High Performance 12-Bit, 6-Channel Output,
Decimating LCD DecDriver®
AD8382
PRODUCT FEATURES
High accuracy, high resolution voltage outputs
12-bit input resolution
Laser trimmed outputs
Fast settling, high voltage drive
33 ns settling time to 0.25% into 200 pF load
Slew rate 390 V/µs
Outputs to within 1.3 V of supply
High update rates
Fast, 120 Ms/s data update rate
Voltage controlled video reference (brightness) and
full-scale (contrast) output levels
Flexible logic
STSQ/XFR allow parallel AD8382 operation
INV bit reverses polarity of video signal
Output overload protection
Low static power dissipation: 743 mW
Includes STBY function
3.3 V logic, 9 V to 18 V analog supplies
Available in 48-lead 7 mm × 7 mm LFCSP
APPLICATIONS
FUNCTIONAL BLOCK DIAGRAM
DB(0:11)
12
12
AD8382
STBY
BYP
BIAS
R/L
E/O
CLK
STSQ
XFR
SEQUENCE
CONTROL
12 2-STAGE 12
LATCH
DAC
VID0
12 2-STAGE 12
LATCH
DAC
VID1
12 2-STAGE 12
LATCH
DAC
VID2
12 2-STAGE 12
LATCH
DAC
VID3
12 2-STAGE 12
LATCH
DAC
VID4
12 2-STAGE 12
LATCH
DAC
VID5
SCALING
CONTROL
VREFHI
VREFLO
INV
V1
V2
Figure 1. Functional Block Diagram
LCD analog column driver
PRODUCT DESCRIPTION
The AD8382 DecDriver provides a fast, 12-bit latched decimating digital input that drives six high voltage outputs.12-bit input
words are sequentially loaded into six separate, high speed,
bipolar DACs. A flexible digital input format allows several
AD8382s to be used in parallel for higher resolution displays.
STSQ synchronizes sequential input loading, XFR controls
synchronous output updating, and R/L controls the direction of
loading as either left-to-right or right-to-left. Six channels of
high voltage output drivers drive to within 1.3 V of the rail. The
output signal can be adjusted for dc reference, signal inversion,
and contrast for maximum flexibility.
The AD8382 is fabricated on Analog Devices’ XFHV, fast
bipolar 26 V process, providing fast input logic bipolar DACs
with trimmed accuracy and fast settling, high voltage, precision
drive amplifiers on the same chip. The AD8382 dissipates
743 mW nominal static power. The STBY pin reduces power to
a minimum, with fast recovery.
Rev. 0
Information furnished by Analog Devices is believed to be accurate and reliable.
However, no responsibility is assumed by Analog Devices for its use, nor for any
infringements of patents or other rights of third parties that may result from its use.
Specifications subject to change without notice. No license is granted by implication
or otherwise under any patent or patent rights of Analog Devices. Trademarks and
registered trademarks are the property of their respective companies.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700
www.analog.com
Fax: 781.326.8703
© 2003 Analog Devices, Inc. All rights reserved.
AD8382
TABLE OF CONTENTS
Specifications..................................................................................... 3
Applications..................................................................................... 15
Absolute Maximum Ratings............................................................ 5
VBIAS Generation—V1, V2 Input Pin Functionality ........... 17
Timing Characteristics..................................................................... 6
Power Supply Sequencing ......................................................... 18
Pin Configuration and Functional Descriptions.......................... 7
PCB Design for Optimized Thermal Performance ............... 18
Typical Performance Characteristics ............................................. 8
Layout Considerations............................................................... 20
Functional Description .................................................................. 13
Outline Dimensions ....................................................................... 21
Transfer Function ....................................................................... 13
Ordering Guide .......................................................................... 21
Accuracy ...................................................................................... 14
REVISION HISTORY
Revision 0: Initial Version
Rev. 0 | Page 2 of 24
AD8382
SPECIFICATIONS
Table 1. @ 25°C, AVCC = 15.5 V, DVCC = 3.3 V, TMIN = 0°C, TMAX = 85°C, VREFHI = 9.5 V, VREFLO = V1 = V2 = 7 V, unless
otherwise noted.
