AD ADV7152 Cmos 220 mhz true-color graphics triple 10-bit video ram-dac Datasheet

a
CMOS 220 MHz True-Color Graphics
Triple 10-Bit Video RAM-DAC
ADV7152
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 and 8-Bit (Pseudo)
Pixel Data Serializer
Multiplexed Pixel Input Ports; 1:1, 2:1
+5 V CMOS Monolithic Construction
100-Lead Plastic Quad Flatpack (QFP)
Thermally Enhanced to Achieve uJC < 1.08C/W
APPLICATIONS
High Resolution, True Color Graphics
Professional Color Prepress Imaging
GENERAL DESCRIPTION
The ADV7152 (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 ADV7152 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
ADV7152 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.
The device consists of three, high speed, 10-bit, video D/A converters (RGB), three 256 3 10 (one 256 3 30) color look-up
tables, palette priority selects, a pixel input data multiplexer/
serializer and a clock generator/divider circuit. The ADV7152
implements 1:1 and 2:1 pixel data multiplexing. The onboard
palette priority select inputs enable multiple palette devices to
be connected together for use in multipalette and window
(Continued on page 10)
MODES OF OPERATION
24-Bit True Color (30-Bit Gamma Corrected)
@ 220 MHz
@ 170 MHz
@ 135 MHz
@ 110 MHz
@ 85 MHz
ADV is a registered trademark of Analog Devices, Inc.
8-Bit Pseudo Color
15-Bit True Color
FUNCTIONAL BLOCK DIAGRAM
VAA
256-COLOR/GAMMA
PALETTE RAM
RED (R0–R7),
GREEN (G0–G7),
BLUE (B0–B7)
COLOR DATA
A
24
B
PALETTE
SELECTS
(PS0, PS1)
4
10
RED
256 x 10
24
P
I
X
E
L
P
O
R
T
10-BIT
RED DAC
IOR
10-BIT
GREEN DAC
IOG
10-BIT
BLUE DAC
IOB
GREEN
256 x 10
8
MUX
2:1
8
BLUE
256 x 10
10
10
2
MUX
2:1
CONTROL REGISTERS
PRGCKOUT
SCKIN
SCKOUT
CLOCK DIVIDE
&
SYNCHRONIZATION
CIRCUIT
PIXEL MASK
REGISTER
÷32, ÷16, ÷8, ÷4, ÷2
ADDRESS
REGISTER
ADDR
(A0-–A15)
IOG
IOB
SYNC
OUTPUT
CLOCK CONTROL
LOADIN
LOADOUT
IOR
8
48
4
ADV7152
COMMAND
REGISTERS
(CR1-CR3)
TEST
REGISTERS
MODE
REGISTER
(MR1)
REVISION
REGISTER
ID
REGISTER
DATA TO
PALETTES
30
RED
REGISTER
IPLL
SYNCOUT
VOLTAGE
REFERENCE
CIRCUIT
VREF
RSET
COMP
COLOR REGISTERS
GREEN
REGISTER
BLUE
REGISTER
SYNC
BLANK
MPU PORT
CLOCK
CLOCK
ECL TO CMOS
10 (8+2)
CE R/W C0
C1
D9 – D0
GND
REV. B
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.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 617/329-4700
Fax: 617/326-8703
ADV7152–SPECIFICATIONS
(VAA1 = +5 V; VREF = +1.235 V; RSET = 280 V. IOR, IOG, IOB (RL = 37.5 V,
CL = 10 pF); IOR, IOG, IOB = GND. All specifications TMIN to TMAX2 unless otherwise noted.)
Parameter
All Versions
Unit
10
Bits
±1
±1
±5
LSB max
LSB max
% Gray Scale max
Binary
DIGITAL INPUTS (Excluding CLOCK, CLOCK)
Input High Voltage, V INH
Input Low Voltage, VINL
Input Current, IIN
Input Capacitance, CIN
2
0.8
± 10
10
V min
V max
µA max
pF typ
CLOCK INPUTS (CLOCK, CLOCK)
Input High Voltage, V INH
Input Low Voltage, VINL
Input Current, IIN
Input Capacitance, CIN
VAA – 1.0
VAA – 1.6
± 10
10
V min
V max
µA max
pF typ
DIGITAL OUTPUTS
Output High Voltage, V OH
Output Low Voltage, V OL
Floating-State Leakage Current
Floating-State Output Capacitance
2.4
0.4
20
20
V min
V max
µA max
pF typ
15/22
mA min/max
17.69/20.40
16.74/18.50
0.95/1.90
0/50
6.29/8.96
0/50
17.22
3
0/+1.4
100
30
mA min/max
mA min/max
mA min/max
µA min
mA min/max
µA min/max
µA typ
% max
V min/V max
kΩ typ
pF max
Typically 19.05 mA
Typically 17.62 mA
Typically 1.44 mA
Typically 5 µA
Typically 7.62 mA
Typically 5 µA
VOLTAGE REFERENCE
Voltage Reference Range, V REF
Input Current, I VREF
1.14/1.26
+5
V min/V max
µA typ
VREF = 1.235 V for Specified Performance
POWER REQUIREMENTS
VAA
IAA3
IAA
IAA
IAA
IAA
Power Supply Rejection Ratio
5
400
370
350
330
315
0.5
V nom
mA max
mA max
mA max
mA max
mA max
%/% max
DYNAMIC PERFORMANCE
Clock and Data Feedthrough 4, 5
–30
dB typ
50
–23
pV secs typ
dB typ
STATIC PERFORMANCE
Resolution (Each DAC)
Accuracy (Each DAC)
Integral Nonlinearity
Differential Nonlinearity
Gray Scale Error
Coding
ANALOG OUTPUTS
Gray Scale Current Range
Output Current
White Level Relative to Blank
White Level Relative to Black
Black Level Relative to Blank
Blank Level on IOR, IOB
Blank Level on IOG
Sync Level on IOG
LSB Size
DAC-to-DAC Matching
Output Compliance, V OC
Output Impedance, R OUT
Output Capacitance, C OUT
Glitch Impulse
DAC-to-DAC Crosstalk6
Test Conditions/Comments
Guaranteed Monotonic
VIN = 0.4 V or 2.4 V
VIN = 0.4 V or 2.4 V
ISOURCE = 400 µA
ISINK = 3.2 mA
Typically 1%
IOUT = 0 mA
220 MHz Parts
170 MHz Parts
135 MHz Parts
110 MHz Parts
85 MHz Parts
Typically 0.12%/%, COMP = 0.1 µF
NOTES
1
± 5% for all versions.
2
Temperature range (T MIN to TMAX): 0°C to +70°C; TJ (Silicon Junction Temperature) ≤ 100°C.
3
Pixel Port is continuously clocked with data corresponding to a linear ramp. T J = 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.
–2–
REV. B
ADV7152
TIMING CHARACTERISTICS1
(VAA2 = +5 V; VREF = +1.235 V; RSET = 280 V. IOR, IOG, IOB (RL = 37.5 V, CL = 10 pF);
IOR, IOG, I0B = GND. All specifications TMIN to TMAX3 unless otherwise noted.)
CLOCK CONTROL AND PIXEL PORT 4
Parameter
fCLOCK
t1
t2
t3
t4
fLOADIN
1:1 Multiplexing
2:1 Multiplexing
t5
1:1 Multiplexing
2:1 Multiplexing
t6
1:1 Multiplexing
2:1 Multiplexing
t7
1:1 Multiplexing
2:1 Multiplexing
t8
t9
t10
τ–t115
tPD6
1:1 Multiplexing
2:1 Multiplexing
t12
t13
t14
t15
220 MHz
Version
170 MHz
Version
135 MHz
Version
110 MHz
Version
85 MHz
Version
Units
Conditions/Comments
220
4.55
2
2
10
170
5.88
2.5
2.5
10
135
7.4
3
3.2
10
110
9.1
4
4
10
85
11.77
4
4
10
MHz max
ns min
ns min
ns min
ns max
Pixel CLOCK Rate
Pixel CLOCK Cycle Time
Pixel CLOCK High Time
Pixel CLOCK Low Time
Pixel CLOCK to LOADOUT Delay
LOADIN Clocking Rate
110
110
110
85
110
67.5
110
55
85
42.5
MHz max
MHz max
9.1
9.1
9.1
11.76
9.1
14.8
9.1
18.18
11.76
23.53
ns min
ns min
4
4
4
5
4
6
4
8
4
9
ns min
ns min
4
4
0
5
0
τ–5
4
5
0
5
0
τ–5
4
6
0
5
0
τ–5
4
8
0
5
0
τ–5
4
9
0
5
0
τ–5
ns min
ns min
ns min
ns min
ns min
ns max
5
6
10
5
5
1
5
6
10
5
5
1
5
6
10
5
5
1
5
6
10
5
5
1
5
6
10
5
5
1
CLOCKs
CLOCKs
ns max
ns max
ns min
ns min
LOADIN Cycle Time
LOADIN High Time
LOADIN Low Time
Pixel Data Setup Time
Pixel Data Hold Time
LOADOUT to LOADIN Delay
LOADOUT to LOADIN Delay
Pipeline Delay
(1 × CLOCK = t1)
Pixel CLOCK to PRGCKOUT Delay
SCKIN to SCKOUT Delay
BLANK to SCKIN Setup Time
BLANK to SCKIN Hold Time
ANALOG OUTPUTS7
Parameter
t16
t17
t18
tSK
220 MHz 170 MHz 135 MHz 110 MHz 85 MHz
Version Version Version Version Version Units
15
1
15
2
0
15
1
15
2
0
15
1
15
2
0
15
1
15
2
0
15
1
15
2
0
ns typ
ns typ
ns typ
ns max
ns typ
Conditions/Comments
Analog Output Delay
Analog Output Rise/Fall Time
Analog Output Transition Time
Analog Output Skew (IOR, IOG, IOB)
MPU PORTS8, 9
Parameter
t19
t20
t21
t22
t238
t249
t259
t26
t27
REV. B
220 MHz 170 MHz 135 MHz 110 MHz 85 MHz
Version Version Version Version Version Units
3
10
45
25
5
45
20
5
20
5
3
10
45
25
5
45
20
5
20
5
3
10
45
25
5
45
20
5
20
5
3
10
45
25
5
45
20
5
20
5
3
10
45
25
5
45
20
5
20
5
–3–
ns min
ns min
ns min
ns min
ns min
ns max
ns max
ns min
ns min
ns min
Conditions/Comments
R/W, C0, C1 to CE Setup Time
R/W, C0, C1 to CE Hold Time
CE Low Time
CE High Time
CE Asserted to Databus Driven
CE Asserted to Data Valid
CE Disabled to Databus Three-Stated
Write Data (D0–D9) Setup Time
Write Data (D0–D9) Hold Time
ADV7152
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 V AA–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 PLL and
SYNCOUT ≤ 30 pF.
