FEATURES
220 MHz, 24-Bit (30-Bit Gamma Corrected) True Color
Triple 10-Bit “Gamma Correcting” D/A Converters
Triple 256 3 10 (256 3 30) Color Palette RAM
On-Chip Clock Control Circuit
Palette Priority Select Registers
RS-343A/RS-170 Compatible Analog Outputs
TTL Compatible Digital Inputs
Standard MPU l/O Interface
10-Bit Parallel Structure
8+2 Byte Structure
Programmable Pixel Port: 24-Bit, 15-Bit and
Programmable Pixel Port: 8-Bit (Pseudo)
Pixel Data Serializer
Multiplexed Pixel Input Ports; 1:1, 2:1, 4:1
+5 V CMOS Monolithic Construction
160-Lead Plastic Quad Flatpack (QFP)
Thermally Enhanced to Achieve u
MODES OF OPERATION
24-Bit True Color (30-Bit Gamma Corrected)
@ 220 MHz
@ 170 MHz
@ 135 MHz
@ 110 MHz
< 1.08C/W
JC
Triple 10-Bit Video RAM-DAC
ADV7150
@ 85 MHz
8-Bit Pseudo Color
15-Bit True Color
APPLICATIONS
High Resolution, True Color Graphics
Professional Color Prepress Imaging
GENERAL DESCRIPTION
The ADV7150 (ADV®) is a complete analog output, Video
RAM-DAC on a single CMOS monolithic chip. The part is specifically designed for use in high performance, color graphics
workstations. The ADV7150 integrates a number of graphic
functions onto one device allowing 24-bit direct True-Color operation at the maximum screen update rate of 220 MHz. The
ADV7150 implements 30-bit True Color in 24-bit frame buffer
designs. The part also supports other modes, including 15-bit
True Color and 8-bit Pseudo or Indexed Color. Either the Red,
Green or Blue input pixel ports can be used for Pseudo Color.
(Continued on page 12)
ADV is a registered trademark of Analog Devices, Inc.
FUNCTIONAL BLOCK DIAGRAM
V
AA
RED (R7–R0),
GREEN (G7–G0),
BLUE (B7–B0)
COLOR DATA
PALETTE
SELECTS
(PS0, PS1)
LOADIN
LOADOUT
PRGCKOUT
SCKOUT
SCKIN
SYNC
BLANK
CLOCK
CLOCK
24
A
B
C
D
P
24
X
E
24
L
P
O
24
R
T
8
CLOCK CONTROL
CLOCK DIVIDE
SYNCHRONIZATION
÷32 ÷16, ÷8, ÷4, ÷2
ECL TO CMOS
I
&
CIRCUIT
96
8
ADV7150
MUX
4:1
MUX
4:1
ADDRESS
REGISTER
8
8
8
2
MODE
REGISTER
ADDR
(A7–A0)
CE R/W C0 C1
(MR1)
REV. A
Information furnished by Analog Devices is believed to be accurate and
reliable. However, no responsibility is assumed by Analog Devices for its
use, nor for any infringements of patents or other rights of third parties
which may result from its use. No license is granted by implication or
otherwise under any patent or patent rights of Analog Devices.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 617/329-4700Fax: 617/326-8703
ADV7150–SPECIFICA TIONS
CL = 10 pF); IOR, IOG, IOB = GND. All specifications T
MIN
1
(V
= +5 V; V
AA
2
to T
unless otherwise noted.)
MAX
= +1.235 V; R
REF
= 280 V. IOR, IOG, IOB (RL = 37.5 V,
SET
ParameterAll VersionsUnitTest Conditions/Comments
STATIC PERFORMANCE
Resolution (Each DAC)10Bits
Accuracy (Each DAC)
Integral Nonlinearity±1LSB max
Differential Nonlinearity±1LSB maxGuaranteed Monotonic
Gray Scale Error±5% Gray Scale max
CodingBinary
DIGITAL INPUTS (Excluding CLOCK, CLOCK)
Input High Voltage, V
Input Low Voltage, V
Input Current, I
Input Capacitance, C
INL
IN
IN
INH
2V min
0.8V max
±10µA maxVIN = 0.4 V or 2.4 V
10pF max
CLOCK INPUTS (CLOCK, CLOCK)
Input High Voltage, V
Input Low Voltage, V
Input Current, I
Input Capacitance, C
INL
IN
IN
INH
VAA – 1.0V min
VAA – 1.6V max
±10µA maxVIN = 0.4 V or 2.4 V
10pF typ
DIGITAL OUTPUT
Output High Voltage, V
Output Low Voltage, V
OL
OH
2.4V minI
0.4V maxI
SOURCE
= 3.2 mA
SINK
= 400 µA
Floating-State Leakage Current20µA max
Floating-State Output Capacitance20pF typ
ANALOG OUTPUTS
Gray Scale Current Range15/22mA min/max
Output Current
White Level Relative to Blank17.69/20.40mA min/maxTypically 19.05 mA
White Level Relative to Black16.74/18.50mA min/maxTypically 17.62 mA
Black Level Relative to Blank0.95/1.90mA min/maxTypically 1.44 mA
Blank Level on IOR, IOB0/50µA min/maxTypically 5 µA
Blank Level on IOG6.29/8.96mA min/maxTypically 7.62 mA
Sync Level on IOG0/50µA min/maxTypically 5 µA
LSB Size17.22µA typ
DAC-to-DAC Matching3% maxTypically 1%
Output Compliance, V
Output Impedance, R
OUT
Output Capacitance, C
OC
OUT
0/+1.4V min/V max
100kΩ typ
30pF maxI
OUT
= 0 mA
VOLTAGE REFERENCE
Voltage Reference Range, V
Input Current, I
VREF
REF
1.14/1.26V min/V maxV
+5µA typ
= 1.235 V for Specified Performance
REF
POWER REQUIREMENTS
V
AA
3
I
AA
3
I
AA
I
AA
I
AA
I
AA
5V nom
400mA max220 MHz Parts
370mA max170 MHz Parts
350mA max135 MHz Parts
330mA max110 MHz Parts
315mA max85 MHz Parts
Power Supply Rejection Ratio0.5%/% maxTypically 0.12%/%: COMP = 0.1 µF
DYNAMIC PERFORMANCE
Clock and Data Feedthrough
Glitch Impulse50pV secs typ
DAC-to-DAC Crosstalk
NOTES
1
±5% for all versions.
2
Temperature range (T
3
Pixel Port is continuously clocked with data corresponding to a linear ramp. TJ = 100°C.
4
Clock and data feedthrough is a function of the amount of overshoot and undershoot on the digital inputs. Glitch impulse includes clock and data feedthrough.
5
TTL input values are 0 to 3 volts, with input rise/fall times ≤ 3 ns, measured the 10% and 90% points. Timing reference points at 50% for inputs and outputs.
6
DAC-to-DAC crosstalk is measured by holding one DAC high while the other two are making low-to-high and high-to-low transitions.
Specifications subject to change without notice.
MIN
to T
4, 5
6
): 0°C to +70°C; TJ (Silicon Junction Temperature) ≤ 100°C.
MAX
–30dB typ
–23dB typ
–2–
REV. A
TIMING CHARACTERISTICS
IOR, IOG, I0B = GND. All specifications T
CLOCK CONTROL AND PIXEL PORT
MIN
to T
4
= +5 V; V
AA
3
unless otherwise noted.)
MAX
= +1.235 V; R
REF
= 280 V. IOR, IOG, IOB (RL = 37.5 V, CL = 10 pF);
SET
1
2
(V
220 MHz 170 MHz 135 MHz 110 MHz 85 MHz
ParameterVersionVersionVersionVersionVersionUnits
ADV7150
Conditions/Comments
f
CLOCK
t
1
t
2
t
3
t
4
f
LOADIN
22017013511085MHz max Pixel CLOCK Rate
4.555.887.49.111.77ns minPixel CLOCK Cycle Time
22.53.244ns minPixel CLOCK High Time
22.5344ns minPixel CLOCK Low Time
1010101010ns maxPixel CLOCK to LOADOUT Delay
LOADIN Clocking Rate
1:1 Multiplexing11011011011085MHz max
2:1 Multiplexing1108567.55542.5MHz max
4:1 Multiplexing5542.533.7527.521.25MHz max
t
5
LOADIN Cycle Time
1:1 Multiplexing9.19.19.19.19.1ns min
2:1 Multiplexing9.111.7614.818.1823.53ns min
4:1 Multiplexing18.1823.5329.6336.3647.1ns min
t
6
LOADIN High Time
1:1 Multiplexing44444ns min
2:1 Multiplexing45689ns min
4:1 Multiplexing89121518ns min
t
7
LOADIN Low Time
1:1 Multiplexing44444ns min
2:1 Multiplexing45689ns min
4:1 Multiplexing89121518ns min
t
8
t
9
t
10
5
τ–t
11
6
t
PD
1:1 Multiplexing55555CLOCKs(1 × CLOCK = t
00000ns minPixel Data Setup Time
55555ns minPixel Data Hold Time
00000ns minLOADOUT to LOADIN Delay
τ–5τ–5τ–5τ–5τ–5ns maxLOADOUT to LOADIN Delay
1010101010ns maxPixel CLOCK to PRGCKOUT Delay
55555ns maxSCKIN to SCKOUT Delay
55555ns minBLANK to SCKIN Setup Time
11111ns minBLANK to SCKIN Hold Time
33333ns minR/W, C0, C1 to CE Setup Time
1010101010ns minR/W, C0, C1 to CE Hold Time
4545454545ns minCE Low Time
2525252525ns minCE High Time
55555ns minCE Asserted to Databus Driven
4545454545ns maxCE Asserted to Data Valid
2020202020ns maxCE Disabled to Databus Three-Stated
55555ns min
t
26
t
27
REV. A
2020202020ns minWrite Data (D0–D9) Setup Time
55555ns minWrite Data (D0–D9) Hold Time
–3–
ADV7150
NOTES
1
TTL input values are 0 to 3 volts, with input rise/fall times ≤ 3 ns, measured between the 10% and 90% points. ECL inputs (CLOCK, CLOCK) are
VAA–0.8 V to VAA–1.8 V, with input rise/fall times ≤ 2 ns, measured between the 10% and 90% points. Timing reference points at 50% for inputs and outputs. Analog output load ≤ 10 pF. Databus (D0–D9) loaded as shown in Figure 1. Digital output load for LOADOUT, PRGCKOUT, SCKOUT, I
SYNCOUT≤ 30 pF.
2
±5% for all versions.
3
Temperature range (T
4
Pixel Port consists of the following inputs: Pixel Inputs: RED [A, B, C, D]; GREEN [A, B, C, D]; BLUE [A, B, C, D], Palette Selects: PS0 [A, B, C, D]; PS1
These fixed values for Pipeline Delay are valid under conditions where t10 and τ-t11 are met. If either t10 or τ-t11 are not met, the part will operate but the Pipe line De-
lay is increased by 2 additional CLOCK cycles for 2:1 Mode and is increased by 4 additional CLOCK cycles for 4:1 Mode, after calibration is performed.
7
Output delay measured from the 50% point of the rising edge of CLOCK to the 50% point of full-scale transition. Output rise/fall time measured between the 10%
MIN
to T
): 0°C to +70°C; TJ (Silicon Junction Temperature) ≤ 100°C.
MAX
and 90% points of full-scale transition. Transition time measured from the 50% point of full-scale transition to the output remaining within 2% of the final output
value (Transition time does not include clock and data feedthrough).
8
t23 and t24 are measured with the load circuit of Figure 1 and defined as the time required for an output to cross 0.4 V or 2.4 V.
9
t25 is derived from the measured time taken by the data outputs to change by 0.5 V when loaded with the circuit of Figure 1. The measured number is then extrapo-
lated back to remove the effects of charging the 100 pF capacitor. This means that the time, t25, quoted in the Timing Characteristics is the true value for the device
and as such is independent of external databus loading capacitances.