Parameter
Conditions
1
VIDEO DC PERFORMANCE
VDE
VCME
∆V
∆V
VIDEO OUTPUT DYNAMIC PERFORMANCE
Data Switching Slew Rate
Invert Switching Slew Rate
Data Switching Settling Time to 1%
Data Switching Settling Time to 0.25%
Invert Switching Settling Time to 1%
Invert Switching Settling Time to 0.25%
Invert Switch Overshoot
CLK and Data Feedthrough2
All-Hostile Crosstalk3
Amplitude
Duration
DAC Transition Glitch Energy
VIDEO OUTPUT CHARACTERISTICS
Output Voltage Swing
CLK to VID Delay: t94
INV to VID Delay: t10
Output Current
Output Resistance
RESOLUTION
Coding
DIGITAL INPUT CHARACTERISTICS
Max. Input Data Update Rate
Data Setup Time: t1
STSQ Setup Time: t3
XFR Setup Time: t5
Data Hold Time: t2
STSQ Hold Time: t4
XFR Hold Time: t6
CLK High Time: t7
CLK Low Time: t8
CIN
IIH
IIL—All Inputs except CLK
IIL—CLK
VIH
VIL
VTH
TMIN to TMAX
DAC Code 1500 to 3200
DAC Code 1500 to 3200
DAC Code 2048
DAC Code 0 to 4095
Min
–5
–3.5
TC, MIN to TC, MAX,
VO = 5 V Step, CL = 200 pF
20% to 80%
20% to 80%
Binary
Max
Unit
+0.5
2.5
4
+5
+3.5
7.5
15
mV
mV
mV
mV
390
530
22
33
34
130
100
10
27
50
100
300
200
40
30
0.3
Code 2047 to Code 2048
AVCC – VOH, VOL– AGND
50% of VIDx
50% of VIDx
Typ
10
10.4
1.1
12
12.4
100
22
V/µs
V/µs
ns
ns
ns
ns
mV
mV p-p
mV p-p
ns
nV-s
1.3
14
14.4
V
ns
ns
mA
Ω
12
Bits
120
0
1
1
3
3
3
3
2.5
Ms/s
ns
ns
ns
ns
ns
ns
ns
ns
pF
Input tr, tf = 2 ns (10% to 90%)
3
0.05
0.6
1.2
2
0.8
1.6
Rev. 0 | Page 3 of 24
µA
µA
µA
V
V
V
AD8382
Parameter
REFERENCE INPUTS1
V1 Range
V2 Range
V1 Input Current
V2 Input Current
VREFLO Range
VREFHI Range
(VREFHI – VREFLO) Range
VREFHI Input Resistance
VREFLO Bias Current
VREFHI Input Current
VFS Range
POWER SUPPLY
DVCC, Operating Range
DVCC, Quiescent Current
AVCC, Operating Range
Total AVCC Quiescent Current
STBY AVCC Current
STBY DVCC Current
OPERATING TEMPERATURE RANGE
Ambient Temperature Range, TA
Ambient Temperature Range, TA5
Junction Temperature Range, TJ
Conditions
V2 ≥ (V1 – 0.25 V)
V2 ≥ (V1 – 0.25 V)
Min
Typ
5
5
Max
Unit
AVCC – 4
AVCC – 4
V
V
AVCC – 1.3
AVCC
2.75
µA
µA
V
V
V
5.5
kΩ
µA
µA
V
3.6
31
18
52
0.45
5
V
mA
V
mA
mA
mA
75
85
125
°C
°C
°C
0.2
–7.5
VREFHI ≤ (VREFLO + 2.75 V)
VREFHI ≤ (VREFLO + 2.75 V)
V1 – 0.5
VREFLO
0
20
–0.2
125
VFS = 2 × (VREFHI – VREFLO)
3
3.3
23
9
43
0.15
3.5
STBY = HIGH
STBY = HIGH
Still Air
100% Tested
0
0
25
VDE = differential error voltage. VCME = common-mode error voltage. ∆V = maximum deviation between outputs.
Full-scale output voltage = VFS = 2 × (VREFHI – VREFLO). See the Accuracy section on page 14.
2
Measured on two outputs differentially as CLK and DB(0:11) are driven and STSQ and XFR are held LOW.
3
Measured on two outputs differentially as the other four are transitioning by 5 V. Measured for both states of INV.
4
Measured from 50% of falling CLK edge to 50% of output change. Measurement is made for both states of INV.
5
Operation at 85°C ambient temperature requires a thermally optimized PCB layout (see Applications section), minimum airflow of 200 lfm, input clock rate not
exceeding 120 MHz, black-to-white transition ≤ 4 V, and CL ≤ 200 pF.