2
± 5% for all versions.
3
Temperature range (T MIN to TMAX): 0°C to +70°C; TJ (Silicon Junction Temperature) ≤ 100°C.
4
Pixel Port consists of the following inputs: Pixel Inputs: RED [A, B]; GREEN [A, B]; BLUE [A, B], Palette Selects: PS0 [A, B]; PS1 [A, B]; 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 multiplexing, τ = CLOCK × 2 = 2 × t1 ns.
6
These fixed values for Pipeline Delay are valid under conditions where t 10 and τ–t11 are met. If either t10 or τ–t11 are not met, the part will operate but the Pipeline Delay is increased by 2 clock cycles for 2:1 mode after calibration cycle 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% and 90% points of full-scale transition. Settling time measured from the 50% point of full-scale transition to the output remaining within ± 1 LSB. (Settling 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 extrapolated back to remove the effects of charging the 100 pF capacitor. This means that the time, t 25, 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.
ISINK
TO
OUTPUT
PIN
+2.1V
100pF
ISOURCE
Figure 1. Load Circuit for Databus Access and Relinquish Times
t2
t3
t1
CLOCK
CLOCK
t4
LOADOUT
(1:1 MULTIPLEXING)
LOADOUT
(2:1 MULTIPLEXING)
Figure 2. LOADOUT vs. Pixel Clock Input (CLOCK, CLOCK)
t6
t5
t7
LOADIN
t8
PIXEL INPUT
DATA*
VALID
DATA
t9
VALID
DATA
VALID
DATA
*INCLUDES PIXEL DATA (R0-R7, G0-G7, B0-B7); PALETTE SELECT INPUTS (PS0-PS1); SYNC; BLANK
Figure 3. LOADIN vs. Pixel Input Data
–4–
REV. B
ADV7152
CLOCK
t10
LOADOUT
LOADIN
PIXEL INPUT
DATA*
AN BN
AN+1 BN+1
DIGITAL
ANALOG
OUTPUT
DATA
INPUT TO
AN+2 BN+2
ANALOG
OUTPUT
IOR, IOR
IOG, IOG
IOB, IOB
IPLL , SYNCOUT
PIPELINE
AN-1
BN-1
AN
BN
AN+1
BN+1
AN+2
BN+2
tPD
*INCLUDES PIXEL DATA (R0–R7, G0–G7, B0–B7); PALETTE SELECT INPUTS (PS0-PS1); SYNC; BLANK
Figure 4. Pixel Input to Analog Output Pipeline with Minimum LOADOUT to LOADIN Delay (2:1 Multiplex Mode)
CLOCK
τ
τ- t11
LOADOUT
LOADIN
PIXEL INPUT
DATA*
AN+1 BN+1
AN BN
DIGIT
AL IN
ANALOG
OUTPUT
DATA
PUT T
O ANA
IOR, IOR
IOG, IOG
IOB, IOB
IPLL , SYNCOUT
LOG O
AN+2BN+2
UTPU
T PIPE
AN-1
LINE
BN-1
AN
BN
AN+1
BN+1
AN+2
BN+2
tPD
*INCLUDES PIXEL DATA (R0–R7, G0–G7, B0–B7); PALETTE SELECT INPUTS (PS0–PS1); SYNC; BLANK
Figure 5. Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (2:1 Multiplex Mode)
CLOCK
PRGCKOUT
(CLOCK/4)
PRGCKOUT
(CLOCK/8)
PRGCKOUT
(CLOCK/16)
PRGCKOUT
(CLOCK/32)
t12
*INLCUDES PIXEL DATA (R0-R7, G0-G7, B0-B7); PALETTE SELECT INPUTS (PS0-PS1); SYNC; BLANK
Figure 6. Pixel Clock Input vs. Programmable Clock Output (PRGCKOUT)
REV. B
–5–
ADV7152
t14
t13
SCKIN
t15
BLANKING PERIOD
BLANK
SCKOUT
START OF SCAN
LINE (N+1)
END OF SCAN
LINE (N)
*INCLUDES PIXEL DATA (R0-R7, G0-G7, B0-B7); PALETTE SELECT INPUTS (PS0-PS1); SYNC; BLANK
Figure 7. Video Data Shift Clock Input (SCKIN) & BLANK vs. Video Data Shift Clock Output (SCKOUT)
CLOCK
t18
t16
ANALOG
OUTPUTS
WHITE LEVEL
90 %
IOR, IOR
IOG, IOG
IOB, IOB
IPLL , SYNCOUT
50 %
FULL SCALE
TRANSITION
10 %
BLACK LEVEL
t17
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.
IPLL AND SYNCOUT ARE DIGITAL OUTPUT SIGNALS. t 16 IS THE ONLY
RELEVANT OUTPUT TIMING SPECIFICATION FOR IPLL AND SYNCOUT.
AA
AA
AAAAAAAAAAA
AA
AAAAAAAAAAA
AA
AA
AA
AAAAAA AAAAAAAA
Figure 8. Analog Output Response vs. CLOCK
t19
R/W, C0, C1
t20
VALID
CONTROL DATA
t21
CE
t22
t24
D0–D9
(READ MODE)
D0–D9
(WRITE MODE)
t25
t23
R/W = 1
R/W = 0
t26
t27
Figure 9. Microprocessor Port (MPU) Interface Timing
–6–
REV. B
ADV7152
RECOMMENDED OPERATING CONDITIONS
Parameter
Symbol
Min
Typ
Max
Units
Power Supply
Ambient Operating Temperature
Reference Voltage
Output Load
VAA
TA
VREF
RL
4.75
0
1.14
5.00
5.25
+70
1.26
Volts
°C
Volts
Ω
1.235
37.5
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 ADV7152 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
NOTES
1
Stresses above those listed under “Absolute Maximum Ratings” may cause
permanent damage to the device. This is a stress rating only and functional
operation of the device at these or any other conditions above those listed in the
operational sections of this specification is not implied. Exposure to absolute
maximum rating conditions for extended periods may affect device reliability.
2
Analog Output Short Circuit to any Power Supply or Common can be of an
indefinite duration.
ORDERING GUIDE1, 2, 3
Speed
ADV7152LS220
ADV7152LS170
ADV7152LS135
110 MHz ADV7152LS110
85 MHz ADV7152LS85
NOTES
1
ADV7152 is packaged in a 100-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.
100-Lead QFP Configuration
51
80
ROW C
ADV7152
QFP
Top View
(Not to Scale)
50
ROW B
ROW D
81
PIN NO. 1 IDENTIFIER
31
ROW A
100
1
30
Pin
No. Mnemonic
Pin
No.
Mnemonic
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
SCKOUT
SYNCOUT
GND
GND
GND
GND
G6A
G6B
G7A
G7B
PS0A
PS0B
PS1A
PS1B
B0A
B0B
B1A
B1B
B2A
B2B
B3A
B3B
B4A
B4B
B5A
B5B
B6A
B6B
B7A
B7B
CE
R/W
C0
C1
D0
D1
D2
GND
D3
D4
SYNC
BLANK
R0A
R0B
GND
R1A
R1B
R2A
R2B
R3A
R3B
R4A
R4B
R5A
R5B
R6A
R6B
R7A
R7B
G0A
G0B
G1A
G1B
G2A
G2B
NC
G3A
G3B
G4A
G4B
G5A
G5B
CLOCK
CLOCK
LOADIN
LOADOUT
VAA
VAA
PRGCKOUT
SCKIN
NC = NO CONNECT.
REV. B
ESD SENSITIVE DEVICE
PIN ASSIGNMENTS
VAA to GND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 V
Voltage on Any Digital Pin . . . . GND – 0.5 V to VAA + 0.5 V
Ambient Operating Temperature (TA) . . . . . –55°C to +125°C
Storage Temperature (TS) . . . . . . . . . . . . . . –65°C to +150°C
Junction Temperature (TJ) . . . . . . . . . . . . . . . . . . . . +150°C
Lead Temperature (Soldering, 10 secs) . . . . . . . . . . . +260°C
Vapor Phase Soldering (1 minute) . . . . . . . . . . . . . . . +220°C
Analog Outputs to GND2 . . . . . . . . . . . . . GND – 0.5 to VAA
220 MHz
170 MHz
135 MHz
WARNING!