Specifications subject to change without notice.
I
SINK
TO
OUTPUT
PIN
100pF
+2.1V
PLL
and
CLOCK
CLOCK
LOADOUT
(1:1 MULTIPLEXING)
LOADOUT
(2:1 MULTIPLEXING)
LOADOUT
(4:1 MULTIPLEXING)
LOADIN
PIXEL INPUT
DATA*
I
SOURCE
Figure 1. Load Circuit for Databus Access and Relinquish Times
Figure 7. Pixel Input to Analog Output Pipeline with Maximum LOADOUT to LOADIN Delay (2:1 Multiplex Mode)
–6–
REV. A
CLOCK
PRGCKOUT
(CLOCK/4)
PRGCKOUT
(CLOCK/8)
PRGCKOUT
(CLOCK/16)
PRGCKOUT
(CLOCK/32)
SCKIN
BLANK
ADV7150
t
12
Figure 8. Pixel Clock Input vs. Programmable Clock Output (PRGCKOUT)
t
13
t
15
t
14
BLANKING
PERIOD
SCKOUT
END OF SCAN LINE (N)
START OF SCAN LINE (N+1)
Figure 9. Video Data Shift Clock Input (SCKIN) & BLANK vs. Video Data Shift Clock Output (SCKOUT)
CLOCK
t
18
WHITE LEVEL
90 %
50 %
10 %
BLACK LEVEL
NOTE:
THIS DIAGRAM IS NOT TO SCALE. FOR THE PURPOSES OF
CLARITY, THE ANALOG OUTPUT WAVEFORM IS MAGNIFIED IN
TIME AND AMPLITUDE W.R.T THE CLOCK WAVEFORM.
I
AND SYNCOUT ARE DIGITAL VIDEO OUTPUT SIGNALS.
PLL
t
IS THE ONLY RELEVENT OUTPUT TIMING SPECIFICATION
16
FOR I
AND SYNCOUT.
PLL
FULL-SCALE
TRANSITION
ANALOG
OUTPUTS
IOR, IOR
IOG, IOG
IOB, IOB
I
PLL, SYNCOUT
t
16
t
17
REV. A
Figure 10. Analog Output Response vs. CLOCK
–7–
ADV7150
R/W, C0, C1
(READ MODE)
D0–D9
t
19
CONTROL DATA
CE
VALID
t
23
t
20
t
21
t
22
t
24
R/W = 1
t
25
D0–D9
(WRITE MODE)
R/W = 0
t
26
t
27
Figure 11. Microprocessor Port (MPU) Interface Timing
RECOMMENDED OPERATING CONDITION
ParameterSymbolMinTypMaxUnits
Power SupplyV
Ambient Operating TemperatureT
Reference VoltageV
Output LoadR
AA
A
REF
L
CAUTION
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily
accumulate on the human body and test equipment and can discharge without detection.
Although the ADV7150 features proprietary ESD protection circuitry, permanent damage may
occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD
precautions are recommended to avoid performance degradation or loss of functionality.
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.
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
REV. A
–9–
ADV7150
PIN FUNCTION DESCRIPTION
Mnemonic Function
RED (R0A . . . R0D–R7A . . . R7D),Pixel Port (TTL Compatible Inputs): 96 pixel select inputs, with 8 bits each for Red, 8
GREEN (G0
BLUE (B0
PS0
. . . PS0D, PS1A . . . PS1
A
LOADINPixel Data Load Input (TTL Compatible Input). This input latches the multiplexed
LOADOUTPixel Data Load Output (TTL Compatible Output). This output control signal runs at
PRGCKOUTProgrammable Clock Output (TTL Compatible Output). This output control signal
SCKINVideo Shift Clock Input (TTL Compatible Input). The signal on this input is internally
SCKOUTVideo Shift Clock Output (TTL Compatible Output). This output is a synchronously
CLOCK,
BLANKComposite Blank (TTL Compatible Input). This video control signal drives the analog
SYNCComposite-Sync Input (TTL Compatible Input). This video control signal drives the
SYNCOUTComposite-Sync Output (TTL Compatible Output). This video output is a delayed
D0–D9Databus (TTL Compatible Input/Output Bus). Data, including color palette values and
CEChip Enable (TTL Compatible Input). This input must be at Logic “0,” when writing
. . . G0D–G7A. . . G7D),bits for Green and 8 bits for Blue. Each bit is multiplexed [A-D] 4:1, 2:1 or 1:1. It can
A
. . . B0D–B7A . . . B7D)be configured for 24-Bit True-Color Data, 8-Bit Pseudo-Color Data and 15-Bit True-Color
A
Data formats. Pixel Data is latched into the device on the rising edge of LOADIN.
D
Palette Priority Selects (TTL Compatible Inputs): These pixel port select inputs determine whether or not the device’s pixel data port is selected on a pixel by pixel basis.
The palette selects allow switching between multiple palette devices. The device can be
preprogrammed to completely shut off the DAC analog outputs. If the values of PS0
and PS1 match the values programmed into bits MR16 and MR17 of the Mode Register, then the device is selected. Each bit is multiplexed [A-D] 4:1, 2:1 or 1:1. PS0 and
PS1 are latched into the device on the rising edge of LOADIN.
pixel data, including PS0–PS1,
BLANK and SYNC into the device.
a divided down frequency of the pixel CLOCK input. Its frequency is a function of the
multiplex rate. It can be used to directly or indirectly drive LOADIN
f
LOADOUT
= f
CLOCK
/M
where M = 1 for 1:1 Multiplex Mode
whereM = 2 for 2:1 Multiplex Mode
whereM = 4 for 4:1 Multiplex Mode.
runs at a divided down frequency of the pixel CLOCK input. Its frequency is user
programmable and is determined by bits CR30 and CR31 of Command Register 3
f
PRGCKOUT
= f
CLOCK
/N
where N = 4, 8, 16 and 32.
gated synchronously with the
BLANK signal. The resultant output, SCKOUT, is a
video clocking signal that is stopped during video blanking periods.
gated version of SCKIN and
BLANK. SCKOUT, is a video clocking signal that is
stopped during video blanking periods.
CLOCKClock Inputs (ECL Compatible Inputs). These differential clock inputs are designed to
be driven by ECL logic levels configured for single supply (+5 V) operation. The clock
rate is normally the pixel clock rate of the system.
outputs to the blanking level.
IOG analog output to the
CR22 in Command Re
output, otherwise the
version of
SYNC. The delay corresponds to the number of pipeline stages of the device.
SYNC level. It is only asserted during the blanking period.
gister 2 must be set if SYNC is to be decoded onto the analog
SYNC input is ignored.
device control information is written to and read from the device over this 10-bit, bidirectional databus. 10-bit data or 8-bit data can be used. The databus can be configured
for either 10-bit parallel data or byte data (8+2) as well as standard 8-bit data. Any unused bits of the databus should be terminated through a resistor to either the digital
power plane (V
) or GND.
CC
to or reading from the device over the databus (D0–D9). Internally, data is latched on
the rising edge of CE.
–10–
REV. A
ADV7150
MnemonicFunction
R/
WRead/Write Control (TTL Compatible Input). This input determines whether data is
written to or read from the device’s registers and color palette RAM. R/
be at Logic “0” to write data to the part. R/
W must be at Logic “1” and CE at Logic
“0” to read from the device.
C0, C1Command Controls (TTL Compatible Inputs). These inputs determine the type of read
or write operation being performed on the device over the databus (see Interface Truth
Table). Data on these inputs is latched on the falling edge of
IOR, IOG; IOG, IOB;Red, Green and Blue Current Outputs (High Impedance Current Sources). These RGB
IOR;
CE.
IOBvideo outputs are specified to directly drive RS-343A and RS-170 video levels into dou-
bly terminated 75 Ω loads.
IOR, IOG and IOB are the complementary outputs of IOR, IOG and IOB. These out-
puts can be tied to GND if it is not required to use differential outputs.
V
REF
Voltage Reference Input (Analog Input). An external 1.235 V voltage reference is required to drive this input. An AD589 (2-terminal voltage reference) or equivalent is recommended. (Note: It is not recommended to use a resistor network to generate the
voltage reference.)
R
SET
Output Full-Scale Adjust Control (Analog Input). A resistor connected between this pin
and analog ground controls the absolute amplitude of the output video signal. The value
of R
is derived from the full-scale output current on IOG according to the following
SET
equations:
R
(Ω) = C1 × V
SET
R
(Ω) = C2 × V
SET
/IOG (mA); SYNC on GREEN
REF
/IOG (mA); NO SYNC on GREEN.
REF
Full-Scale output currents on IOR and IOB for a particular value of R
IOR (mA)= C2 ×V
REF
(V)/R
SET
(Ω)
and
IOB (mA) = C2 ×V
REF
(V)/R
SET
(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.
COMPCompensation Pin. A 0.1 µF capacitor should be connected between this pin and VAA.
I
PLL
Phase Lock Loop Output Current (High Impedance Current Source). This output is
used to enable multiple ADV7150s along with ADV7151s to be synchronized together
with pixel resolution when using an external PLL. This output is triggered either from
the falling edge of
SYNC or BLANK as determined by bit CR21 of Command Register
2. When activated, it supplies a current corresponding to:
I
(mA) = 1,728 × V
PLL
When not using the I
V
AA
Power Supply (+5 V ± 5%). The part contains multiple power supply pins, all should be
function, this output pin should be tied to GND.
PLL
REF
(V)/R
SET
(Ω)
connected together to one common +5 V filtered analog power supply.
GNDAnalog Ground. The part contains multiple ground pins, all should be connected
together to the system’s ground plane.
W and CE must
are given by:
SET
REV. A
–11–
ADV7150
A
B
C
D
MULTIPLEXER
24
24
24
24
24
RED
GREEN
BLUE
8
8
8
(Continued from page 1)
The device consists of three, high speed, 10-bit, video D/A converters (RGB), three 256 × 10 (one 256 × 30) color look-up
tables, palette priority selects, a pixel input data multiplexer/
serializer and a clock generator/divider circuit. The ADV7150 is
capable of 1:1, 2:1 and 4:1 multiplexing. The onboard palette
priority select inputs enable multiple palette devices to be connected together for use in multipalette and window applications.
The part is controlled and programmed through the microprocessor (MPU) port. The part also contains a number of onboard
test registers, associated with self diagnostic testing of the device. The individual Red, Green and Blue pixel input ports allow True-Color, image rendition. True-Color image rendition,
at speeds of up to 220 MHz, is achieved through the use of the
onboard data multiplexer/serializer. The pixel input port’s flexibility allows for direct interface to most standard frame buffer
memory configurations.
The 30 bits of resolution, associated with the color look-up table
and triple 10-bit DAC, realizes 24-bit True-Color resolution,
while also allowing for the onboard implementation of linearization algorithms, such as Gamma-Correction. This allows effective 30-Bit True-Color operation.
CIRCUIT DETAILS AND OPERATION
OVERVIEW
Digital video or pixel data is latched into the ADV7150 over the
devices Pixel Port. This data acts as a pointer to the onboard
Color Palette RAM. The data at the RAM address pointed to is
latched into the digital-to-analog converters (DACs) and output
as an RGB analog video signal.
For the purposes of clarity of description, the ADV7150 is broken down into three separate functional blocks. These are:
1. Pixel port and clock control circuit
2. MPU port, registers and color palette
3. Digital-to-analog converters and video outputs
Table I shows the architectural and packaging differences be-
tween other devices in the ADV715x series of workstation parts.
(For more details consult the relevant data sheets.)
Table I. Architectural and Packaging Differences of the
ADV715x Series
*See ADV7151 and ADV7150 data sheets for more information on these parts.
The on-chip video clock controller circuit generates all the internal clocking and some additional external clocking signals. An
external ECL oscillator source with differential outputs is all
that is required to drive the CLOCK and
CLOCK inputs of the
ADV7150. The part can also be driven by an external clock generator chip circuit, such as the AD730.