1
Rev. 0 | Page 4 of 24
AD8382
ABSOLUTE MAXIMUM RATINGS
Table 2. Absolute Maximum Ratings1
Parameter
Supply Voltages
AVCCx to AGNDx
DVCC to DGND
Input Voltages
Maximum Digital Input Voltage
Minimum Digital Input Voltage
Maximum Analog Input Voltage
Minimum Analog Input Voltage
Internal Power Dissipation2
LFCSP Package @ 25°C Ambient
Operating Temperature Range
Storage Temperature Range
Lead Temperature Range (Soldering 10 sec)
MAXIMUM POWER DISSIPATION
Rating
18 V
4.5 V
DVCC + 0.5 V
DGND – 0.5 V
AVCC + 0.5 V
AGND – 0.5 V
3.84 W
0°C to 85°C
–65°C to +125°C
300°C
1
Stresses above those listed under the Absolute Maximum Ratings may cause
permanent damage to the device. This is a stress rating only; functional
operation of the device at these or any other conditions above those
indicated in the operational section of this specification is not implied.
Exposure to absolute maximum ratings for extended periods may reduce
device reliability.
2
48-lead LFCSP Package:
θJA = 26°C/W (JEDEC STD, 4-layer PCB in still air)
θJC = 20°C/W.
ΨJB = 11°C/W in Still Air
The maximum power that the AD8382 can safely dissipate is
limited by its junction temperature. The maximum safe junction
temperature for plastic encapsulated devices, as determined by
the plastic’s glass transition temperature, is approximately
150°C. Temporarily exceeding this limit may cause a shift in
parametric performance due to a change in stresses exerted on
the die by the package. Exceeding a junction temperature of
175°C for extended periods can result in device failure.
OPERATING TEMPERATURE RANGE
Although the maximum safe operating junction temperature is
higher, the AD8382 is 100% tested at a junction temperature of
125°C. Consequently, the maximum guaranteed operating
junction temperature is 125°C. To ensure operation within the
specified operating temperature range, it is necessary to limit
the maximum power dissipation to:
PDMAX ≈
where TJMAX = 125°C
AD8382 ON A 4–LAYER JEDEC PCB WITH THERMALLY OPTIMIZED
LANDING PATTERN AS DESCRIBED IN THE APPLICATION NOTES
2.00
OVERLOAD PROTECTION
The thermal shutdown “debiases” the output amplifier when the
junction temperature reaches the internally set trip point. In the
event of an extended short circuit between a video output and
power supply rail, the output amplifier current continues to
switch between 0 mA and 100 mA typ. with a period set by the
thermal time constant and hysteresis of the thermal trip point.
The thermal shutdown provides long-term protection by
limiting average junction temperature to a safe level.
200 lfm
1.75
POWER DISSIPATION (W)
The AD8382 employs a two-stage overload protection circuit
that consists of an output current limiter and a thermal
shutdown. The maximum current at any output of the AD8382
is internally limited to 100 mA average. In the event of a
momentary short circuit between a video output and a power
supply rail (VCC or AGND), the output current limit is
sufficiently low to provide temporary protection.
(TJMAX–TA )
(θ JA –0.9 × 3 Airflow in lfm)
120MHz
500 lfm
1.50
STILL AIR
1.25
60Hz XGA
1.00
0.75
0.50
Quiescent
65
70
75
80
85
90
95
100
105
110
115
MAXIMUM AMBIENT TEMPERATURE (°C)
Figure 2. Maximum Power Dissipation vs. Temperature.
Note: Quiescent power dissipation is 0.74 W when operating
under the conditions specified in this data sheet.
To ensure a high degree of reliability, the exposed paddle must
be electrically connected to AVCC.
When driving a 6-channel XGA panel with an input capacitance
of 200 pF, the AD8382 dissipates a total of 1.14 W when
displaying 1 pixel wide alternating white and black vertical lines
generated by a standard 60 Hz XGA input video.
To ensure optimized thermal performance, the exposed paddle
must be thermally connected to the AVCC plane as described in
the Applications section.
The total power dissipation of the AD8382 is 1.67 W when
operating at the maximum specified frequency of 120 MHz,
under the conditions specified in this data sheet (Figure 2).