–7–
Pin
No. Mnemonic
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
D5
VAA
D6
D7
D8
D9
GND
GND
IOB
IOR
IOG
IOB
IOG
VAA
IPLL
IOR
COMP
VREF
RSET
VAA
ADV7152
PIN FUNCTION DESCRIPTION
Mnemonic
Function
RED (R0A . . . R0B–R7A . . . R7B),
GREEN (G0A . . . G0B–G7A. . . G7B),
BLUE (B0A . . . B0B–B7A . . . B7B)
Pixel Port (TTL Compatible Inputs). 48 pixel select inputs, with 8 bits each for Red, 8
bits for Green and 8 bits for Blue. Each bit is multiplexed [A-B] 2:1 or 1:1. It can be
configured for 24-Bit True-Color Data, 8-Bit Pseudo-Color Data and 15-Bit True-Color
Data formats. Pixel Data is latched into the device on the rising edge of LOADIN.
PS0A . . . PS0B, PS1A . . . PS1B
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-B] 2:1 or 1:1. PS0 and PS1 are
latched into the device on the rising edge of LOADIN.
LOADIN
Pixel Data Load Input (TTL Compatible Input). This input latches the multiplexed
pixel data, including PS0–PS1, BLANK and SYNC into the device.
LOADOUT
Pixel Data Load Output (TTL Compatible Output). This output control signal runs at a
divided down frequency of the pixel CLOCK input. Its frequency is a function of the
multiplex rate. It can be used to directly or indirectly drive LOADIN
fLOADOUT = fCLOCK/M
where (M = 1 for 1:1 Multiplex Mode)
where (M = 2 for 2:1 Multiplex Mode).
PRGCKOUT
Programmable Clock Output (TTL Compatible Output). This output control signal
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
fPRGCKOUT = fCLOCK/N
where N = 4, 8, 16 and 32.
SCKIN
Video Shift Clock Input (TTL Compatible Input). The signal on this input is internally
gated synchronously with the BLANK signal. The resultant output, SCKOUT, is a
video clocking signal that is stopped during video blanking periods.
SCKOUT
Video Shift Clock Output (TTL Compatible Output). This output is a synchronously
gated version of SCKIN and BLANK. SCKOUT, is a video clocking signal that is
stopped during video blanking periods.
CLOCK, CLOCK
Clock Inputs (ECL Compatible Inputs). These differential clock inputs are designed to
be driven by ECL logic levels configured for single supply (+5 V) operation. The clock
rate is normally the pixel clock rate of the system.
BLANK
Composite Blank (TTL Compatible Input). This video control signal drives the analog
outputs to the blanking level.
SYNC
Composite-Sync Input (TTL Compatible Input). This video control signal drives the
IOG analog output to the SYNC level. It is only asserted during the blanking period.
CR22 in Command Register 2 must be set if SYNC is to be decoded onto the analog
output, otherwise the SYNC input is ignored.
SYNCOUT
Composite SYNC O/P (TTL Compatible Output). This video output is a delayed version of SYNC. The delay corresponds to the number of pipeline stages of the device.
D0–D9
Databus (TTL Compatible Input/Output Bus). Data, including color palette values and
device control information is written to and read from the device over this 10-bit, bidirectional databus. 10-bit data or 8-bit data can be used. The databus can be configured
for either 10-bit parallel data or byte data (8+2) as well as standard 8-bit data. Any unused bits of the databus should be terminated through a resistor to either he digital
power plane (VCC) or GND.
CE
Chip Enable (TTL Compatible Input). This input must be at Logic “0” when writing to
or reading from the device over the databus (D0–D9). Internally, data is latched on the
rising edge of CE.
–8–
REV. B
ADV7152
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/W and CE must
be at Logic “0” to write data to the part. R/W must be at Logic “1” and CE at Logic
“0” to read from the device.
C0, C1
Command Controls (TTL Compatible Inputs). These inputs determine the type of read
or write operation being performed on the device over the databus (see Interface Truth
Table). Data on these inputs is latched on the falling edge of CE.
IOR; IOR, IOG; IOG, IOB;
IOB
Red, Green and Blue Current Outputs (High Impedance Current Sources). These RGB
video outputs are specified to directly drive RS-343A and RS-170 video levels into doubly terminated 75 Ω loads.
IOR, IOG and IOB are the complementary outputs of IOR, IOG and IOB. These outputs can be tied to GND if it is not required to use differential outputs.
VREF
Voltage Reference Input (Analog Input). An external 1.235 V voltage reference is required to drive this input. An AD589 (2-terminal voltage reference) or equivalent is recommended. (Note: It is not recommended to use a resistor network to generate the
voltage reference.)
RSET
Output Full-Scale Adjust Control (Analog Input). A resistor connected between this pin
and analog ground controls the absolute amplitude of the output video signal. The value
of RSET is derived from the full-scale output current on IOG according to the following
equations:
RSET (Ω) = C1 × RREF/IOG (mA); SYNC on GREEN
RSET (Ω) = C2 × RREF/IOG (mA); No SYNC on GREEN.
Full-Scale output currents on IOR and IOB for a particular value of RSET are given by:
IOR (mA)= C2 × RREF(V)/RSET (Ω)
and
IOB (mA) = C2 × RREF (V)/RSET (Ω)
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.
IPLL
Phase Lock Loop Output Current (High Impedance Current Source). This output is
used to enable multiple ADV7150/ADV7152s along with ADV7151s to be synchronized
together with subpixel 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
IPLL (mA) = 1,728 × RREF(V)/RSET (Ω)
When not using the IPLL function, this output pin should be tied to GND.
VAA
Power Supply (+5 V ± 5%). The part contains multiple power supply pins, all should be
connected together to one common +5 V filtered analog power supply.
GND
Analog Ground. The part contains multiple ground pins, all should be connected
together to the system’s ground plane.
REV. B
–9–
ADV7152
(Continued from page 1)
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 ports flexibility allows for direct interface to most standard frame buffer
memory configurations.
An external ECL oscillator source with differential outputs is all
that is required to drive the CLOCK and CLOCK inputs of the
ADV7152. The part can also be driven by an external clock
generator chip circuit, such as the AD730.
The ADV7152 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 ADV7152 is fabricated in a +5 V CMOS process. Its
monolithic CMOS construction ensures greater functionality
with low power dissipation.
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.
The ADV7152 is packaged in a plastic 100-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.
The on-chip video clock controller circuit generates all the internal clocking and some additional external clocking signals.
CIRCUIT DETAILS AND OPERATION
Pixel Port and Clock Control Circuit
OVERVIEW
The Pixel Port of the ADV7152 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:
Digital video or pixel data is latched into the ADV7152 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 ADV7152 is broken down into three separate functional blocks. These are:
Color Data
Pixel Controls
Palette Selects
1. Pixel port and clock control circuit
The associated clocking signals for the pixel port include:
2. MPU port, registers and color palette
Clock Inputs
3. Digital-to-analog converters and video outputs
Clock Outputs
Table I shows the architectural and packaging differences between 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
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
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
ADV7151*
RED, GREEN, BLUE
SYNC, BLANK
PS0–PS1
CLOCK, CLOCK,
LOADIN, SCKIN
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 ADV7152 has 48 color data inputs. The part has two (for
2:1 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 2:1
and 1:1 multiplex modes.
RED
8
A
24
GREEN 8
•
•
BLUE
*See ADV7151 and ADV7150 data sheets for more information on these parts.
MULTIPLEXER
8
B
24
24
Figure 10. Multiplexed Color Inputs for the ADV7152
–10–
REV. B
ADV7152
resultant pixel or dot clock rate of 100 MHz. As mentioned in
the previous section, the ADV7152 supports a number of color
data formats in 2:1 and 1:1 multiplex modes.
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
fLOADIN = fCLOCK/2
fLOADIN = fCLOCK
In 1:1 multiplex mode, the ADV7152 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.)
2:1 Multiplex Mode
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 ADV7152, allowing the CLOCK to occur
anywhere during the LOADIN cycle.
VIDEO MEMORY/
FRAME BUFFER
ADV7152
24
VRAM (BANK A)
A
50MHz
MULTIPLEXER
Alternatively, the LOADOUT signal of the ADV7152 can be
used. LOADOUT can be connected either directly or indirectly
to LOADIN. Its frequency is automatically set to the correct
LOADIN requirement.
24
100MHz
(2 x 50MHz)
24
VRAM (BANK B)
SYNC, BLANK
B
50MHz
Figure 11. Direct Interfacing of Video Memory to
ADV7152
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.”
CLOCK CONTROL CIRCUIT
SYNCOUT
In some applications where it is not permissible to encode
SYNC on green (IOG), SYNCOUT can be used as a separate
TTL digital SYNC output. This has the advantage over an independent (of the ADV7150) 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 SYNC from being encoded onto IOG. Bit CR12 of
Command Register 1 is set to “1,” enabling SYNCOUT. The
output signal generates a TTL SYNCOUT with correct pipeline
delay that is capable of directly driving the composite SYNC
signal of a computer monitor.
The ADV7152 has an integrated Clock Control Circuit (Figure
12). This circuit is capable of both generating the ADV7152’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.
CLOCK, CLOCK Inputs
The Clock Control Circuit is driven by the pixel clock inputs,
CLOCK and CLOCK. These inputs can be driven by a differential ECL oscillator running from a +5 V supply.
CLOCK
CLOCK
PRGCKOUT
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 PSI 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.