The ADV7150 is capable of generating RGB video output signals which are compatible with RS-343A and RS-170 video
standards, without requiring external buffering.
Test diagnostic circuitry has been included to complement the
users system level debugging.
The ADV7150 is fabricated in a +5 V CMOS process. Its
monolithic CMOS construction ensures greater functionality
with low power dissipation.
The ADV7150 is packaged in a plastic 160-pin power quad flatpack (QFP). Superior thermal dissipation is achieved by inclusion of a copper heatslug, within the standard package outline to
which the die is attached.
Pixel Port and Clock Control Circuit
The Pixel Port of the ADV7150 is directly interfaced to the
video/graphics pipeline of a computer graphics subsystem. It is
connected directly or through a gate array to the video RAM of
the systems Frame-Buffer (video memory). The pixel port on
the device consists of:
Color DataRED, GREEN, BLUE
Pixel Controls
SYNC, BLANK
Palette SelectsPS0–PS1
The associated clocking signals for the pixel port include:
Clock InputsCLOCK, CLOCK,
LOADIN, SCKIN
Clock OutputsLOADOUT, 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 & ClockControl Circuit of the part operational.
Pixel Port (Color Data)
The ADV7150 has 96 color data inputs. The part has four (for
4:1 multiplexing) 24-bit wide direct color data inputs. These are
user programmed to support a number of color data formats including 24-Bit True Color, 15-Bit True Color and 8-Bit Pseudo
Color (see “Color Data Formats” section) in 4:1, 2:1 and 1:1
multiplex modes.
Figure 12. Multiplexed Color Inputs for the ADV7150
–12–
REV. A
ADV7150
CLOCK
ADV7150
CLOCK
DIVIDE BY
N (÷ N)
LOADOUT
DIVIDE BY
M (÷ M)
PRGCKOUT
LOADIN
SCKOUT
SCKIN
BLANK
LATCH
ENABLE
SYNC
TO COLOR DATA
MULTIPLEXER
ECL
TO
TTL
M IS A FUNCTION OF MULTIPLEX RATE
M = 4 IN 4:1 MULTIPLEX MODE
M = 2 IN 2:1 MULTIPLEX MODE
M = 1 IN 1:1 MULTIPLEX MODE
N IS INDEPENDENTLY
PROGRAMMABLE
N= (4, 8, 16, 32)
Color data is latched into the parts pixel port on every rising
edge of LOADIN (see Timing Waveform, Figure 3). The
required frequency of LOADIN is determined by the multiplex
rate, where:
f
LOADIN
f
LOADIN
f
LOADIN
= f
= f
= f
/44:1 Multiplex Mode
CLOCK
/22:1 Multiplex Mode
CLOCK
CLOCK
1:1 Multiplex Mode
Other pixel data signals latched into the device by LOADIN
include
SYNC, BLANK and PS0–PS1.
Internally, data is pipelined through the part by the differential
pixel clock inputs, CLOCK and
CLOCK. The LOADIN control signal needs only have a frequency synchronous relationship
to the pixel CLOCK (see “Pipeline Delay & Onboard Calibration” section). A completely phase independent LOADIN signal
can be used with the ADV7150, allowing the CLOCK to occur
anywhere during the LOADIN cycle.
Alternatively, the LOADOUT signal of the ADV7150 can be
used. LOADOUT can be connected either directly or indirectly
to LOADIN. Its frequency is automatically set to the correct
LOADIN requirement.
SYNC, BLANK
The BLANK and SYNC video control signals drive the analog
outputs to the blanking and
SYNC levels respectively. These
signals are latched into the part on the rising edge of LOADIN.
The
SYNC information is encoded onto the IOG analog signal
when Bit CR22 of Command Register 2 is set to a Logic “1.”
The
SYNC input is ignored if CR22 is set to “0.”
SYNCOUT
In some applications where it is not permissible to encode
SYNC on green (IOG), SYNCOUT can be used as a separate
TTL digital
pendent (of the ADV7150)
SYNC output. This has the advantage over an inde-
SYNC in that it does not necessitate
knowing the absolute pipeline delay of the part. This allows
complete independence between LOADIN/Pixel Data and
CLOCK. The
SYNC input is connected to the device as normal
with Bit CR22 of Command Register 2 set to “0” thereby preventing
Command Register 1 is set to “1,” enabling
output signal generates a TTL
delay that is capable of directly driving the composite
SYNC from being encoded onto IOG. Bit CR12 of
SYNCOUT. The
SYNCOUT with correct pipeline
SYNC
signal of a computer monitor.
PS0–PS1 (Palette Priority Select Inputs)
These pixel port select inputs determine whether or not the device is selected. These controls effectively determine whether the
devices RGB analog outputs are turned-on or shut down. When
the analog outputs are shut down, IOR, IOG and IOB are
forced to 0 mA regardless of the state of the pixel and control
data inputs. This state is determined on a pixel by pixel basis as
the PS0–PS1 inputs are multiplexed in exactly the same format
as the pixel port color data. These controls allow for switching
between multiple palette devices (see Appendix 4). If the values
of PS0 and PS1 match the values programmed into bits MR16
and MR17 of the Mode Register, then the device is selected, if
there is no match the device is effectively shut down.
Multiplexing
The onboard multiplexers of the ADV7150 eliminate the need
for external data serializer circuits. Multiple video memory
devices can be connected, in parallel, directly to the device.
VIDEO MEMORY/ FRAME BUFFER
VRAM (BANK A)
VRAM (BANK B)
VRAM (BANK C)
VRAM (BANK D)
24
33MHz
24
33MHz
24
33MHz
24
33MHz
ADV7150
MULTIPLEXER
24
132 MHz
(4 x 33 MHz)
Figure 13. Direct Interfacing of Video Memory to ADV7150
Figure 13 shows four memory banks of 33 MHz memory connected to the ADV7150, running in 4:1 multiplex mode, giving
a resultant pixel or dot clock rate of 132 MHz. As mentioned in
the previous section, the ADV7150 supports a number of color
data formats in 4:1, 2:1 and 1:1 multiplex modes.
In 1:1 multiplex mode, the ADV7150 is clocked using the
LOADIN signal. This means that there is no requirement for differential ECL inputs on CLOCK and
CLOCK. The pixel clock is
connected directly to LOADIN. (Note: The ECL CLOCK can
still be used to generate LOADOUT PRGCKOUT, etc.)
CLOCK CONTROL CIRCUIT
The ADV7150 has an integrated Clock Control Circuit (Figure
14). This circuit is capable of both generating the ADV7150’s
internal clocking signals as well as external graphics subsystem
clocking signals. Total system synchronization can be attained
by using the parts output clocking signals to drive the controlling graphics processor’s master clock as well as the video frame
buffers shift clock signals.
Figure 14. Clock Control Circuit of the ADV7150
REV. A
–13–
ADV7150
LOADOUT
LOADIN
ADV7150
VIDEO
FRAME
BUFFER
LOADOUT
LOADIN
ADV7150
VIDEO
FRAME
BUFFER
LOADOUT(1)
LOADOUT(2)
PIXEL
DATA
PIXEL
DATA
LOADIN
LOADOUTLOADOUT(1)
LOADOUT(2)
DELAY
SCKOUT
SCKIN
BLANK
LATCH
ENABLE
SYNC
LOADOUT
SCKOUT
ADV7150
VIDEO
FRAME
BUFFER
PIXEL
DATA
LOADIN
SCKIN
BLANK
CLOCK, CLOCK Inputs
The Clock Control Circuit is driven by the pixel clock inputs,
CLOCK and
ential ECL oscillator running from a +5 V supply.
Alternatively, the ADV7150 CLOCK inputs can be driven by a
Programmable Clock Generator (Figure 15), such as the
ICS1562. The ICS1562 is a monolithic, phase-locked-loop,
clock generator chip. It is capable of synthesizing differential
ECL output frequencies in a range up to 220 MHz from a single
low frequency reference crystal.
LOW FREQUENCY
OSCILLATOR
V
CLOCK
Figure 15. PLL Generator Driving CLOCK,
ADV7150
CLOCK CONTROL SIGNALS
LOADOUT
The ADV7150 generates a LOADOUT control signal which
runs at a divided down frequency of the pixel CLOCK. The
frequency is automatically set to the programmed multiplex
rate, controlled by CR37 and CR36 of Command Register 3.
f
LOADOUT
f
LOADOUT
f
LOADOUT
The LOADOUT signal is used to directly drive the LOADIN
pixel latch signal of the ADV7150. This is most simply achieved
by tying the LOADOUT and LOADIN pins together. Alternatively, the LOADOUT signal can be used to drive the frame
buffer’s shift clock signals, returning to the LOADIN input delayed with respect to LOADOUT.
If it is not necessary to have a known fixed number of pipeline
delays, then there is no limitation on the delay between LOADOUT and LOADIN (LOADOUT(1) and LOADOUT(2)).
LOADIN and Pixel Data must conform to the setup and hold
times (t
If, however, it is required that the ADV7150 has a fixed number
of pipeline delays (t
form to timing specifications t
ures 4 to 7.
CLOCK. These inputs can be driven by a differ-
GND
+5V
ECL
ECL
CLOCK
GENERATOR
V
REF
D0-D3CSR/W
= f
CLOCK
= f
CLOCK
= f
CLOCK
and t9).
8
OUT+
OUT–
OUT V
/44:1 Multiplex Mode
/22:1 Multiplex Mode
), LOADOUT and LOADIN must con-
PD
V
GND
GND
1:1 Multiplex Mode
10
V
V
CC
220Ω
330Ω
CC
GND
220Ω
330Ω
0.1 µF
V
AA
CLOCK
and τ-t11 as illustrated in Fig-
AA
+5V
CLOCK
CLOCK
ADV7150
REF
GND
of the
Figure 16. LOADOUT vs. Pixel Clock Input (CLOCK,
CLOCK
PRGCKOUT
The PRGCKOUT control signal outputs a user programmable
clock frequency. It is a divided down frequency of the pixel
CLOCK (see Figure 8). The rising edge of PRGCKOUT is
synchronous to the rising edge of LOADOUT
f
PRGCKOUT
= f
CLOCK
/N
where N = 4, 8, 16 or 32.
One application of the PRGCKOUT is to use it as the master
clock frequency of the graphics subsystems processor or
controller.
SCKIN, SCKOUT
These video memory signals are used to minimize external support chips. Figure 17 illustrates the function that is provided.
An input signal applied to SCKIN is synchronously AND-ed
with the video blanking signal (
BLANK). The resulting signal is
output on SCKOUT. Figure 9 of the Timing Waveform section
shows the relationship between SCKOUT, SCKIN and
BLANK.
Figure 17. SCKOUT Generation Circuit
The SCKOUT signal is essentially the video memory shift control signal. It is stopped during the screen retrace. Figure 18
shows a suggested frame buffer to ADV7150 interface. This is a
minimum chip solution and allows the ADV7150 control the
overall graphics system clocking and synchronization.
