs of power dissipation in watts vs. pixel clock frequency for the ADV7150.
1.50
V
= 5V
AA
= 1.2V
V
1.25
1.00
0.75
POWER DISSIPATION – Watts
0.50
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 THE PIXEL INPUT (I. E., A BLACK TO WHITE RAMP). IN GENERAL, COLOR IMAGES TEND TO APPROXIMATE THIS CHARACTERISTIC.
REF
= +25°C
T
A
60 22080 180 200160140120100
PIXEL CLOCK FREQUENCY – MHz
Table A. Thermal Characteristics vs. Airflow
Air Velocity 0 50 100 200 (Linear feet/min) (Still Air)
θ
(°C/W)
JA
No Heatsink 25.5 23 21 19 EG&G D10100-28 Heatsink 23 20 18 16 Thermalloy 2290 Heatsink 19 17 15 12
Thermal Model
The junction temperature of the device in a specific application is given by:
T
= TA + PD (
J
θ
+
θ
JC
) (1)
CA
or
T
= TA + PD (
J
θ
) (2)
JA
where:
= Junction Temperature of Silicon (°C)
T
J
T
= Ambient Temperature (°C)
A
P
= Power Dissipation (W)
D
θ
= Junction to Case Thermal Resistance (°C/W)
JC
θ
= Case to Ambient Thermal Resistance (°C/W)
CA
θ
= Junction to Ambient Thermal Resistance (°C/W)
JA
Package Enhancements
The standard QFP package has been enhanced to a PowerQuad2 package. This supports an improved thermal performance com­pared to standard QFP. In this case, the die is attached to heatslug 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 an efficient thermal path from the ADV7150 to the Printed Circuit Board via the leads.
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 θ
as shown
JA
in the “Thermal Characteristics vs. Airflow” table.
Typical Power Dissipation vs. Pixel Rate
Package Characteristics
The table of thermal characteristics shows typical information for the ADV7150 (160-Lead Plastic Power QFP) using various values of Airflow.
Junction to Case (θ
) Thermal Resistance for this particular
JC
part is:
θ
(160-Lead Plastic Power QFP) = 1.0°C/W
JC
(Note: θ
REV. A
is independent of airflow.)
JC
–33–
ADV7150
APPENDIX 8
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
S-160
160-Lead Plastic Power Quad Flatpack
0.037 (0.95)
0.026 (0.65)
SEATING
PLANE
0.004 (0.10) MAX
0.070 (1.77)
0.062 (1.57)
0.160 (4.07)
4°±4° MAX
0.145 (3.67)
0.125 (3.17)
MAX
6°±4°
10°
0.070 (1.77)
0.062 (1.57)
121
160
120
1
1.239 (31.45)
1.219 (30.95)
1.107 (28.10)
1.100 (27.90)
PIN 1
0.026 (0.65) MIN
TOP VIEW
(PINS DOWN)
SQ
SQ
0.014 (0.35)
0.011 (0.27)
81
80
41
40
–34–
REV. A
–35–
C1695–10–8/94
–36–
PRINTED IN U.S.A.
REV. A
Loading...