ECL
TO
TTL
DIVIDE BY N
(÷ N)
DIVIDE BY M
(÷ M)
LOADOUT
SCKOUT
LATCH
BLANK
ENABLE
SYNC
SCKIN
ADV7152
LOADIN
TO COLOR DATA
MULTIPLEXER
Multiplexing
The onboard multiplexers of the ADV7152 eliminate the need
for external data serializer circuits. Multiple video memory devices can be connected, in parallel, directly to the device. Figure 11 shows two memory banks of 50 MHz memory connected
to the ADV7152, running in 2:1 multiplex mode, giving a
REV. B
M IS A FUNCTION OF MULTIPLEX RATE
M = 2 IN 2:1 MULTIPLEX MODE
M = 1 IN 1:1 MULTIPLEX MODE
N IS INDEPENDENTLY
PROGRAMMABLE
N= (4, 8, 16, 32)
Figure 12. Clock Control Circuit of the ADV7152
–11–
ADV7152
Alternatively, the ADV7152 CLOCK inputs can be driven by a
Programmable Clock Generator (Figure 13), 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
VAA
GND
VCC
VCC
220Ω
+5V
CLOCK
ECL OUT–
CLOCK
330Ω
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
220Ω
ECL OUT+
If however, it is required that the ADV7152 has a fixed number
of pipeline delays (tPD), LOADOUT and LOADIN must conform to timing specifications t10 and τ-t11 as illustrated in Figures 4 and 5.
PRGCKOUT
+5V
VCLOCK
If it is not necessary to have a known fixed number of pipeline
delays, then there is no limitation on the delay between LOADOUT and LOADIN (LOADOUT(1) and LOADOUT(2)).
LOADIN and Pixel Data must conform to the setup and hold
times (t8 and t9).
330Ω
fPRGCKOUT = f CLOCK/N
GND
CLOCK
GENERATOR
GND
VAA
ADV7152
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.
0.1 µF
VREF
VREF OUT
SCKIN, SCKOUT
GND
These video memory signals are used to minimize external support chips. Figure 15 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 7 of the Timing Waveform section
shows the relationship between SCKOUT, SCKIN and
BLANK.
GND
D0–D3 CS R/W
Figure 13. PLL Generator Driving CLOCK, CLOCK of the
ADV7152
CLOCK CONTROL SIGNALS LOADOUT
The ADV7152 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 CR36 of Command Register 3.
fLOADOUT = fCLOCK/2
fLOADOUT = fCLOCK
SCKOUT
LATCH
BLANK
2:1 Multiplex Mode
1:1 Multiplex Mode
ENABLE
SYNC
SCKIN
The LOADOUT signal is used to directly drive the LOADIN
pixel latch signal of the ADV7152. 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.
LOADOUT(1)
LOADOUT
VIDEO
FRAME
BUFFER
ADV7152
LOADIN
LOADOUT
VIDEO
FRAME
BUFFER LOADOUT(2)
Figure 15. SCKOUT Generation Circuit
The SCKOUT signal is essentially the video memory shift control signal. It is stopped during the screen retrace. Figure 16
shows a suggested frame buffer to ADV7152 interface. This is a
minimum chip solution and allows the ADV7152 control the
overall graphics system clocking and synchronization.
ADV7152
LOADOUT
LOADIN
LOADIN
PIXEL
DATA
PIXEL
DATA
VIDEO
FRAME
BUFFER
SCKIN
ADV7152
BLANK
LOADOUT
SCKOUT
LOADOUT(1)
DELAY
LOADIN
PIXEL
DATA
LOADOUT(2)
Figure 16. ADV7152 Interface Using SCKIN and SCKOUT
Figure 14. LOADOUT vs. Pixel Clock Input (CLOCK, CLOCK)
–12–
REV. B
ADV7152
Pipeline Delay and On-Board Calibration
24-BIT
PIXEL DATA
The ADV7152 has a fixed number of pipeline delays (tPD), so
long as timings t10 and τ-t11 are met. However, if a fixed pipeline
delay is not a requirement, timings t10 and τ-t11 can be ignored,
a calibration cycle must be run and there is no restriction on
LOADIN to LOADOUT timing. If timings t10 and τ-t11 are not
met, the part will function correctly though with an increased
number of pipeline delays, tPD + N CLOCKS (for 2:1 mode
N = 2, for 1:1 mode N = 0). The ADV7152 has onboard calibration circuitry which synchronizes pixel data and LOADIN
with the internal ADV7152 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.
8
8
BLUE
256 x 8
ANALOG VIDEO
OUTPUTS
8
8-BIT
RED DAC
8
8-BIT
GREEN DAC
8
8-BIT
BLUE DAC
RED
OUT
GREEN
OUT
BLUE
OUT
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).
24-Bit “Gamma” True Color
(CR25, CR26, CR27 = 1, 1, 1 and MR11 = 1)
8-BIT
PIXEL DATA
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.
30-BIT
COLOR DATA
GREEN
256 x 8
24-BIT
COLOR DATA
Figure 18. 24-Bit to 24-Bit Direct True-Color Configuration
The ADV7152 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 TO 30-BIT
LOOK-UP-TABLE
RED
256 x 8
8
COLOR VIDEO MODES
24-BIT
PIXEL DATA
24-BIT TO 24-BIT
LOOK-UP-TABLE
8
8-BIT TO 30-BIT
LOOK-UP-TABLE
RED
256 x 10
GREEN
256 x 10
BLUE
256 x 10
30-BIT
COLOR DATA
ANALOG VIDEO
OUTPUTS
10
10-BIT
RED DAC
10
10-BIT
GREEN DAC
10
10-BIT
BLUE DAC
RED
OUT
GREEN
OUT
BLUE
OUT
ANALOG VIDEO
OUTPUTS
Figure 19. 8-Bit to 30-Bit Pseudo-Color Configuration
8
8
8
RED
256 x 10
GREEN
256 x 10
BLUE
256 x 10
10
10-BIT
RED DAC
10
10-BIT
GREEN DAC
10
10-BIT
BLUE DAC
RED
OUT
This mode allows for the display of 256 simultaneous colors out
of a total palette of millions of addressable colors.
GREEN
OUT
8-Bit “Standard” Pseudo Color
(CR25, CR26, CR27 = X, 0, 0 or X, 1, 0 or X, 0, 1 and
MR11 = 0)
BLUE
OUT
Figure 17. 24-Bit to 30-Bit True-Color Configuration
This mode allows for the display of full 24-bit, GammaCorrected True-Color Images.
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).
8-BIT
PIXEL DATA
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.
8
RED
256 x 8
GREEN
256 x 8
BLUE
256 x 8
This mode allows for the display of full 24-bit True-Color
Images.
REV. B
8-BIT TO 24-BIT
LOOK-UP-TABLE
24-BIT
COLOR DATA
ANALOG VIDEO
OUTPUTS
8
8-BIT
RED DAC
8
8-BIT
GREEN DAC
8
8-BIT
BLUE DAC
RED
OUT
GREEN
OUT
BLUE
OUT
Figure 20. 8-Bit to 24-Bit Pseudo-Color Configuration
–13–
ADV7152
This mode allows for the display of 256 simultaneous colors out
of a total palette of millions of addressable colors.
15-Bit “Gamma” True Color
(CR24, CR25, CR26, CR27 = 0, 0, 1, 1 or 1, 0, 1, 1 and
MR11 = 1)
5
15-BIT TO 30-BIT
LOOK-UP-TABLE
RED
32 x 10
30-BIT
COLOR DATA
The pixel data to the ADV7152 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.
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
PIXEL DATA
PIXEL PORT MAPPING
R4
R3
R2
R1
R0
x
ANALOG VIDEO
OUTPUTS
x
x
10
10-BIT
RED DAC
10
10-BIT
GREEN DAC
RED
OUT
G4
5
GREEN
32 x 10
BLUE
32 x 10
5
10
GREEN
OUT
10-BIT
BLUE DAC
G3
G2
G1
BLUE
OUT
G0
x
x
Figure 21. 15-Bit to 30-Bit True-Color Configuration
x
This mode allows for the display of 15-bit, Gamma-Corrected
True-Color Images.
B4
B3
15-Bit “Standard” True Color
(CR24, CR25, CR26, CR27 = 0, 0, 1, 1 or 1, 0, 1, 1 and
MR11 = 0)
B2
B1
B0
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).
15-BIT
PIXEL 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
ANALOG VIDEO
OUTPUTS
8
8-BIT
RED DAC
8
8-BIT
GREEN DAC
8
8-BIT
BLUE DAC
RED
OUT
GREEN
OUT
BLUE
OUT
Figure 22. 15-Bit to 24-Bit True-Color Configuration
This mode allows for the display of 15-bit True-Color Images.
x
x
x
PIXEL
INPUT
DATA
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
0
R3
0
R2
0
R1
R4
R0
R3
x
R2
256 x 10 RAM
(RED LUT)
5
x
R1
x
R0
G4
0
G3
0
G2
0
G1
G4
G0
G3
x
G2
x
G1
x
G0
B4
0
B3
0
LOCATION "0"
0
B1
B4
B0
B3
x
B2
10
TO
RED
DAC
256 x 10 RAM
(GREEN LUT)
5
B2
LOCATION "31"
LOCATION "31"
LOCATION "0"
10
TO
GREEN
DAC
256 x 10 RAM
(BLUE LUT)
5
x
B1
x
B0
DATA
LATCHED
TO
PIXEL
PORT
DATA
INTERNALLY
SHIFTED
TO 5 LSBS
LOCATION "31"
LOCATION "0"
10
TO
BLUE
DAC
DATA LATCHES
FIRST 32
LOCATIONS
OF RAM
Figure 23. 15-Bit True-Color Mapping Using R3–R7, G3–G7
and B3–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.
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 23. 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 24. (Note: Data on unused pixel inputs is
ignored.)