Figure 18. ADV7150 Interface Using SCKIN and SCKOUT
–14–
REV. A
)
8
8-BIT
RED DAC
8-BIT
BLUE DAC
RED
256 x 8
GREEN
256 x 8
BLUE
256 x 8
8
8-BIT
GREEN DAC
8
8
8
8
RED
OUT
GREEN
OUT
BLUE
OUT
24-BIT
COLOR DATA
24-BIT TO 24-BIT
LOOK-UP TABLE
24-BIT
COLOR DATA
ANALOG VIDEO
OUTPUTS
10
10-BIT
RED DAC
10-BIT
BLUE DAC
RED
256 x 10
GREEN
256 x 10
BLUE
256 x 10
10-BIT
GREEN DAC
10
10
8
RED
OUT
GREEN
OUT
BLUE
OUT
8-BIT PIXEL
DATA
8-BIT TO 30-BIT
LOOK-UP TABLE
30-BIT
COLOR DATA
ANALOG VIDEO
OUTPUTS
8
8-BIT
RED DAC
8-BIT
BLUE DAC
RED
256 x 8
GREEN
256 x 8
BLUE
256 x 8
8-BIT
GREEN DAC
8
8
8
RED
OUT
GREEN
OUT
BLUE
OUT
8-BIT PIXEL
DATA
8-BIT TO 24-BIT
LOOK-UP TABLE
24-BIT
COLOR DATA
ANALOG VIDEO
OUTPUTS
Pipeline Delay and Onboard Calibration
The ADV7150 has a fixed number of pipeline delays (tPD), so
long as timings t
delay is not a requirement, timings t
and τ-t11 are met. However, if a fixed pipeline
10
and τ-t11 can be ignored,
10
a calibration cycle must be run and there is no restriction on
LOADIN to LOADOUT timing. If timings t
and τ-t11 are not
10
met, the part will function correctly though with an increased
number of pipeline delays, t
+ N CLOCKS (for 4:1 mode
PD
N = 4, for 2:1 mode N = 2, for 1:1 mode N = 0). The ADV7150
has onboard calibration circuitry which synchronizes pixel data and
LOADIN with the internal ADV7150 clocking signals. Calibration can be performed in two ways: during the devices initialization sequence by toggling two bits of the Mode Register, MR10
followed by MR15, or by writing a “1” to Bit CR10 of Command
Register 1 which executes a calibration on every Vertical Sync.
COLOR VIDEO MODES
The ADV7150 supports a number of color video modes all at
the maximum video rate.
Command bits CR24–CR27 of Command Register 2 along with
Bit MR11 of Mode Register 1 determine the color mode.
24-Bit “Gamma” True Color
(CR25, CR26, CR27 = 1, 1, 1 and MR11 = 1)
The part is set to 24-bit/30-bit True-Color operation. The pixel
port accepts 24 bits of color data which is directly mapped to
the Look-Up Table RAM. The Look-Up Table is configured as
a 256 location by 30 bits deep RAM (10 bits each for Red,
Green and Blue). The output of the RAM drives the DACs with
30-bit data (10 bits each for Red, Green and Blue). The RAM is
preloaded with a user determined, nonlinear function, such as a
gamma correction curve.
10-BIT
RED DAC
10-BIT
GREEN DAC
10-BIT
BLUE DAC
ANALOG VIDEO
OUTPUTS
RED
OUT
GREEN
OUT
BLUE
OUT
24-BIT
COLOR DATA
8
8
8
24-BIT TO 30-BIT
LOOK-UP TABLE
RED
256 x 10
GREEN
256 x 10
BLUE
256 x 10
30-BIT
COLOR DATA
10
10
10
Figure 19. 24-Bit to 30-Bit True-Color Configuration
This mode allows for the display of full 24-bit, GammaCorrected True-Color Images.
24-Bit “Standard” True Color
(CR25, CR26, CR27 = 1, 1, 1 and MR11 = 0)
This mode sets the part into direct 24-bit True-Color operation.
The pixel port accepts 24 bits of color data which is directly
mapped to Look-Up Table RAM. The Look-Up Table is configured as a 256 location by 24 bits deep RAM (8 bits each for
Red, Green and Blue) and essentially acts as a bypass RAM.
The output of the RAM drives the DACs with 24-bit data (8
bits each for Red, Green and Blue). The RAM is preloaded with
a linear function.
This mode allows for the display of full 24-bit True-Color
Images.
REV. A
–15–
ADV7150
Figure 20. 24-Bit to 24-Bit Direct True-Color Configuration
8-Bit “Gamma” Pseudo Color
(CR25, CR26, CR27 = X, 0, 0 or X, 1, 0 or X, 0, 1 and
MR11 = 1)
This mode sets the part into 8-bit Pseudo-Color operation. The
pixel port accepts 8 bits of pixel data which indexes a 30-bit
word in the Look-Up Table RAM. The Look-Up Table is configured as a 256 location by 30 bits deep RAM (10 bits each for
Red, Green and Blue). The output of the RAM drives the
DACs with 30-bit data (10 bits each for Red, Green and Blue).
Figure 21. 8-Bit to 30-Bit Pseudo-Color Configuration
This mode allows for the display of 256 simultaneous colors out
of a total palette of millions of addressable colors.
8-Bit “Standard” Pseudo Color
(CR25, CR26, CR27 = X, 0, 0 or X, 1, 0 or X, 0, 1 and
MR11 = 0)
This mode sets the part into 8-bit Pseudo-Color operation. The
pixel port accepts 8 bits of pixel data which indexes a 24-bit
word in the Look-Up Table RAM. The Look-Up Table is configured as a 256 location by 24 bits deep RAM (10 bits each for
Red, Green and Blue). The output of the RAM drives the
DACs with 24-bit data (8 bits each for Red, Green and Blue).
Figure 22. 8-Bit to 24-Bit Pseudo-Color Configuration
This mode allows for the display of 256 simultaneous colors out
of a total palette of millions of addressable colors.
ADV7150
15-Bit “Gamma” True Color
(CR24, CR25, CR26, CR27 = 0, 0, 1, 1 or 1, 0, 1, 1 and
MR11 = 1)
The part is set to 15-bit True-Color operation. The pixel port
accepts 15-bits of color data which is mapped to the 5 LSBs of
each of the red, green and blue palettes of the Look-Up Table
RAM. The Look-Up Table is configured as a 32 location by
30 bits deep RAM (10 bits each for Red, Green and Blue). The
output of the RAM drives the DACs with 30-bit data (10 bits
each for Red, Green and Blue).
15-BIT
COLOR DATA
5
5
5
15-BIT TO 30-BIT
LOOK-UP TABLE
RED
32 x 10
GREEN
32 x 10
BLUE
32 x 10
30-BIT
COLOR DATA
10
10
10
ANALOG VIDEO
10-BIT
RED DAC
10-BIT
GREEN DAC
10-BIT
BLUE DAC
OUTPUTS
RED
OUT
GREEN
OUT
BLUE
OUT
Figure 23. 15-Bit to 30-Bit True-Color Configuration
This mode allows for the display of 15-bit, Gamma-Corrected
True-Color Images.
15-Bit “Standard” True Color
(CR24, CR25, CR26, CR27 = 0, 0, 1, 1 or 1, 0, 1, 1 and
MR11 = 0)
The part is set to 15-bit True-Color operation. The pixel port
accepts 15 bits of color data which is mapped to the 5 LSBs of
each of the red, green and blue palettes of the Look-Up Table
RAM. The Look-Up Table is configured as a 32 location by
24 bits deep RAM (8 bits each for Red, Green and Blue). The
output of the RAM drives the DACs with 24-bit data (8 bits
each for Red, Green and Blue).
8-BIT
RED DAC
8-BIT
GREEN DAC
8-BIT
BLUE DAC
ANALOG VIDEO
OUTPUTS
RED
OUT
GREEN
OUT
BLUE
OUT
15-BIT
COLOR DATA
5
5
5
15-BIT TO 24-BIT
LOOK-UP TABLE
RED
32 x 8
GREEN
32 x 8
BLUE
32 x 8
24-BIT
COLOR DATA
8
8
8
Figure 24. 15-Bit to 24-Bit True-Color Configuration
R7
R4
R3
R2
R1
R0
G4
G3
G2
G1
G0
B4
B3
B2
B1
B0
PIXEL
INPUT
DATA
x
x
x
x
x
x
x
x
x
R6
R5
R4
R3
R2
R1
R0
G7
G6
G5
G4
G3
G2
G1
G0
B7
B6
B5
B4
B3
B2
B1
B0
PIN
ASSIGNMENTS
R4
R3
R2
R1
R0
x
x
x
G4
G3
G2
G1
G0
x
x
x
B4
B3
B2
B1
B0
x
x
x
DATA
LATCHED
TO
PIXEL
PORT
0
0
0
R4
R3
5
R2
R1
R0
0
0
0
G4
G3
5
G2
G1
G0
0
0
0
B4
B3
5
B2
B1
B0
DATA
INTERNALLY
SHIFTED
TO 5 LSBS
256 x 10
RAM
(RED LUT)
LOCATION "31"
LOCATION "0"
256 x 10
RAM
(GREEN LUT)
LOCATION "31"
LOCATION "0"
256 x 10
RAM
(BLUE LUT)
LOCATION "31"
LOCATION "0"
DATA LATCHES
FIRST 32
LOCATIONS
OF RAM
10
TO
RED
DAC
10
TO
GREEN
DAC
10
TO
BLUE
DAC
Figure 25. 15-Bit True-Color Mapping Using R3–R7,
G3–G7 and B3–B7
This mode allows for the display of 15-bit True-Color Images.
PIXEL PORT MAPPING
The pixel data to the ADV7150 is automatically mapped in the
parts pixel port as determined by the pixel data mode programmed (Bits CR24–CR27 of Command Register 2).
Pixel data in the 24-bit True-Color modes is directly mapped to
the 24 color inputs R0–R7, G0–G7 and B0–B7.
There are three modes of operation for 8-bit Pseudo Color.
Each mode maps the input pixel data differently. Data can be
input one of the three color channels, R0–R7 or G0–G7 or
B0–B7.
–16–
REV. A
ADV7150
R7
R6
R5
R4
R3
R2
R1
R0
G7
G6
G5
G4
G3
G2
G1
G0
B7
B6
B5
B4
B3
B2
B1
B0
PIN
ASSIGNMENTS
R4
R3
R2
R1
R0
G4
G3
G2
x
G1
G0
G4
B4
B3
B2
B1
x
x
x
x
x
x
x
x
DATA
LATCHED
TO
PIXEL
PORT
0
0
0
R4
R3
5
R2
R1
R0
0
0
0
G4
G3
5
G2
G1
G0
0
0
0
B4
B3
5
B2
B1
B0
DATA
INTERNALLY
SHIFTED
TO 5 LSBS
256 x 10
RAM
(RED LUT)
LOCATION "31"
LOCATION "0"
256 x 10
RAM
(GREEN LUT)
LOCATION "31"
LOCATION "0"
256 x 10
RAM
(BLUE LUT)
LOCATION "31"
LOCATION "0"
DATA LATCHES
FIRST 32
LOCATIONS
OF RAM
10
TO
RED
DAC
10
TO
GREEN
DAC
10
TO
BLUE
DAC
R4
R3
R2
R1
R0
G4
G3
G2
x
G1
G0
B4
B3
B2
B1
B0
x
x
x
x
x
x
x
x
PIXEL
INPUT
DATA
Figure 26. 15-Bit True-Color Mapping Using R0–R7 and
G0–G6
The part has two modes of operation for 15-bit True Color. In
the first mode, data is input to the device over the red, green
and blue channel (R3–R7, G3–G7 and B3–B7) and is internally
mapped to locations 0 to 31 of the Look-Up Table (LUT) according to Figure 25. In the second mode, data is input to the
device over just two of the color ports, red and green (R0–R7
and G0–G6) and is internally mapped to LUT locations 0 to 31
according to Figure 26. (Note: Data on unused pixel inputs is
ignored.)
MICROPROCESSOR (MPU) PORT
The ADV7150 supports a standard MPU Interface. All the
functions of the part are controlled via this MPU port. Direct
access is gained to the Address Register, Mode Register and all
the Control Registers as well as the Color Palette. The following
sections describe the setup for reading and writing to all of the
devices registers.
MPU Interface
The MPU interface (Figure 27) consists of a bidirectional,
10-bit wide databus and interface control signals
and R/
W. The 10-bit wide databus is user configurable as
The ADV7150 contains a number of onboard registers including the Mode Register (MR17–MR10), Address Register (A7–
A0) and nine Control Registers as well as Red (R9–R0), Green
(G9–G0) and Blue (B9–B0) Color Registers. These registers
control the entire operation of the part. Figure 28 shows the
internal register configuration.
Control lines C1 and C0 determine which register the MPU is
accessing. C1 and C0 also determine whether the Address Register is pointing to the color registers and look-up table RAM or
the control registers. If C1, C0 = 1, 0 the MPU has access to
whatever control register is pointed to by the Address Register
(A7–A0). If C1, C0 = 0, 1 the MPU has access to the Look-Up
Table RAM (Color Palette) through the associated color registers. The
The R/
CE input latches data to or from the part.
W control input determines between read or write ac-
cesses. The Truth Tables III and IV show all modes of access to
the various registers and color palette for both the 8-bit wide
databus configuration and 10-bit wide databus configuration. It
should be noted that after power-up, the devices MPU port is
automatically set to 10-bit wide operation (see Power-On Reset
section).
Color Palette Accesses
Data is written to the color palette by first writing to the address
register of the color palette location to be modified. The MPU
performs three successive write cycles for each of the red, green
and blue registers (10-bit or 8-bit). An internal pointer moves
from red to green to blue after each write is completed. This
pointer is reset to red after a blue write or whenever the address
register is written. During the blue write cycle, the three bytes of
red, green and blue are concatenated into a single 30-bit/24-bit
word and written to the RAM location as specified in the address register (A7–A0). The address register then automatically
increments to point to the next RAM location and a similar red,
green and blue palette write sequence is performed. The address
register resets to 00H following a blue write cycle to color palette RAM location FFH.
CE, C0, C1
REV. A
–17–
ADV7150
ADDRESS
REGISTER
ADDR
(A7–A0)
CE
CONTROL REGISTERS
R/W
MODE
REGISTER
(MR1)
C0 C1
PIXEL MASK
REGISTER
TEST
REGISTERS
ID
REGISTER
MPU PORT
10 (8+2)
D9 – D0
COMMAND
REGISTERS
(CR1–CR3)
REVISION
REGISTER
DATA TO
PALETTES
REGISTER
Figure 27. MPU Port and Register Configuration
RED
30
COLOR REGISTERS
GREEN
REGISTER
BLUE
REGISTER
Data is read from the color palette by first writing to the address
register of the color palette location to be read. The MPU performs three successive read cycles from each of the red, green
and blue locations (10-bit or 8-bit) of the RAM. An internal
pointer moves from red to green to blue after each read is completed. This pointer is reset to red after a blue read or whenever
the address register is written. The address register then automatically increments to point to the next RAM location, and a
similar red, green and blue palette read sequence is performed.
The address register resets to 00H following a blue read cycle of
color palette RAM location FFH.
MODE REGISTER
(MR17–MR10)
C1 = 1
C0 = 0
ADDRESS
REGISTER
(A15–A0)
00H
01H
02H
03H
04H
05H
06H
07H
08H
09H
0AH
0BH
* THIS REGISTER IS READ ONLY.
A READ CYCLE WILL RETURN ZEROS "00".
The MPU can write to or read from all of the ADV7150s registers. C0 and C1 determine whether the Mode Register or Address Register is being accessed. Access to these registers is
direct. The Control Registers are accessed indirectly. The
Address Register must point to the desired Control Register.
Figure 28 along with the 8-bit and 10-bit Interface Truth Tables
illustrate the structure and protocol for device communication
over the MPU port.
C1 = 1
C0 = 1
C1 = 0
C0 = 0
C1 = 0
C0 = 1
RED
REGISTER
(R9–R0)
POINTS TO LOCATION
CORRESPONDING TO
ADDRESS REG (A7–A0)
GREEN
REGISTER
(G9–G0)
LOOK-UP TABLE RAM
(256 x 30)
ADDRESS REG = ADDRESS REG + 1
BLUE
REGISTER
(B9–B0)
Figure 28. Internal Register Configuration and Address Decoding
–18–
REV. A
ADV7150
Table III. Interface Truth Table (10-Bit Databus Mode)
R/WC1C0Databus (D9–D0)OperationResult
011DB7–DB0Write to Mode RegisterDB7–DB0 → MR17–MR10
000DB7–DB0Write to Address RegisterDB7–DB0 → A7–A0
010DB7–DB0Write to Control RegistersDB7–DB0 → Control Register
(Particular Control Register Determined by Address Register)
001DB9–DB0Write to RED RegisterDB9–DB0 → R9–R0
001DB9–DB0Write to GREEN RegisterDB9–DB0 → G9–G0
001DB9–DB0Write to BLUE RegisterDB9–DB0 → B9–B0
Write RGB Data to RAM Location Pointed to by Address Register (A7–A0)
Address Register = Address Register + 1
111DB7–DB0Read Mode RegisterMR17–MR10 → DB7–DB0
100DB7–DB0Read Address RegisterA7–A0 → DB7–DB0
110DB7–DB0Read Control RegistersRegister Data → DB7–DB0
(Particular Control Register Determined by Address Register)
101DB9–DB0Read RED RAM LocationR9–R0 → DB9–DB0
101DB9–DB0Read GREEN RAM LocationG9–G0→ DB9–DB0
101DB9–DB0Read BLUE RAM LocationB9–B0 → DB9–DB0
(RAM Location Pointed to by Address Register(A7–A0))
Address Register = Address Register + 1
DB = Data Bit.
Table IV. Interface Truth Table (8-Bit Databus Mode)*
R/WC1C0Databus (D7–D0)OperationResult
011DB7–DB0Write to Mode RegisterDB7–DB0 → MR17–MR10
000DB7–DB0Write to Address RegisterDB7–DB0 → A7–A0
010DB7–DB0Write to Control RegistersDB7–DB0 → Control Registers
(Particular Control Register Determined by Address Register (A7–A0))
001DB9–DB2Write to RED RegisterDB9–DB2 → R9–R2
001DB1–DB0Write to RED RegisterDB1–DB0 → R1–R0
001DB9–DB2Write to GREEN RegisterDB9–DB2 → G9–G2
001DB1–DB0Write to GREEN RegisterDB1–DB0 → G1–G0
001DB9–DB2Write to BLUE RegisterDB9–DB2 → B9–B2
001DB1–DB0Write to BLUE RegisterDB1–DB0 → B1–B0
Write RGB Data to RAM Location Pointed to by Address Register (A7-A0)
Address Register = Address Register + 1
111DB7–DB0Read Mode RegisterMR17–MR10 → DB7–DB0
100DB7–DB0Read Address RegisterA7–A0 → DB7–DB0
110DB7–DB0Read Control RegistersRegister Data → DB7–DB0
(Particular Control Register Determined by Address Register)
101DB9–DB2Read RED RAM LocationR9–R2 → DB9–DB2
101DB1–DB0Read RED RAM LocationR1–R0 → DB1–DB0
101DB9–DB2Read GREEN RAM LocationG9–G2 → DB9–DB2
101DB1–DB0Read GREEN RAM LocationG1–G0 → DB1–DB0
101DB9–DB2Read BLUE RAM LocationB9–B2 → DB9–DB2
101DB1–DB0Read BLUE RAM LocationB1–B0 → DB1–DB0
(RAM Location Pointed to by Address Register (A7–A0))
Address Register = Address Register + 1
*Writing or reading 10-bit data (DB9–DB0) over an 8-bit databus (D7–D0) requires two write or two read cycles.
:DB9–DB2 is mapped to D7–D0 on the first cycle.
:DB1–DB0 is mapped to D1–D0 on the second cycle.
DB = Data Bit.
REV. A
–19–
ADV7150
Power-On Reset
On power-up of the ADV7150 executes a power-on reset operation. This initializes the pixel port such that the pixel sequence
ABCD starts at A. The Mode Register (MR17–MR10), Command Register 2 (CR27–CR20) and Command Register 3
(CR37–CR30) have all bits set to a Logic “1.” Command Register 1 (CR17–CR10) has all bits set to a Logic “0.”
The output clocking signals are also set during this reset period.
PRGCKOUT = CLOCK/32
LOADOUT = CLOCK/4
The power-on reset is activated when V
goes from 0 V to
AA
5 V. This reset is active for 1 µs. The ADV7150 should not be
accessed during this reset period. The pixel clock should be
applied at power-up.
MR19
RESERVED*
* THESE BITS ARE READ-ONLY RESERVED BITS.
A READ CYCLE WILL RETURN ZEROS "00."
The following section describes each register, including Address
Register, Mode Register and each of the nine Control Registers
in terms of its configuration.
Address Register (A7–A0)
As illustrated in the previous tables, the C0 and C1 control inputs, in conjunction with this address register specify which
control register, or color palette location is accessed by the
MPU port. The address register is 8-bits wide and can be read
from as well as written to. When writing to or reading from the
color palette on a sequential basis, only the start address needs
to be written. After a red, green and blue write sequence, the
address register is automatically incremented.
MR11MR10
MPU DATA BUS WIDTH
MR12
0 8-BIT (D7–D0)
1 10-BIT (D9–D0)
RAM-DAC
RESOLUTION CONTROL
MR11
0 8-BIT
1 10-BIT
RESET CONTROL
MR10
Mode Register 1 (MR1) (MR19–MR10)
MODE REGISTER MR1 (MR19–MR10)
The mode register is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register (MR18 and MR19 are both reserved). It is denoted as
MR17–MR10 for simplification purposes.
The diagram shows the various operations under the control of
the mode register. This register can be read from as well written
to. In read mode, if MR18 and MR19 are read back, they are
both returned as zeros.
Mode Register (MR17–MR10) Bit Description
Reset Control (MR10)
This bit is used to reset the pixel port sampling sequence. This
ensures that the pixel sequence ABCD starts at A. It is reset by
writing a “1” followed by a “0” followed by a “1.” This bit must
be run through this cycle during the initialization sequence.
RAM-DAC Resolution Control (MR11)
When this is programmed with a “1,” the RAM is 30 bits deep
(10 bits each for red, green and blue) and each of the three
DACs is configured for 10-bit resolution. When MR11 is
–20–
programmed with a “0,” the RAM is 24-bits deep (8 bits each
for red, green and blue) and the DACs are configured for 8-bit
resolution. The two LSBs of the 10-bit DACs are pulled down
to zero in 8-bit RAM-DAC mode.
MPU Databus Width (MR12)
This bit determines the width of the MPU port. It is configured
as either a 10-bit wide (D9–D0) or 8-bit wide (D7–D0) bus.
10-bit data can be written to the device when configured in
8-bit wide mode. The 8 MSBs are first written on D7–D0, then
the two LSBs are written over D1–D0. Bits D9–D8 are zeros in
8-bit mode.
Operational Mode Control (MR14–MR13)
When MR14 is “0” and MR13 is “1,” the part operates in
normal mode.
Calibrate LOADIN (MR15)
This bit automatically calibrates the onboard LOADIN/
LOADOUT synchronization circuit. A “0” to “1” transition
initiates calibration. This bit is set to “0” in normal operation.
See “Pipeline Delay and Calibration” section. This bit must be
run through this cycle during the initialization sequence.
REV. A
ADV7150
Palette Select Match Bits Control (MR17–MR16)
These bits allow multiple palette devices to work together.
When bits PS1 and PS0 match MR17 and MR16 respectively,
the device is selected. If these bits do not match, the device is
not selected and the analog video outputs drive 0 mA, see
“Palette Priority Select Inputs” section.
CONTROL REGISTERS
The ADV7150 has 9 control registers. To access each register,
two write operations must be performed. The first write to the
address register specifies which of the 9 registers is to be accessed. The second access determines the value written to that
particular control register.
Pixel Test Register
(Address Reg (A7–A0) = 00H)
This register is used when the device is in test/diagnostic mode.
It is a 24-bit (8 bits each for RED, GREEN and BLUE) wide
read-only register which allows the MPU to read data on the
pixel port, see “Test Diagnostic” section.
DAC Test Register
(Address Reg (A7–A0) = 01H)
This register is used when the device is in test/diagnostic mode.
It is a 30-bit (10 bits each for RED, GREEN and BLUE) wide
read-only register which allows MPU access to the DAC port,
see “Test Diagnostic” section.