–14–
REV. B
ADV7152
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
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
Table II. Databus Width Table
0
R3
0
R2
0
R1
R4
R0
R3
256 x 10 RAM
(RED LUT)
5
G4
LOCATION "31"
R2
G3
R1
G2
R0
x
0
G1
0
B4
G4
B3
G3
B2
G2
B1
G1
G0
x
0
x
0
x
0
x
B4
x
B3
x
B2
x
B1
x
B0
LOCATION "31"
Read/Write
Mode
10 Bit
10 Bit
8 Bit
8 Bit
10 Bit
8 Bit
10 Bit
8 Bit
10-Bit Parallel
8-Bit Parallel
8+2 Byte
8-Bit Parallel
The ADV7152 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 26 shows the
internal register configuration.
10
TO
GREEN
DAC
LOCATION "0"
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 CE input latches data to or from the part.
256 x 10 RAM
(BLUE LUT)
5
DATA
LATCHED
TO
PIXEL
PORT
RAM/DAC
Resolution
Register Mapping
256 x 10 RAM
(GREEN LUT)
5
B0
TO
RED
DAC
LOCATION "0"
0
G0
10
Databus
Width
LOCATION "31"
The R/W control input determines between read or write accesses. 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).
TO
BLUE
DAC
LOCATION "0"
DATA
INTERNALLY
SHIFTED
TO 5 LSBS
10
DATA LATCHES
FIRST 32
LOCATIONS
OF RAM
Figure 24. 15-Bit True-Color Mapping Using R0–R7
and G0–G6
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.
MICROPROCESSOR (MPU) PORT
The ADV7152 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 25) consists of a bidirectional,
10-bit wide databus and interface control signals CE, C0, C1
and R/W. The 10-bit wide databus is user configurable as
illustrated.
CONTROL REGISTERS
PIXEL MASK
REGISTER
ADDRESS
REGISTER
MODE
REGISTER
TEST
REGISTERS
ADDR
(A7–A0)
(MR1)
ID
REGISTER
COMMAND
REGISTERS
(CR1–CR3)
REVISION
REGISTER
DATA TO
PALETTES
30
RED
REGISTER
COLOR REGISTERS
GREEN
REGISTER
MPU PORT
10 (8+2)
CE
R/W
C0
C1
D0 – D9
Figure 25. MPU Port and Register Configuration
REV. B
–15–
BLUE
REGISTER
ADV7152
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.
Register Accesses
The MPU can write to or read from all of the ADV7152s 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.
MODE REGISTER
(MR17–MR10)
C1 = 1
C0 = 1
ADDRESS REGISTER
(A7–A0)
C1 = 0
C0 = 0
C1 = 1
C0 = 0
ADDRESS
REGISTER
(A7–A0)
00H
01H
02H
03H
C1 = 0
C0 = 1
CONTROL
REGISTERS
PIXEL TEST REGISTER
RED
REGISTER
(R9–R0)
R
G
B
DAC TEST REGISTER
R
G
GREEN
REGISTER
(G9–G0)
BLUE
REGISTER
(B9–B0)
B
SYNC, BLANK & I PLL
TEST REGISTER
ID REGISTER (READ ONLY)
04H
PIXEL MASK REGISTER
05H
COMMAND REGISTER 1
06H
COMMAND REGISTER 2
07H
COMMAND REGISTER 3
08H
RESERVED * (READ ONLY)
09H
RESERVED * (READ ONLY)
0AH
RESERVED * (READ ONLY)
0BH
REVISION REGISTER
POINTS TO LOCATION
CORRESPONDING TO
ADDRESS REG (A7–A0)
LOOK-UP TABLE RAM
(256 x 30)
ADDRESS REG =
ADDRESS REG + 1
* THIS REGISTER IS READ ONLY.
Figure 26. Internal Register Configuration and Address Decoding
Table III. Interface Truth Table (10-Bit Databus Mode)
R/W
C1
C0
Databus (D9–D0)
Operation
Result
0
0
0
1
0
1
1
0
0
DB7–DB0
DB7–DB0
DB7–DB0
Write to Mode Register
DB7–DB0 → MR17–MR10
Write to Address Register
DB7–DB0 → A7–A0
Write to Control Registers
DB7–DB0 → Control Register
(Particular Control Register Determined by Address Register)
0
0
0
0
0
0
1
1
1
DB9–DB0
DB9–DB0
DB9–DB0
Write to RED Register
DB9–DB0 → R9–R0
Write to GREEN Register
DB9–DB0 → G9–G0
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
1
0
1
1
0
0
DB7–DB0
DB7–DB0
DB7–DB0
Read Mode Register
MR17–MR10 → DB7–DB0
Read Address Register
A7–A0 → DB7–DB0
Read Control Registers
Register Data → DB7–DB0
(Particular Control Register Determined by Address Register)
1
1
1
0
0
0
1
1
1
DB9–DB0
DB9–DB0
DB9–DB0
Read RED RAM Location
R9–R0 → DB9–DB0
Read GREEN RAM Location
G9–G0 → DB9–DB0
Read BLUE RAM Location
B9–B0 → DB9–DB0
(RAM Location Pointed to by Address Reg (A7–A0))
Address Register = Address Register + 1
DB = Data Bit.
–16–
REV. B
ADV7152
Table IV. Interface Truth Table (8-Bit Databus Mode)*
R/W
C1
C0
Databus
(D9–D0)
0
0
0
1
0
1
1
0
0
DB7–DB0
DB7–DB0
DB7–DB0
Write to Mode Register
DB7–DB0 → MR17–MR10
Write to Address Register
DB7–DB0 → A7–A0
Write to Control Registers
DB7–DB0 → Control Registers
(Particular Control Register Determined by Address Register (A7–A0))
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
DB9–DB2
DB1–DB0
DB9–DB2
DB1–DB0
DB9–DB2
DB1–DB0
Write to RED Register
DB9–DB2 → R9–R2
Write to RED Register
DB1–DB0 → R1–R0
Write to GREEN Register
DB9–DB2 → G9–G2
Write to GREEN Register
DB1–DB0 → G1–G0
Write to BLUE Register
DB9–DB2 → B9–B2
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
1
0
1
1
0
0
DB7–DB0
DB7–DB0
DB7–DB0
Read Mode Register
MR17–MR10 → DB7–DB0
Read Address Register
A7–A0 → DB7–DB0
Read Control Registers
Register Data → DB7–DB0
(Particular Control Register Determined by Address Register)
1
1
1
1
1
1
0
0
0
0
0
0
1
1
1
1
1
1
DB9–DB2
DB1–DB0
DB9–DB2
DB1–DB0
DB9–DB2
DB1–DB0
Read RED RAM Location
R9–R2 → DB9–DB2
Read RED RAM Location
R1–R0 → DB1–DB0
Read GREEN RAM Location
G9–G2 → DB9–DB2
Read GREEN RAM Location
G1–G0 → DB1–DB0
Read BLUE RAM Location
B9–B2 → DB9–DB2
Read BLUE RAM Location
B1–B0 → DB1–DB0
(RAM Location Pointed to by Address Register (A7–A0))
Address Register = Address Register + 1
Operation
Result
*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.
Power-On Reset
On power-up of the ADV7152 executes a power-on reset operation. This initializes the pixel port such that the pixel sequence
AB 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.”
REV. B
The output clocking signals are also set during this reset period.
PRGCKOUT = CLOCK/32
LOADOUT = CLOCK/2
The power-on reset is activated when VAA goes from 0 V to
5 V. This reset is active for 1 µs. The ADV7152 should not be
accessed during this reset period. The pixel clock should be
applied at power-up.
–17–
ADV7152
REGISTER PROGRAMMING
MPU Databus Width (MR12)
The following section describes each register, including Address
Register, Mode Register and each of the nine Control Registers
in terms of its configuration.
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.
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.
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.
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.
Palette Select Match Bits Control (MR17–MR16)
These bits allow multiple palette devices to work together.
When bits PSI 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.
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.
Control Registers
The ADV7152 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.
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 AB 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.
Pixel Test Register
(Address Reg (A7–A0) = 00H)
RAM-DAC Resolution Control (MR11)
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.
When this is programmed with a “1,” the RAM is 30 bits deep
(10 bits each for red, green and blue) and each of the three
DACs is configured for 10-bit resolution. When MR11 is programmed with a “0,” the RAM is 24 bits deep (8 bits each for
red, green and blue) and the DACs are configured for 8-bit
resolution. The two LSBs of the 10-bit DACs are pulled down
to zero in 8-bit RAM-DAC mode.
MR19
MR18
MR17
MR16
MR15
MR14
CALIBRATE
LOADIN
RESERVED*
PALETTE SELECT
MATCH BITS CONTROL
MR16
PS0
MR17
PS1
MR10
MPU DATA BUS
WIDTH
0
1
MR14 MR13
0
1
0
1
MR11
MR12
MR15
OPERATIONAL MODE CONTROL
0
0
1
1
MR12
MR13
RESERVED
NORMAL OPERATION
RESERVED
RESERVED
8-BIT (D7–D0)
10-BIT (D9–D0)
RAM-DAC
RESOLUTION CONTROL
MR11
0
1
* THESE BITS ARE READ-ONLY RESERVED BITS.
8-BIT
10-BIT
RESET CONTROL
MR10
A READ CYCLE WILL RETURN ZEROS "00."
Mode Register 1 (MR1) (MR19–MR10)
–18–
REV. B
ADV7152
DAC Test Register
(Address Reg (A7–A0) = 01H)
COMMAND REGISTER 1 (CR1)
(Address Reg (A7–A0) = 05H)
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.
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).
SYNC, BLANK and IPLL Test Register
(Address Reg (A7–A0) = 02H)
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–CR17.
In read mode, CR11 and CR13–CR19 are returned as zeros.
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.
COMMAND REGISTER 1 BIT DESCRIPTION
Calibration Control (CR10)
ID Register
(Address Reg (A7–A0) = 03H)
This is an 8-bit wide “Identification” read-only register. For the
ADV7152 it will always return the hexadecimal value 8CH.