SYNC, BLANK and I
Test Register
PLL
(Address Reg (A7–A0) = 02H)
This register is used when the device is in test/diagnostic mode.
It is a 3-bit wide (3 LSBs) read/write register which allows MPU
access to these particular pixel control bits, see “Test Diagnostic” section.
ID Register
(Address Reg (A7–A0) = 03H)
This is an 8-bit wide “Identification” read-only register. For the
ADV7150 it will always return the hexadecimal value 8EH.
Pixel Mask Register
(Address Reg (A7–A0) = 04H)
The contents of the pixel mask register are individually bit-wise
logically AND-ed with the Red, Green and Blue pixel input
stream of data. It is an 8-bit read/write register with D0 corresponding to R0, G0 and B0. For normal operation, this register
is set with FFH.
This register contains a number of control bits as shown in the
diagram. CR1 is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register
(CR18 to CR19 are reserved).
The diagram below shows the various operations under the control of CR1. This register can be read from as well as written to. In
write mode, “0” should be written to CR11 and CR13 to CR17.
In read mode, CR11 and CR13 to CR19 are returned as zeros.
COMMAND REGISTER 1-BIT DESCRIPTION
Calibration Control (CR10)
This bit automatically calibrates the onboard LOADIN/
LOADOUT synchronization circuit. MR15 of Mode Register
MR1 must be set to “0.”
SYNCOUT Control (CR12)
This bit specified whether the video SYNCOUT signal is to be
enabled. On power up a “0” is written to the bit and
“
SYNCOUT” is set three-state.
CR18CR19
RESERVED*
*THESE BITS ARE
READ–ONLY
RESERVED BITS.
A READ CYCLE WILL
RETURN ZEROS "00."
This register contains a number of control bits as shown in the
diagram. CR2 is a 10-bit wide register. However, for programming purposes, it may be considered as an 8-bit wide register
(CR28 and CR29 are both reserved).
The diagram shows the various operations under the control of
CR2. This register can be read from as well written to. In read
mode, CR28 and CR29 are both returned as zeros.
COMMAND REGISTER 2-BIT DESCRIPTION
R7 Trigger Polarity Control (CR20)
This bit is used when the device is in test/diagnostic mode. It
determines whether the pixel data is latched into the test registers in the rising or falling edge of R7. (See “Test Diagnostics”
section.)
I
Trigger Control (CR21)
PLL
This bit specifies whether the I
output is triggered from
PLL
BLANK or SYNC.
SYNC Recognition Control (CR22)
This bit specifies whether the video SYNC input is to be
encoded onto the IOG analog output or ignored.
Pedestal Enable Control (CR23)
This bit specifies whether a 0 IRE or a 7.5 IRE blanking pedestal is to be generated on the video outputs.
True-Color/Pseudo-Color Mode Control (CR27–CR24)
These 4 bits specify the various color modes. These include a
24-bit true-color mode, two 15-bit true-color modes and three
8-bit pseudo color modes.
CR29CR28
RESERVED*
*THESE BITS ARE READONLY RESERVED BITS.
A READ CYCLE WILL
RETURN ZEROS "00."
TRUE COLOR/PSEUDO-COLOR MODE CONTROL
CR27CR26CR24CR25
0
0
1
1100
1101
00
1000
000
CR26CR25CR24CR27
PEDESTAL ENABLE
CONTROL
CR23
0 0 IRE
1 7.5 IRE
MODE
8-BIT PSEUDO COLOR ON R7–R0
8-BIT PSEUDO COLOR ON G7–G0
8-BIT PSEUDO COLOR ON B7–B0
15-BIT TRUE COLOR ON
R7–R3, G7–G3, B7–B3
15-BIT TRUE COLOR ON
R7–R0, G6–G0
This register contains a number of control bits as shown in the
diagram. CR3 is a 10-bit wide register. However for programming purposes, it may be considered as an 8-bit wide register
(CR38 and CR39 are both reserved).
The diagram shows the various operations under the control of
CR3. This register can be read from as well written to. In read
mode, CR38 and CR39 are both returned as zeros.
COMMAND REGISTER 3-BIT DESCRIPTION
PRGCKOUT Frequency Control (CR31–CR30)
These bits specify the output frequency of the PRGCKOUT
output. PRGCKOUT is a divided down version of the pixel
CLOCK.
CR39CR38CR37CR36CR35CR34CR32CR31CR30CR33
RESERVED*
*THESE BITS ARE READONLY RESERVED BITS.
A READ CYCLE WILL
RETURN ZEROS "00."
EXTRA BLANK PIPELINE DELAY CONTROL
(ADDS TO PIXEL PIPELINE DELAY; tPD)
CR35 CR34 CR33 CR32
0
0
0
0
0
0
·
·
·
·
1
1
BLANK Pipeline Delay Control (CR35–CR32)
These bits specify the additional pipeline delay that can be
added to the
pipeline delay (t
BLANK function, relative to the overall device
). As the BLANK control normally enters the
PD
video DAC from a shorter pipeline than the video pixel data,
this control is useful in deskewing the pipeline differential.
Pixel Multiplex Control (CR37–CR36)
These bits specify the device’s multiplex mode. It, therefore,
also determines the frequency of the LOADOUT signal.
LOADOUT is a divided down version of the pixel CLOCK.
Revision Register
(Address Reg (A7–A0) = 0BH)
This register is a read only register containing the revision of
silicon.
0
0
1
·
·
1
t
0
1
0
·
·
1
PD
t
+ 1 x LOADOUT
PD
t
+ 2 x LOADOUT
PD
t
+ 15 x LOADOUT
PD
·
·
PIXEL MULTIPLEX CONTROL
CR37 CR36
1:1 MUXING: LOADOUT = CLOCK ÷ 1
0
0
0
1
1
2:1 MUXING LOADOUT = CLOCK ÷ 2
1
RESERVED
0
4:1 MUXING :LOADOUT = CLOCK÷ 4
1
Command Register 3 (CR3) (CR39–CR30)
PRGCKOUT FREQUENCY
CONTROL
CR31 CR30
0
1
0
1
CLOCK ÷ 4
CLOCK ÷ 8
CLOCK ÷ 16
CLOCK ÷ 32
0
0
1
1
REV. A
–23–
ADV7150
DIGITAL-TO-ANALOG CONVERTERS
(DACS) AND VIDEO OUTPUTS
The ADV7150 contains three high speed video DACs. The
DAC outputs are represented as the three primary analog color
signals IOR (red video), IOG (green video) and IOB (blue video).
Other analog signals on the part include I
complementary video outputs
IOR, IOG, IOB. These complementary outputs can be used to drive differentially terminated
video loads, they will have equal but opposite output levels to
IOR, IOG and IOB when loaded with a resistive load similar to
IOR, IOG and IOB.
DACs and Analog Outputs
The part contains three matched 10-bit digital-to-analog converters. The DACs are designed using an advanced, high speed,
segmented architecture. The bit currents corresponding to each
digital input are routed to either IOR, IOG, IOB (bit = “l”) or
IOR, IOG, IOB (bit = “0”). (Normally IOR, IOG, IOB = GND.)
The analog video outputs are high impedance current sources.
Each of the these three RGB current outputs are specified to
directly drive a 37.5 Ω load (doubly terminated 75 Ω).
An external 1.23 V voltage reference is required to drive the
analog outputs of the ADV7150. The reference voltage is connected to the V
A resistor R
and ground. For specified performance, R
input.
REF
is connected between the R
SET
input of the part
SET
has a value of
SET
280 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 30 illustrates the resulting
video waveform, and the Video Output Truth Table shows the
corresponding control input stimuli.
IOR, IOB IOG
mA mAVV
1.44 0.054 9.05 0.340
00
26.67 1.0000.71419.05
0.286
7.62
00
92.5 IRE
7.5 IRE
40 IRE
GRAY SCALE
Figure 30. Composite Video Waveform (
on IOG; Pedestal = 7.5 IRE; R
= 280 Ω)
SET
SYNC
WHITE
LEVEL
BLACK
LEVEL
BLANK
LEVEL
SYNC
LEVEL
Decoded
Variations on RS-343A
Various other video output configurations can be implemented
by the ADV7150, including RS-170. Values of R
for particu-
SET
lar output video formats/levels are calculated by using the equations for R
table shows calculated values of R
given in the “Pin Configuration” section. The
SET
for some of the most com-
SET
mon variants on the RS-343A standard. The associated waveforms are shown in the diagrams.
Table V. Video Output Truth Table
IOGIOR, IOBDAC
Description(mA)(mA)SYNCBLANKInput Data
WHITE LEVEL26.6719.05113FFH
VIDEOVideo + 9.05Video + 1.4411Data
VIDEO to
BLANKVideo + 1.44Video + 1.4401Data
BLACK LEVEL9.051.4411000H
BLACK to
Figure 32. Composite Video Waveform
(Pedestal = 7.5 IRE; R
IOR, IOB, IOG
mAV
19.05 0.714
= 280 Ω)
SET
WHITE LEVEL
WHITE
LEVEL
BLACK/
BLANK
LEVEL
SYNC
LEVEL
R
(V)Video Signal
SET
265
280No
SYNC decoded on IOG; Pedestal = 0 IRE
SYNC decoded; Pedestal = 7.5 IRE
259No SYNC decoded; Pedestal = 0 IRE
I
Synchronization Output Control
PLL
This output synchronization signal is used in applications where
it is necessary to synchronize multiple palette devices (ADV7150
+ ADV7151) to subpixel resolution. Each devices I
PLL
output
signal is in phase with its analog RGB output signal. If multiple
devices have differing output delays, the time difference can be
derived from the I
to phase shift the
signals. This time difference is then used
PLL
CLOCK inputs on one or other of the devices
inputs.
The I
of
signal is internally triggered by either the falling edge
PLL
SYNC or BLANK as determined by CR21 of Command
Register 2.
100 IRE
GRAY SCALE
00
Figure 33. Composite Video Waveform
(Pedestal = 0 IRE; R
= 259 Ω)
SET
BLACK/ BLANK
LEVEL
REV. A
–25–
ADV7150
APPENDIX 1
BOARD DESIGN AND LAYOUT CONSIDERATIONS
POWER SUPPLY DECOUPLING (0.1µF AND 0.01µF CAPACITOR FOR EACH VAA GROUP)
+5V (VAA)
0.1µF
COMP
V
AA
ADV7150
GND
0.1µF0.01µF
ANALOG POWER PLANE
IOR
IOG
IOB
IOR
IOG
IOB
I
REF
PLL
SET
(1% METAL)
R
SET
280Ω
75Ω75Ω
V
R
)
+5V (V
AA
1kΩ
AD589
(1.2V REF)
75Ω
COMPLIMENTARY
OUTPUTS
0.1µF
0.1µF
0.01µF
+5V (VAA)
33µF
CO-AXIAL CABLE
(75Ω)
BNC
CONNECTORS
NOTES:
1. ALL RESISTORS ARE 1% METAL FILM
2. 0.1µF AND 0.01µF CAPACITORS ARE CERAMIC
3. ADDITIONAL DIGITALCIRCUITRY OMITTED FOR CLARITY
0.01µF
0.1µF
L1
(FERRITE BEAD)
75Ω
75Ω
75Ω
MONITOR
(CRT)
0.1µF
+5V (VCC)
0.01µF
0.1µF
Recommended Analog Circuit Layout
The ADV7150 is a highly integrated circuit containing both
precision analog and high speed digital circuitry. It has been
designed to minimize interference effects on the integrity of the
analog circuitry by the high speed digital circuitry. It is imperative that these same design and layout techniques be applied to
the system level design such that high speed, accurate performance is achieved. The “Recommended Analog Circuit Layout”
shows the analog interface between the device and monitor.
The layout should be optimized for lowest noise on the ADV7150
power and ground lines by shielding the digital inputs and providing good decoupling. The lead length between groups of V
AA
and GND pins should by minimized so as to minimize inductive
ringing.