This bit automatically calibrates the onboard LOADIN/
LOADOUT synchronization circuit. MR15 of Mode Register
MR1 must be set to “0.”
Pixel Mask Register (Address Reg (A7–A0) = 04H)
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.
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.
CR19
CR18
CR17
CR16
CR15
CR14
CR13
CR12
CR11
CR10
RESERVED *
SYNCOUT CONTROL
CR17–CR13
(00000)
THESE BITS
SHOULD BE
SET TO ZERO
CR11
(0)
CR12
0
DISABLE
1
ENABLE
SYNCOUT
THIS BIT
SHOULD BE
SET TO ZERO
CALIBRATION CONTROL
* THESE BITS ARE READ-ONLY RESERVED BITS.
A READ CYCLE WILL RETURN ZEROS "00."
Command Register 1 (CR1) (CR19–CR10)
REV. B
–19–
CR10
0
DISABLE
1
CALIBRATES
ON EVERY
VERTICAL
SYNC
(MR15=0)
ADV7152
IPLL Trigger Control (CR21)
COMMAND REGISTER 2 (CR2)
(Address Reg (A7–A0) = 06H)
This bit specifies whether the IPLL output is triggered from
BLANK or SYNC.
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).
SYNC Recognition Control (CR22)
This bit specifies whether the video SYNC input is to be
encoded onto the IOG analog output or ignored.
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.
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)
COMMAND REGISTER 2 BIT DESCRIPTION
R7 Trigger Polarity Control (CR20)
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.
This bit determines whether the pixel data is latched into the
test registers on the rising or falling edge of R7, see “Test Diagnostics” section in Appendix 3.
CR29 CR28
CR27
CR26
CR25
CR24
PEDESTAL
ENABLE
CONTROL
RESERVED *
* THESE BITS ARE READ–ONLY
RESERVED BITS.
A READ CYCLE WILL
RETURN ZEROS "00."
0 IRE
7.5 IRE
TRUE-COLOR / PSEUDO-COLOR MODE CONTROL
CR27 CR26 CR25 CR24
0
0
1
1
0
1
0
1
0
0
0
0
0
0
0
0
1
1
0
1
1
1
1
0
CR22
CR21
CR20
SYNC
RECOGNITION
CONTROL
CR22
CR23
0
1
CR23
COLOR 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
24-BIT TRUE COLOR
R7–R0, G7–G0, B7–B0
0
1
IGNORE
DECODE
IPLL TRIGGER
CONTROL
CR21
0
SYNC
1
BLANK
R7 TRIGGER
POLARITY CONTROL
CR20
0
1
Command Register 2 (CR2) (CR29–CR20)
–20–
REV. B
ADV7152
COMMAND REGISTER 3 (CR3)
(Address Reg (A7–A0) = 07H)
BLANK Pipeline Delay Control (CR35–CR32)
These bits specify the additional pipeline delay that can be
added to the BLANK function, relative to the overall device
pipeline delay (tPD). As the BLANK control normally enters the
video DAC from a shorter pipeline than the video pixel data,
this control is useful in deskewing the pipeline differential.
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.
Pixel Multiplex Control (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.
COMMAND REGISTER 3-BIT DESCRIPTION
PRGCKOUT Frequency Control (CR31–CR30)
Revision Register
(Address Reg (A7–A0) = 0BH)
These bits specify the output frequency of the PRGCKOUT
output. PRGCKOUT is a divided down version of the pixel
CLOCK.
CR39
CR38
RESERVED*
* THESE BITS ARE READ-ONLY
RESERVED BITS.
A READ CYCLE WILL RETURN
ZEROS "00".
CR37
CR36
This register is a read only register containing the revision of
silicon.
CR35
CR34
CR33
CR32
CR31
CR30
EXTRA BLANK PIPELINE DELAY CONTROL
(ADDS TO PIXEL PIPELINE DELAY; tPD )
CR35 CR34 CR33 CR32
0
0
0
–
–
1
0
0
0
–
–
1
0
0
1
–
–
1
0
1
0
–
–
1
BLANK PIPELINE DELAY
tPD
tPD
tPD
–
–
tPD
+ 1 x LOADOUT
+ 2 x LOADOUT
–
–
–
–
+ 15 x LOADOUT
PRGCKOUT
FREQUENCY CONTROL
PIXEL MULTIPLEX CONTROL
CR37 CR36
CR31 CR30
0
0
1
1
0
1
0
1
1:1 MUXING: LOADOUT = CLOCK ÷1
2:1 MUXING: LOADOUT = CLOCK ÷2
RESERVED
RESERVED
Command Register 3 (CR3) (CR39–CR30)
REV. B
–21–
0
0
1
1
0
1
0
1
CLOCK ÷4
CLOCK ÷8
CLOCK ÷16
CLOCK ÷32
ADV7152
DIGITAL-TO-ANALOG CONVERTERS
(DACS) AND VIDEO OUTPUTS
IOR, IOG, IOB
DACs
The ADV7152 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 IPLL and VREF as well as
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 = “1”) 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 Ω load (doubly terminated 75 Ω).
IOR, IOB
mA
ZS =75Ω
ZL = 75Ω
(CABLE)
(SOURCE
TERMINATION)
(MONITOR)
Figure 27. DAC Output Termination (Doubly Terminated
75 Ω Load)
Reference Input and RSET
An external 1.23 V voltage reference is required to drive the
analog outputs of the ADV7152. The reference voltage is connected to the VREF input.
A resistor RSET is connected between the RSET input of the part
and ground. For specified performance, RSET has a value of
280 Ω. This corresponds to the generation of RS-343A video
levels (with SYNC on IOG and Pedestal = 7.5 IRE) into a doubly terminated 75 Ω load. Figure 28 illustrates the resulting
video waveform, and the Video Output Truth Table shows the
corresponding control input stimuli.
IOG
V
mA
V
0.714
26.67
1.000
WHITE LEVEL
EY
SC
A
LE
19.05
ZO = 75Ω
1.44
0.054
9.05
G
R
92.5 IRE
BLACK LEVEL
0.340
7.5 IRE
0
0
7.62
0.286
0
0
BLANK LEVEL
40 IRE
SYNC LEVEL
Figure 28. Composite Video Waveform (SYNC Decoded on IOG; Pedestal = 7.5 IRE; RSET = 280 Ω)
Table V. Video Output Truth Table
Description
IOG
(mA)
IOR, IOB
(mA)
SYNC
BLANK
DAC
Input Data
WHITE LEVEL
VIDEO
VIDEO to BLANK
BLACK LEVEL
BLACK to BLANK
BLANK LEVEL
SYNC LEVEL
26.67
Video + 9.05
Video + 1.44
9.05
1.44
7.62
0
19.05
Video + 1.44
Video + 1.44
1.44
1.44
0
0
1
1
0
1
0
1
0
1
1
1
1
1
0
0
3FFH
Data
Data
000H
000H
xxxH
xxxH
–22–
REV. B
ADV7152
Variations on RS-343A
IOR, IOB, IOG
mA
Various other video output configurations can be implemented
by the ADV7152, including RS-170. Values of RSET for particular output video formats/levels are calculated by using the equations for RSET given in the “Pin Configuration” section. The
table shows calculated values of RSET for some of the most common variants on the RS-343A standard. The associated waveforms are shown in the diagrams.
RSET (V)
Video Signal
265
280
259
SYNC decoded on IOG; Pedestal = 0 IRE
No SYNC decoded; Pedestal = 7.5 IRE
No SYNC decoded; Pedestal = 0 IRE
IOR, IOB
mA
V
WHITE
LEVEL
SC
AL
E
8.05
GR
EY
100 IRE
0
BLACK/BLANK
LEVEL
0.302
43 IRE
0
0
SYNC
LEVEL
0
The IPLL signal is internally triggered by either the falling edge
of SYNC or BLANK as determined by CR21 of Command
Register 2.
WHITE LEVEL
GR
EY
92.5 IRE
1.44
BLACK LEVEL
0.054
7.5 IRE
0
0
BLANK LEVEL
Figure 30. Composite Video Waveform
(Pedestal = 7.5 IRE; RSET = 280 Ω)
REV. B
BLACK/ BLANK
LEVEL
This output synchronization signal is used in applications where
it is necessary to synchronize multiple palette devices (ADV7150
+ ADV7151) to subpixel resolution. Each devices IPLL 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 IPLL signals. This time difference is then used
to phase shift the CLOCK inputs on one or other of the devices
inputs.
SC
AL
E
V
0.714
0
Figure 31. Composite Video Waveform
(Pedestal = 0 IRE; RSET = 259 Ω)
IOR, IOB, IOG
mA
GR
EY
SC
AL
E
100 IRE
Figure 29. Composite Video Waveform (SYNC
Decoded on IOG; Pedestal = 0 IRE; RSET = 265 Ω)
19.05
WHITE LEVEL
IPLL Synchronization Output Control
IOG
mA
V
18.62 0.698 26.67 1.000
0
V
19.05 0.714
–23–
ADV7152
APPENDIX 1
BOARD DESIGN AND LAYOUT CONSIDERATIONS
The ADV7152 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.
It is important to note that while the ADV7152 contains circuitry to reject power supply noise, this rejection decreases with
frequency. If a high frequency switching power supply is used,
the designer should pay close attention to reducing power supply noise and consider using a three terminal voltage regulator
for supplying power to the analog power plane.
Digital Signal Interconnect
The digital inputs to the ADV7152 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.
The layout should be optimized for lowest noise on the ADV7152
power and ground lines by shielding the digital inputs and providing good decoupling. The lead length between groups of VAA
and GND pins should by minimized so as to minimize inductive
ringing.
Due to the high clock rates involved, long clock lines to the
ADV7152 should be avoided to reduce noise pickup.
Ground Planes
The ground plane should encompass all ADV7152 ground pins,
voltage reference circuitry, power supply bypass circuitry for the
ADV7152, the analog output traces, and all the digital signal
traces leading up to the ADV7152. The ground plane is the
graphics board’s common ground plane.
Power Planes
The ADV7152 and any associated analog circuitry should have
its own power plane, referred to as the analog power plane (VAA).
This power plane should be connected to the regular PCB
power plane (VCC) at a single point through a ferrite bead. This
bead should be located within three inches of the ADV7152.
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 ADV7152 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 VAA
pins on the ADV7152 must have at least one 0.1 µF decoupling
capacitor to GND. These capacitors should be placed as close
as possible to the device.
Any active termination resistors for the digital inputs should be
connected to the regular PCB power plane (VCC), and not the
analog power plane.
Analog Signal Interconnect
The ADV7152 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.
Digital Inputs, especially Pixel Data Inputs and clocking signals
(CLOCK, LOADOUT, LOADIN, etc.) should never overlay
any of the analog signal circuitry and should be kept as far away
as possible.
For best performance, the analog outputs (IOR, IOG, IOB)
should each have a 75 Ω load resistor connected to GND.
These resistors should be placed as close as possible to the
ADV7152 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 Ω load, this means that IOR, IOG, IOB are each
terminated with 37.5 Ω resistors.
–24–
REV. B
ADV7152
POWER SUPPLY DECOUPLING (0.1µF AND 0.01µF CAPACITOR FOR EACH V AA GROUP)
0.1µF
0.01µF
0.1µF
+5V (V AA )
1kΩ
(1% METAL)
VAA
0.1µF
COMP
0.1µF
0.01µF
0.1µF
33µF
0.01µF
+5V (VCC )
+5V (VAA )
ANALOG POWER PLANE
+5V (VAA )
0.01µF
L1
(FERRITE BEAD)
0.1µF
0.1µF
VREF
AD589
(1.2V REF)
RSET
RSET
280Ω
ADV7152
CO-AXIAL CABLE
(75Ω)
75Ω
IOR
75Ω
MONITOR
(CRT)
IOG
75Ω
IOB
75Ω
75Ω
75Ω
BNC
CONNECTORS
IOR
COMPLIMENTARY
OUTPUTS
IOG
NOTES:
1. ALL RESISTORS ARE 1% METAL FILM
2. 0.1µF AND 0.01µF CAPACITORS ARE CERAMIC
3. ADDITIONAL DIGITAL CIRCUITRY OMITTED FOR CLARITY
IOB
GND
IPLL
Recommended Analog Circuit Layout
APPENDIX 2
TYPICAL FRAME BUFFER INTERFACE
CLOCK
CLOCK GENERATOR
ECL
TO
TTL
CLOCK
PRGCKOUT
DIVIDE BY N
(÷ N)
DIVIDE BY M
(+M)
LOADOUT
SCKOUT
CLOCK
LATCH
GRAPHICS
PROCESSOR/
CONTROLLER
BLANK
SYNC
BLANK
SYNC
ENABLE
SCKIN
LOADIN
ADV7152
24
24
24
24
VRAM
(BANK A)
FRAME
BUFFER/
VIDEO MEMORY
REV. B
24
VRAM
(BANK B)
–25–
MULTIPLEXER
TO
PALETTE/RAM
& DAC
ADV7152
APPENDIX 3
10-BIT DACS AND GAMMA CORRECTION
10-Bit DACs
Gamma Correction 8 Bits vs. 10 Bits
10-bit RAM-DAC resolution allows for nonlinear video correction, in particular Gamma Correction. The ADV7152 allows for
an increase in color resolution from 24-bit to 30-bit effective
color without the necessity of a 30-bit deep frame buffer. In
true-color mode, for example, the part effectively operates as a
24-bit to 30-bit color look-up table.
Up to now we have assumed that there exists a linear relationship between the actual RGB values input to a monitor and the
intensity produced on the screen. This, however, is not the case.
Half scale digital input (1000 0000) might correspond to only
20% output intensity on the CRT (Cathode Ray Tube). The intensity (ICRT) produced on a CRT by an input value IIN is given
by:
ICRT = (IIN)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 (IIN);
therefore:
Quantized to
8 Bits
Quantized to
10 Bits
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
0.977797
0.979304
0.980807
0.982306
0.983801
0.985292
0.986780
0.988264
0.989744
0.991220
0.992693
0.994161
0.995626
0.997088
0.998546
1.000000
250
250
251
251
25 1
252
252
252
253
253
254
254
254
255
255
255
1001
1002
1004
1005
1007
1008
1010
1011
1013
1015
1016
1018
1019
1021
1022
1023
(k = 1, normally)
1.00
Traditionally, there has been a tradeoff between implementing a
nonlinear graphics function, such as gamma correction, and
color dynamic range. The ADV7152 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 ADV7152, 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.
0.90
DAC OUTPUT – Normalized to 1
IIN(corrected) = k(IIN)1/χ
8-Bit Data
Gamma
Corrected
(2.7)
VE
UR
NC
TIO
C
E
E
RR
EY
E
CO
TH
MA
M
Y
B
GA
ED
IV
CE
E
PR
SE
ON
P
S
RE
AR
SE
NE
I
L
ON
SP
E
R
T
CR
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
0
32
64
96
128
160
192
224
256
INPUT CODE – Decimal
Gamma Correction Curve (Gamma Value = 2.7)
Other applications of the 10-bit RAM-DAC include closed-loop
monitor color calibration.
–26–
REV. B
ADV7152
APPENDIX 4
MULTIPLE PALETTE APPLICATIONS
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. 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. When PS0 and PS1 do not match these
bits, the DACs of the particular device are shut down, driving
RGB outputs to 0 mA. PS0 and PS1 are multiplexed similar to
the pixel data, thus allowing for subpixel resolution. The diagrams show an example of one ADV7152 operating in conjunction with three ADV7151’s (Pseudo-Color RAM-DACs). Each
displayed window on the monitor is driven by one of the four
devices. 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 (two ADV7152s or one ADV7152 and one
ADV7151) include:
Cursor Overlay on 24-Bit Graphics
Active Live Video Overlay (from Frame Grabber)
Text/Character Generation and Overlay
(DEVICE: 2)
IOR, IOG, IOB
DACs
(DEVICE: 1)
IOR, IOG, IOB
ZO =75Ω
DACs
ZS = 75Ω
(CABLE)
(SOURCE
TERMINATION)
ZL = 75Ω
(MONITOR)
Multiple Devices Termination for a Single Monitor
ADV7152
R0–R7
G0–G7
B0–B7
256 x 30
RAM
RGB
ANALOG
VIDEO
PALETTE SELECT BITS
PS0, PS1
MR16 MR17
0
0
ADV7151 (1)
P0–P7
256 x 30
PALETTE
RGB
ANALOG
VIDEO
MONITOR
TRUE-COLOR BACKGROUND
PS0 = 0: PS1 = 0
PALETTE SELECT BITS
MR16 MR17
0
1
VIDEO TO MONITOR
ADV7151 (2)
WINDOW 2
(Pseudo-Color)
PS0 = 1: PS1 = 0
256 x 30
PALETTE
PALETTE SELECT BITS
MR16 MR17
1
0
WINDOW 1
(Pseudo-Color)
PS0 = 0: PS1 = 1
RGB
ANALOG
VIDEO
ADV7151 (3)
256 x 30
PALETTE
PALETTE SELECT BITS
MR16 MR17
1
1
RGB
ANALOG
VIDEO
Multiple Devices Driving a Multiwindow Application
REV. B
–27–
WINDOW 3
(Pseudo-Color)
PS0 = 1: PS1 = 1
ADV7152
APPENDIX 5
INITIALIZATION AND PROGRAMMING
ADV7152 Initialization
The following section gives examples of initialization of the
ADV7152 operating in various modes.
After power has been supplied, the ADV7152 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
Multiplexing
Databus
RAM-DAC Resolution
SYNC
Pedestal
24-Bit True Color
2:1
8-Bit
8-Bit
Enabled on IOG
7.5 IRE
Register Initialization
Write 09H to Mode Register (MR1)
Write 08H to Mode Register (MR1)
Write 09H to Mode Register (MR1)
Write 29H to Mode Register (MR1)
Write 09H to Mode Register (MR1)
Write 04H to Address Register (A7–A0)
Write FFH to Pixel Mask Register
Write 05H to Address Register (A7–A0)
Write 00H to Command Reg 1 (CR1)
Write 06H to Address Register (A7–A0)
Write ECH to Command Reg 2 (CR2)
Write 07H to Address Register (A7–A0)
Write 40H to Command Reg 3 (CR3)
C1
1
1
1
1
1
0
1
0
1
0
1
0
1
C0
1
1
1
1
1
0
0
0
0
0
0
0
0
R/W
0
0
0
0
0
0
0
0
0
0
0
0
0
Comment
Resets to Normal Operation, 8-Bit Bus/RAM-DAC
*(Initializes Pipelining
*( “
*(Calibrates LOADOUT/LOADIN Timing
*( “
Address Reg Points to Pixel Mask Register
Sets the Pixel Mask to All “1s”
Address Reg Points to Command Register 1 (CR1)
Address Reg Points to Command Register 2 (CR2)
Sets 24-Bit Color, 7.5 IRE, SYNC on Green (IOG)
Address Reg Points to Command Register 3 (CR3)
Sets 2:1 Multiplexing, PRGCKOUT = CLOCK/4
Color Palette RAM Initialization
C1
C0
R/W
Comment
Write
Write
Write
Write
Write
Write
Write
•
•
Write
Write
Write
0
0
0
0
0
0
0
•
•
0
0
0
0
1
1
1
1
1
1
•
•
1
1
1
0
0
0
0
0
0
0
•
•
0
0
0
Points to Color Palette RAM
(Initializes Palette RAM
(
to a Linear Ramp**
(
(
(
(
(
(
(
(
(RAM Initialization Complete
00H to Address Register (A7–A0)
00H (Red Data) to RAM Location (00H)
00H (Green Data) to RAM Location (00H)
00H (Blue Data) to RAM Location (00H)
01H (Red Data) to RAM Location (01H)
01H (Green Data) to RAM Location (01H)
01H (Blue Data) to RAM Location (01H)
•
•
•
•
•
•
•
•
FFH (Red Data) to RAM Location (FFH)
FFH (Green Data) to RAM Location (FFH)
FFH (Blue Data) to RAM Location (FFH)