Ground Planes
The ground plane should encompass all ADV7150 ground pins,
voltage reference circuitry, power supply bypass circuitry for the
ADV7150, the analog output traces, and all the digital signal
traces leading up to the ADV7150. The ground plane is the
graphics board’s common ground plane.
Power Planes
The ADV7150 and any associated analog circuitry should have
its own power plane, referred to as the analog power plane (V
AA
).
This power plane should be connected to the regular PCB
–26–
power plane (VCC) at a single point through a ferrite bead. This
bead should be located within three inches of the ADV7150.
The PCB power plane should provide power to all digital logic
on the PC board, and the analog power plane should provide
power to all ADV7150 power pins and voltage reference circuitry.
Plane-to-plane noise coupling can be reduced by ensuring that
portions of the regular PCB power and ground planes do not
overlay portions of the analog power plane, unless they can be
arranged such that the plane-to-plane noise is common mode.
Supply Decoupling
For optimum performance, bypass capacitors should be installed
using the shortest leads possible, consistent with reliable operation, to reduce the lead inductance. Best performance is obtained
with 0.1 µF ceramic capacitor decoupling. Each group of V
AA
pins on the ADV7150 must have at least one 0.1 µF decoupling
capacitor to GND. These capacitors should be placed as close
as possible to the device.
It is important to note that while the ADV7150 contains circuitry to reject power supply noise, this rejection decreases with
frequency. If a high frequency switching power supply is used,
the designer should pay close attention to reducing power supply noise and consider using a three terminal voltage regulator
for supplying power to the analog power plane.
REV. A
ADV7150
Digital Signal Interconnect
The digital inputs to the ADV7150 should be isolated as much
as possible from the analog outputs and other analog circuitry.
Also, these input signals should not overlay the analog power
plane.
Due to the high clock rates involved, long clock lines to the
ADV7150 should be avoided to reduce noise pickup.
Any active termination resistors for the digital inputs should be
connected to the regular PCB power plane (V
), and not the
CC
analog power plane.
Analog Signal Interconnect
The ADV7150 should be located as close as possible to the output connectors to minimize noise pick-up and reflections due to
impedance mismatch.
The video output signals should overlay the ground plane, and
not the analog power plane, to maximize the high frequency
power supply rejection.
APPENDIX 2
TYPICAL FRAME BUFFER INTERFACE
CLOCK
GENERATOR
CLOCK
CLOCK
Digital Inputs, especially Pixel Data Inputs and clocking signals
(CLOCK, LOADOUT, LOADIN, etc.) should never overlay
any of the analog signal circuitry and should be kept as far away
as possible.
For best performance, the analog outputs (IOR, IOG, IOB)
should each have a 75 Ω load resistor connected to GND.
These resistors should be placed as close as possible to the
ADV7150 so as to minimize reflections. Normally, the differential analog outputs (
IOR, IOG, IOB) are connected directly to
GND. In some applications, improvements in performance are
achieved by terminating these differential outputs with a resistive load similar in value to the video load. For a doubly terminated 75 Ω load, this means that
IOR, IOG, IOB are each
terminated with 37.5 Ω resistors.
ECL
TO
TTL
CLOCK
GRAPHICS
PROCESSOR/
CONTROLLER
FRAME
BUFFER/
VIDEO
MEMORY
BLANK
SYNC
VRAM
(BANK A)
VRAM
(BANK B)
VRAM
(BANK C)
VRAM
(BANK D)
33MHz
33MHz
33MHz
33MHz
PRGCKOUT
LOADOUT
SCKOUT
BLANK
SCKIN
LOADIN
DIVIDE BY N
(÷ N)
DIVIDE BY M
LATCH
ENABLESYNC
(÷ M)
ADV7150
24
24
24
24
24
24
MULTIPLEXER
24
24
24TO
PALETTE/RAM
& DAC
REV. A
–27–
ADV7150
APPENDIX 3
10-BIT DACS AND GAMMA CORRECTION
10-Bit DACs
10-Bit RAM-DAC resolution allows for nonlinear video correction, in particular Gamma Correction. The ADV7150 allows for
an increase in color resolution from 24-bit to 30-bit effective
color without the necessity of a 30-bit deep frame buffer. In
true-color mode, for example, the part effectively operates as a
24-bit to 30-bit color look-up table.
Up to now we have assumed that there exists a linear relationship between the actual RGB values input to a monitor and the
intensity produced on the screen. This, however, is not the case.
Half scale digital input (1000 0000) might correspond to only 20%
output intensity on the CRT (Cathode Ray Tube). The intensity
(I
) produced on a CRT by an input value IIN is given by:
CRT
I
CRT
= (IIN)
χ
where χ ranges from 2.0 to 2.8.
If the individual values of χ for red, green and blue are known,
then so called “Gamma Correction” can be applied to each of
the three video input signals (I
);
IN
therefore:
I
IN(corrected)
= k(IIN)
1/χ
(k = 1, normally)
Traditionally, there has been a tradeoff between implementing a
nonlinear graphics function, such as gamma correction, and
color dynamic range. The ADV7150 overcomes this by increasing the individual color resolution of each of the red, green and
blue primary colors from 8 bits per color channel to 10 bits per
channel (24 bits to 30 bits).
The table highlights the loss of resolution when 8-bit data is
gamma-corrected to a value of 2.7 and quantized in a traditional 8-bit system. Note that there is no change in the 8-bit
quantized data for linear changes in the input data over much of
the transfer function. On the other hand, when quantized to 10
bits via the 10-bit RAMs and 10-bit DACs of the ADV7150, all
changes on the input 8-bit data are reflected in corresponding
changes in the 10-bit data.
The graph shows a typical gamma curve corresponding to a
gamma value of 2.7. This is programmed to the red, green and
blue RAMs of the color lookup table instead of the more traditional linear function. Different curves corresponding to any
particular gamma value can be independently programmed to
each of the red, green and blue RAMs.
Other applications of the 10-bit RAM-DAC include closed-loop
monitor color calibration.
The palette priority selection inputs allow up to four separate
palette devices to be used in a single system to drive a single
monitor with subpixel resolution. The IOR, IOG and IOB analog video output signals of each device are connected together,
as shown. Signal inputs (PS0, PS1) determine on a pixel by pixel
basis which palette device drives the monitor. This allows for
implementation of multiple windows applications with each
device acting as an independent palette. During initialization,
each device is assigned two match bits, MR16 (PS0) and MR17
(PS1) in Mode Register MR1. PS0 and PS1 inputs will select
one of the preprogrammed devices at any instant when PS0, PS1
matches MR16, MR17, respectively. PS0 and PS1 are multiplexed similar to the pixel data, thus allowing for subpixel resolution. The diagrams show an example of one ADV7150 operating in
ADV7150
R0–R7
G0–G7
B0–B7
PS0, PS1
256 x 30
RAM
PALETTE SELECT BITS
MR16 MR17
0 0
ANALOG
O/P
conjunction with three ADV7151’s (Pseudo-Color RAM-DACs).
Each displayed window on the monitor is driven by one of the
four devices, as determined on a pixel basis by PS0, PS1. Each
device’s analog output signals are connected together as shown.
Note: Only one palette device is selected at any particular
instant. The analog output levels of the unselected devices will
be 0 mA.
Other applications for the palette priority function using a minimum of two devices (one ADV7150 and one ADV7151) include:
Cursor Overlay on 24-Bit Graphics
Active Live Video Overlay (from Frame Grabber)
Text/Character Generation and Overlay
Multiple Devices Termination for a Single Monitor
REV. A
P0–P7
ADV7151 (1)
256 x 30
PALETTE
PALETTE SELECT BITS
MR16 MR17
0 1
ADV7151 (2)
256 x 30
PALETTE
PALETTE SELECT BITS
MR16 MR17
1 0
ADV7151 (3)
256 x 30
PALETTE
PALETTE SELECT BITS
MR16 MR17
1 1
Multiple Devices Driving a Multiwindow Application
RGB
ANALOG
VIDEO
RGB
ANALOG
VIDEO
RGB
ANALOG
VIDEO
VIDEO TO MONITOR
–29–
MONITOR
TRUE-COLOR BACKGROUND
WINDOW 1
(Pseudo-Color)
PS0=0: PS1=1
WINDOW 2
(Pseudo-Color)
PS0=1: PS1=0
WINDOW 3
(Pseudo-Color)
PS0=1: PS1=1
ADV7150
APPENDIX 5
INITIALIZATION AND PROGRAMMING
ADV7150 Initialization
After power has been supplied, the ADV7150 must be initialized. The Mode Register and Control Registers must be set.
The values written to the various registers will be determined by
the desired operating mode of the part, i.e., True Color/Pseudo
Color, 2:1 Muxing/2:1 Muxing, etc.
Example 1
Color Mode24-Bit True Color
Multiplexing2:1
Databus8-Bit
RAM-DAC Resolution8-Bit
SYNCEnabled on IOG
Pedestal7.5 IRE
Register InitializationC1C0R/
Write09H to Mode Register (MR1)110Resets to Normal Operation, 8-Bit Bus/RAM-DAC
Write08H to Mode Register (MR1)110*(Initializes Pipelining
Write09H to Mode Register (MR1)110*( “
Write29H to Mode Register (MR1)110*(Calibrates LOADOUT/LOADIN Timing
Write09H to Mode Register (MR1)110*( “
Write04H to Address Register (A7–A0)000Address Reg Points to Pixel Mask Register
WriteFFH to Pixel Mask Register100Sets the Pixel Mask to All “1s”
Write05H to Address Register (A7–A0)000Address Reg Points to Command Register 1 (CR1)
Write00H to Command Reg 1 (CR1)100
Write06H to Address Register (A7–A0)000Address Reg Points to Command Register 2 (CR2)
WriteECH to Command Reg 2 (CR2)100Sets 24-Bit Color, 7.5 IRE,
Write07H to Address Register (A7–A0)000Address Reg Points to Command Register 3 (CR3)
WriteC0H to Command Reg 3 (CR3)100Sets 2:1 Multiplexing, PRGCKOUT = CLOCK/4
Color Palette RAM InitializationC1C0R/WComment
Write00H to Address Register (A7–A0)000Points to Color Palette RAM
Write00H (Red Data) to RAM Location (00H)010(Initializes Palette RAM
Write00H (Green Data) to RAM Location (00H)010(to a Linear Ramp**
Write00H (Blue Data) to RAM Location (00H)010(
Write01H (Red Data) to RAM Location (01H)010(
Write01H (Green Data) to RAM Location (01H)010(
Write01H (Blue Data) to RAM Location (01H)010(
•••• ••••(
•••• ••••(
WriteFFH (Red Data) to RAM Location (FFH)010(
WriteFFH (Green Data) to RAM Location (FFH)010(
WriteFFH (Blue Data) to RAM Location (FFH)010(RAM Initialization Complete
The following section gives examples of initialization of the
ADV7150 operating in various modes.
WComment
SYNC on Green (IOG)
**These four command lines reset the ADV7150. The pipelines for each of the Red, Creen and Blue pixel inputs are synchronously reset to the Multiplexer’s
“A” input. Mode Register bit MR10 is written by a “1” followed by “0” followed by “1.” LOADIN/LOADOUT timing is internally synchronized by writing a “0”
followed by a “1” followed by a “0” to Mode Register MR15.