**These four command lines reset the ADV7152. 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.
–28–
REV. B
ADV7152
Example 2
Color Mode
Multiplexing
Databus
RAM-DAC Resolution
SYNC
Pedestal
Calibration
24-Bit Gamma Corrected True Color (30 Bits)
2:1
10-Bit
10-Bit
Ignored
0 IRE
Every Vertical Sync
Register Initialization
C1
C0
R/W
Comment
Write
Write
Write
Write
Write
Write
Write
Write
Write
Write
Write
Write
Write
1
1
1
1
1
0
1
0
0
0
1
0
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Resets to Normal Operation, 10-Bit Bus/RAM-DAC
*(Initializes Pipelining
*( “
*(Calibrates LOADOUT/LOADIN Timing
*( “
Address Reg Points to Pixel Mask Register
Sets the Pixel Mask to All “1s”
Address Reg Points to Command Register 1 (CR1)
Calibrates Every Vertical Sync
Address Reg Points to Command Register 2 (CR2)
Sets 24-Bit Color, 0 IRE, No SYNC
Address Reg Points to Command Register 3 (CR3)
Sets 2:1 Multiplexing, PRGCKOUT = CLOCK/8
C1
0
0
0
0
0
0
0
•
•
0
0
0
C0
0
1
1
1
1
1
1
•
•
1
1
1
R/W
0
0
0
0
0
0
0
•
•
0
0
0
Comment
Points to Color Palette RAM
(Initializes Palette RAM
( to a “Gamma” Ramp**
(
(
(
(
(
(
(
(
(RAM Initialization Complete
0FH to Mode Register (MR1)
0EH to Mode Register (MR1)
0FH to Mode Register (MR1)
2FH to Mode Register (MR1)
0FH to Mode Register (MR1)
04H to Address Register (A7–A0)
FFH to Pixel Mask Register
05H to Address Register (A7–A0)
01H to Command Reg 1 (CR1)
06H to Address Register (A7–A0)
E0H to Command Reg 2 (CR2)
07H to Address Register (A7–A0)
41H to Command Reg 3 (CR3)
Color Palette RAM Initialization
Write 00H to Address Register (A7–A0)
Write 000H (Red Data) to RAM Location (00H)
Write 000H (Green Data) to RAM Location (00H)
Write 000H (Blue Data) to RAM Location (00H)
Write xxxH (Red Data) to RAM Location (01H)
Write xxxH (Green Data) to RAM Location (01H)
Write xxxH (Blue Data) to RAM Location (01H)
•
•
•
•
•
•
•
•
•
•
Write 3FFH (Red Data) to RAM Location (FFH)
Write 3FFH (Green Data) to RAM Location (FFH)
Write 3FFH (Blue Data) to RAM Location (FFH)
**These four command lines reset the ADV7152 The pipelines for each of the Red, Green 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.
**Data for a gamma curve characteristic is obtainable in Appendix 3.
REGISTER DIAGNOSTIC TESTING
The previous examples show the register initialization sequence
for the ADV7152. 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
1
1
0
1
1
0
0
1
0
0
1
06H
E0H
E0H
07H
40H
40H
Select Command Register 2 (CR2)
Sets 24-Bit True-Color
Command Reg 2 Value Read-Back
Select Command Register 3 (CR3)
Set 2:1 Mux Mode
Command Reg 3 Value Read-Back
0
0
0
0
0
0
REV. B
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
1
0
1
0
1
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
06H
E0H
07H
40H
06H
E0H
07H
40H
40H
Select Command Register 2 (CR2)
Sets 24-Bit True-Color
Select Command Register 3 (CR3)
Set 2:1 Mux Mode
Select CR2
CR2 Value Read-Back
Select CR3
CR3 Value Read-Back
CR3 Value Read-Back
It is clear that this latter case requires more command lines
than the previous READ after each WRITE case.
–29–
ADV7152
APPENDIX 6
TEST DIAGNOSTICS
SYNC
BLANK
PIXEL
DATA
GRAPHICS PIPELINE
INPUT
MUX
COLOR
PALETTE
RAM
GRAPHICS PIPELINE
TRIGGER
DECODE
DACs
TRIGGER
DECODE
SYNC BLANK
COLOR
REGISTERS
PIXEL TEST
REGISTER
DAC TEST
REGISTERS
IPLL TEST
REGISTER
MPU PORT
CE
R/W
C0
C1
D9–D0
Test/Diagnostic Block Diagram
The ADV7152 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 ADV7152. 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.
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.
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.
Pixel Test Register
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 ADV7152 preset to sample the graphics pipeline on a
low to high transition of R7.
Pixel 0:
Pixel 1:
Pixel 2:
Pixel 3:
. . . .
. . . .
Pixel n-1:
Pixel n:
Pixel n:
RED
GREEN
BLUE
00000000
0........
1........
0........
. . .
. . .
0........
1........
0........
00000000
........
........
........
00000000
........
........
........
........
........
........
........
........
........
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.
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 IPLL Test Register
This is an 8-bit wide register but with only three effective bits.
The three lower bits correspond to SYNC, BLANK and IPLL
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 SYNC, BLANK and
IPLL is latched into this register. It is read over the MPU Port as
described in the “Microprocessor (MPU) Port” section.
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
(Note: If BLANK is low, the corresponding pixel data to the
DAC Test Register will be all “0s.”)
–30–
REV. B
ADV7152
APPENDIX 7
THERMAL AND ENVIRONMENTAL CONSIDERATIONS
The ADV7152 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 ADV7152 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 published herein may not be the most up to date at the time of reading
this. Advances in package compounds and manufacture will inevitably lead to improvements in the thermal data. Please contact your
local sales office for the most up-to-date information.
Power Dissipation
The diagram shows graphs of power dissipation in watts vs.
pixel clock frequency for the ADV7152.
1.50
POWER DISSIPATION – Watts
VAA = +5V
VREF = +1.2V
TA = 25°C
1.25
Table A. Thermal Characteristics vs. Airflow
Air Velocity
(Linear feet/min)
θJA (°C/W)
No Heatsink
35
EG&G D10100-28 Heatsink 32
Thermalloy 2290 Heatsink
25
50
100
200
31
28
21
28
25
18
25
22
15
Thermal Model
The junction temperature of the device in a specific application
is given by:
TJ = TA + PD (θJC + θCA)
or
TJ = TA + PD (θJA)
(1)
(2)
where:
TJ = Junction Temperature of Silicon (°C)
TA = Ambient Temperature (°C)
PD = Power Dissipation (W)
θJC = Junction to Case Thermal Resistance (°C/W)
θCA = Case to Ambient Thermal Resistance (°C/W)
θJA = Junction to Ambient Thermal Resistance (°C/W)
Package Enhancements
The standard QFP package has been enhanced to a PowerQuad2
package. This supports an improved thermal performance come
pared 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 ADV7152 to the Printed Circuit Board via the leads.
1.00
0.75
Heatsinks
0.50
60
80
100
120
140
160
180
PIXEL CLOCK FREQUENCY – MHz
200
220
* 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.
Typical Power Dissipation vs. Pixel Rate
The maximum silicon junction temperature should be limited to
100°C. Temperatures greater than this will reduce long term
device reliability. To ensure that the silicon junction temperature stays within prescribed limits, the addition of an external
heatsink may be necessary. Heatsinks, will reduce θJA as shown
in the “Thermal Characteristics vs. Airflow” table.
Package Characteristics
The table of thermal characteristics shows typical information
for the ADV7152 (100-Lead Plastic Power QFP) using various
values of Airflow.
Junction to Case (θJC) Thermal Resistance for this particular
part is:
θJC (100-Lead Plastic Power QFP) = 1.0°C/W
(Note: 0°C is independent of airflow.)
REV. B
0
(Still Air)
–31–
ADV7152
APPENDIX 8
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
C1695–10–8/94
100-Lead Plastic Power Quad Flatpack (S-100)
0.952 (24.15)
0.932 (23.65)
0.792 (20.10)
0.784 (19.90)
80
81
51
50
0.715 (18.15)
0.696 (17.65)
TOP VIEW
0.556 (14.10)
0.548 (13.90)
10°
6°±4°
PIN 1
100
4°±4° MAX
0.014 (0.35)
0.011 (0.27)
0.030 (0.75)
0.022 (0.55)
0.054 (1.37)
0.046 (1.17)
0.134
(3.40)
MAX
0.116 (2.94)
0.096 (2.44)
0.054 (1.37)
0.046 (1.17)
0.037 (0.95)
0.026 (0.65)
PRINTED IN U.S.A.
0.004
(0.102)
MAX
31
30
1
–32–
REV. B