**This sequence of instructions would, of course, normally be coded using some form of loop instruction.
–30–
REV. A
ADV7150
Example 2
Color Mode24-Bit Gamma Corrected True Color (30 Bits)
Multiplexing2:1
Databus10 Bit
RAM-DAC Resolution 10 Bit
SYNCIgnored
Pedestal0 IRE
CalibrationEvery Vertical Sync
Register InitializationC1C0R/WComment
Write0FH to Mode Register (MR1)110Resets to Normal Operation, 10-Bit Bus/RAM-DAC
Write0EH to Mode Register (MR1)110*(Initializes Pipelining
Write0FH to Mode Register (MR1)110*( “
Write2FH to Mode Register (MR1)110*(Calibrates LOADOUT/LOADIN Timing
Write0FH to Mode Register (MR1)110*( “
Write04H to Address Register (A7–A0)000Address Reg Points to Pixel Mask Register
WriteFFH to Pixel Mask Register100Sets the Pixel Mask to All “1s”
Write05H to Address Register (A7–A0)000Address Reg Points to Command Register 1 (CR1)
Write01H to Command Reg 1 (CR1)000Calibrates Every Vertical Sync
Write06H to Address Register (A7–A0)000Address Reg Points to Command Register 2 (CR2)
WriteE0H to Command Reg 2 (CR2)100Sets 24-Bit Color, 0 IRE, No
Write07H to Address Register (A7–A0)000Address Reg Points to Command Register 3 (CR3)
Write41H to Command Reg 3 (CR3)100Sets 2:1 Multiplexing, PRGCKOUT =
Color Palette RAM InitializationC1C0R/
WComment
Write00H to Address Register (A7–A0)000Points to Color Palette RAM
Write000H (Red Data) to RAM Location (00H)010(Initializes Palette RAM
Write000H (Green Data) to RAM Location (00H)010( to a “Gamma” Ramp**
Write000H (Blue Data) to RAM Location (00H)010(
WritexxxH (Red Data) to RAM Location (01H)010(
WritexxxH (Green Data) to RAM Location (01H)010(
WritexxxH (Blue Data) to RAM Location (01H)010(
•••• ••••(
•••• ••••(
Write3FFH (Red Data) to RAM Location (FFH)010(
Write3FFH (Green Data) to RAM Location (FFH)010(
Write3FFH (Blue Data) to RAM Location (FFH)010(RAM Initialization Complete
**These four command lines reset the ADV7150 The pipelines for each of the Red, Green and Blue pixel inputs are synchronously reset to the Multiplexer’s “A” in-
put. Mode Register bit MR10 is written by a “1” followed by “0” followed by “1.” LOADIN/LOADOUT timing is internally synchronized by writing a “0” followed
by a “1” followed by a “0” to Mode Register MR15.
**Data for a gamma curve characteristic is obtainable in Appendix 3.
REGISTER DIAGNOSTIC TESTING
The previous examples show the register initialization sequence
for the ADV7150. These show control data going to the registers and palette RAM. As well as this writing function, it may
also be necessary, due to system diagnostic requirements, to
confirm that correct data has been transferred to each register
and palette RAM location. There are two ways to incorporate
register value/RAM value checking:
1. READ after each WRITE: After data is written to a particular
register, it can be read back immediately. The following table
shows an example with Command Registers CR2 and CR3.
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.
C1C0R/WD0–D7 Comment
00006HSelect Command Register 2 (CR2)
100E0HSets 24-Bit True-Color
00007HSelect Command Register 3 (CR3)
10040HSet 2:1 Mux Mode
00006HSelect CR2
101E0HCR2 Value Read-Back
00007HSelect CR3
10140HCR3 Value Read-Back
10140HCR3 Value Read-Back
It is clear that this latter case requires more command lines
than the previous READ after each WRITE case.
SYNC
CLOCK/8
REV. A
–31–
ADV7150
APPENDIX 6
TEST DIAGNOSTICS
SYNC
BLANK
PIXEL
DATA
INPUT
MUX
GRAPHICS PIPELINEGRAPHICS PIPELINE
TRIGGER
DECODE
PIXEL TEST
REGISTER
CER/WC0C1 D0–D9
COLOR
PALETTE
RAM
COLOR
REGISTERS
Test/Diagnostic Block Diagram
The ADV7150 contains onboard circuitry which enables both
device and system level test diagnostics. The test circuitry can
be used to test the frame buffer memory as well as the functionality of the ADV7150. A number of test registers are integrated
into the part which effectively allow for monitoring of the graphics pipeline. Pixel data is read from the graphics pipeline independent of the pixel CLOCK. The pixel data itself contains the
triggering information that latches data into the test registers.
This allows for system diagnostics in a continuously clocked
graphics system. The test register data is then read by the microprocessor over the MPU.
Access to the test registers is as described in the “Microprocessor (MPU) Port” section. This section also gives the address
decode locations for the various test registers.
Test Trigger (R7)
The test trigger is decoded from the pixel data stream. Bit R7 of
the RED channel is assigned the task of latching pixel data into
the test registers. A “0” to “1” or a “1” to “0” (as determined
by bit CR20 of Command Register 2) transition on R7, fills the
test register with the corresponding pixel data. This effectively
means that a sequence of data travels along the graphics pipeline, with the test registers taking a sample only when there is a
transition on Bit R7. The following example shows a sequence
with the ADV7150 preset to sample the graphics pipeline on a
low to high transition of R7.
REDGREENBLUE
Pixel 0:000000000000000000000000
Pixel 1:0........................
Pixel 2:1........................
Pixel 3:0........................
.... ...
.... ...
Pixel n- l:0........................
Pixel n:1........................
Pixel n:0........................
In the above sequence of pixels, there is a rising edge on R7 on
Pixel 2. The Red, Green and Blue data for Pixel 2, therefore,
gets latched into the Pixel Test Register. Pixel 2 continues down
–32–
DACs
TRIGGER
MPU PORT
DECODE
DAC TEST
REGISTERS
SYNC
I
PLL
REGISTER
BLANK
TEST
the graphics pipeline and after a number of clocks get latched
into the DAC Test Register. This data can then be read from
the Pixel Test Register and the DAC Test Registers over the
MPU Port. This data will remain in the Pixel Test Registers and
the DAC Test Registers until the next rising edge of R7 causes
new data to be latched in.
In the above example, the next rising edge of R7 occurs on the
Pixel n input. Therefore the data in the Pixel Test Registers and
DAC Test Registers must be read over the MPU before the
Pixel n data is applied, otherwise they will be overwritten by the
Pixel n data and the Pixel 2 data will be lost.
Pixel Test Register
The read-only Pixel Test Register is 24 bits wide, 8 bits each for
red green and blue. It is situated directly after the Pixel Mask
Register. After data is latched into this register by a transition on
R7, it is read in three cycles over the MPU Port as described in
the “Microprocessor (MPU) Port” section.
DAC Test Register
The DAC Test Register is latched with data some CLOCKs
after the Pixel Test Register. The DAC Test Register is a 30-bit
wide read-only register, corresponding to 10 bits each for red,
green and blue data. It is located the Color Palette RAM. If the
RAM-DAC is in 8-bit after resolution mode, the upper two bits
of the red, green and blue data will be zero. After data is latched
into the DAC Test Register by a transition on R7, it is read
in three or six cycles over the MPU Port as described in the
“Microprocessor (MPU) Port” section.
SYNC, BLANK and I
Test Register
PLL
This is an 8-bit wide register but with only three effective bits.
The three lower bits correspond to
SYNC, BLANK and I
PLL
respectively. The upper bits should be masked in software. This
register is at the same position in the graphics pipeline as the
DAC Test Register. When pixel data is latched into the DAC
Test Register, the corresponding status of
I
is latched into this register. It is read over the MPU Port as
PLL
SYNC, BLANK and
described in the “Microprocessor (MPU) Port” section.
(Note: If
BLANK is low, the corresponding pixel data to the
DAC Test Register will be all “0s.”)
REV. A
APPENDIX 7
THERMAL AND ENVIRONMENTAL CONSIDERATIONS
ADV7150
The ADV7150 is a very highly integrated monolithic silicon
device. This high level of integration, in such a small package,
inevitably leads to consideration of thermal and environmental
conditions in which the ADV7150 must operate. Reliability of
the device is significantly enhanced by keeping it as cool as possible. In order to avoid destructive damage to the device, the
absolute maximum junction temperature of 150°C must never
be exceeded. Certain applications, depending on pixel data
rates, may require forced air cooling, or external heatsinks. The
following data is intended as a guide in evaluating the operating
conditions of a particular application so that optimum device
and system performance is achieved.
It should be noted that information on package characteristics pub-
lished herein may not be the most up to date at the time of reading
this. Advances in package compounds and manufacture will inevitably lead to improvements in the thermal data. Please contact your
local sales office for the most up-to-date information.
Power Dissipation
The diagram shows graphs of power dissipation in watts vs.
pixel clock frequency for the ADV7150.
1.50
V
= 5V
AA
= 1.2V
V
1.25
1.00
0.75
POWER DISSIPATION – Watts
0.50
NOTE: THE "WORST CASE ON-SCREEN PATTERN" CORRESPONDS TO FULL-SCALE
TRANSITION ON EACH PIXEL VALUE FOR EVERY CLOCK EDGE (00H, FFH, 00H, ... ).
THE "TYPICAL ON-SCREEN PATTERN" CORRESPONDS TO LINEAR CHANGES IN THE
PIXEL INPUT (I. E., A BLACK TO WHITE RAMP). IN GENERAL, COLOR IMAGES TEND
TO APPROXIMATE THIS CHARACTERISTIC.
REF
= +25°C
T
A
6022080180200160140120100
PIXEL CLOCK FREQUENCY – MHz
Table A. Thermal Characteristics vs. Airflow
Air Velocity050100200
(Linear feet/min)(Still Air)
θ
(°C/W)
JA
No Heatsink25.5232119
EG&G D10100-28 Heatsink 23201816
Thermalloy 2290 Heatsink19171512
Thermal Model
The junction temperature of the device in a specific application
is given by:
T
= TA + PD (
J
θ
+
θ
JC
)(1)
CA
or
T
= TA + PD (
J
θ
) (2)
JA
where:
= Junction Temperature of Silicon (°C)
T
J
T
= Ambient Temperature (°C)
A
P
= Power Dissipation (W)
D
θ
= Junction to Case Thermal Resistance (°C/W)
JC
θ
= Case to Ambient Thermal Resistance (°C/W)
CA
θ
= Junction to Ambient Thermal Resistance (°C/W)
JA
Package Enhancements
The standard QFP package has been enhanced to a PowerQuad2
package. This supports an improved thermal performance compared to standard QFP. In this case, the die is attached to
heatslug so that the power that is dissipated can be conducted to
the external surface of the package. This provides a highly efficient path for the transfer of heat to the package surface. The
package configuration also provides an efficient thermal path
from the ADV7150 to the Printed Circuit Board via the leads.
Heatsinks
The maximum silicon junction temperature should be limited to
100°C. Temperatures greater than this will reduce long term
device reliability. To ensure that the silicon junction temperature stays within prescribed limits, the addition of an external
heatsink may be necessary. Heatsinks, will reduce θ
as shown
JA
in the “Thermal Characteristics vs. Airflow” table.
Typical Power Dissipation vs. Pixel Rate
Package Characteristics
The table of thermal characteristics shows typical information
for the ADV7150 (160-Lead Plastic Power QFP) using various
values of Airflow.
Junction to Case (θ
) Thermal Resistance for this particular
JC
part is:
θ
(160-Lead Plastic Power QFP) =1.0°C/W
JC
(Note: θ
REV. A
is independent of airflow.)
JC
–33–
ADV7150
APPENDIX 8
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
S-160
160-Lead Plastic Power Quad Flatpack
0.037 (0.95)
0.026 (0.65)
SEATING
PLANE
0.004 (0.10)
MAX
0.070 (1.77)
0.062 (1.57)
0.160 (4.07)
4°±4°
MAX
0.145 (3.67)
0.125 (3.17)
MAX
6°±4°
10°
0.070 (1.77)
0.062 (1.57)
121
160
120
1
1.239 (31.45)
1.219 (30.95)
1.107 (28.10)
1.100 (27.90)
PIN 1
0.026 (0.65) MIN
TOP VIEW
(PINS DOWN)
SQ
SQ
0.014 (0.35)
0.011 (0.27)
81
80
41
40
–34–
REV. A
–35–
C1695–10–8/94
–36–
PRINTED IN U.S.A.
REV. A
Loading...
+ hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.