EXPOSED PADDLE
Rev. 0 | Page 5 of 24
AD8382
TIMING CHARACTERISTICS
Table 3. Timing Parameters and Conditions
Parameter
Conditions
Min
t1, Data Setup Time
tr, tf = 2 ns (10% to 90%)
Typ
Max
Unit
0
ns
t2, Data Hold Time
3
ns
t3, STSQ Setup Time
1
ns
t4, STSQ Hold Time
3
ns
t5, XFR Setup Time
1
ns
t6, XFR Hold Time
3
ns
t7, CLK High Time
3
ns
t8, CLK Low Time
2.5
ns
t9, CLK to VIDx Delay
To 50% of VIDx
10
12
14
ns
t10, INV to VIDx Delay
To 50% of VIDx
10.4
12.4
14.4
ns
tf
tr
t8
CLK
t7
t1
VTH
CLK
VTH
t2
XFR
VTH
VTH
DB(0:11)
VTH
INV
STSQ
t3
t4
t5
t6
VTH
VID(0:5)
VTH
XFR
t9
t10
Figure 3. Timing Requirement E/O = HIGH
t8
Figure 5. Output Timing
t7
VTH
CLK
t1
t2
VTH
DB(0:11)
STSQ
t3
XFR
50%
t5
t6
t4
VTH
VTH
Figure 4. Timing Requirement E/O = LOW
Rev. 0 | Page 6 of 24
AD8382
DB0 1
37 AGND0
38 V2
39 VREFLO
40 VREFHI
41 AGNDDAC
42 AVCCDAC
43 V1
44 NC
45 NC
46 STSQ
47 XFR
48 CLK
PIN CONFIGURATION AND FUNCTIONAL DESCRIPTIONS
36
VID0
35
AVCC0,1
DB2 3
34
VID1
DB3 4
33
AGND1,2
DB4 5
32
VID2
31
AVCC2,3
30
VID3
DB7 8
29
AGND3,4
DB8 9
28
VID4
DB9 10
27
AVCC4,5
DB10 11
26
VID5
DB11 12
25
AGND5
PIN 1
INDICATOR
DB1 2
AD8382
DB5 6
NC 24
NC 23
AGNDBIAS 22
BYP 21
STBY 20
NC 19
AVCCBIAS 18
DGND 16
INV 15
R/L 14
E/O 13
DVCC 17
TOP VIEW
(Not to Scale)
DB6 7
NC = NO CONNECT
Figure 6. 48-Lead LFCSP, 7 mm × 7 mm Package
Table 4. Pin Function Descriptions
Mnemonic
DB(0:11)
CLK
STSQ
Function
Data Input
Clock
Start Sequence
Description
12-Bit Data Input. MSB = DB(11).
Clock Input.
A new data loading sequence begins on the rising edge of CLK when this input was HIGH on
the preceding rising edge of CLK and the E/O input is held HIGH. A new data loading sequence
begins on the falling edge of CLK when this input was HIGH on the preceding falling edge of
CLK and the E/O input is held LOW.
A new data loading sequence begins on the left, with Channel 0, when this input is LOW, and
on the right, with Channel 5, when this input is HIGH.
The active CLK edge is the rising edge when this input is held HIGH and the falling edge when
this input is held LOW. Data is loaded sequentially on the rising edges of CLK when this input is
HIGH and on the falling edges when this input is LOW.
Data is transferred to the outputs on the immediately following falling edge of CLK when this
input is HIGH on the rising edge of CLK.
These pins are directly connected to the analog inputs of the LCD panel.
The voltages applied between these pins and AGND set the reference levels of the analog
outputs.
The voltage applied between these pins sets the full-scale output voltage.
R/L
Right/Left Select
E/O
Even/Odd Select
XFR
Data Transfer
VID0–VID5
V1,V2
Analog Outputs
Reference Voltages
VREFHI,
VREFLO
INV
Full-Scale References
Invert
When this pin is HIGH, the analog output voltages are at or above V2. When this pin is LOW,
the analog output voltages are at or below V1.
DVCC
DGND
AVCCx
AGNDx
BYP
STBY
Digital Power Supply
Digital Supply Return
Analog Power Supplies
Analog Supply Returns
Bypass
Standby
Digital Power Supply.
This pin is normally connected to the analog ground plane.
Analog Power Supplies.
Analog Supply Returns.
A 0.1 µF capacitor connected between this pin and AGND ensures optimum settling time.
When HIGH, the internal circuits are debiased and the power dissipation drops to a minimum.
Rev. 0 | Page 7 of 24
AD8382
TYPICAL PERFORMANCE CHARACTERISTICS
1.00
1.50
0.75
1.25
0.50
1.00
7V
–0.25
–0.50
0.50
0.25
0.00
–0.75
–0.25
–1.00
–0.50
–1.25
–0.75
–1.50
–20
OUTPUT (%)
0.75
OUTPUT (%)
0.00
0
20
40
60
80
100
120
140
160
–1.00
–20
180
40
60
80
100
120
140
160
TIME (ns)
0.75
1.25
0.50
1.00
0.25
0.75
12V
–0.25
–0.50
0.25
0.00
–0.25
–1.00
–0.50
–1.25
–0.75
20
40
60
80
100
120
140
160
–1.00
–20
180
180
0.50
–0.75
0
20
Figure 10. Output Settling Time (Falling Edge),
CL = 200 pF, 5 V Step, INV = LOW
1.50
–1.50
–20
0
TIME (ns)
1.00
0.00
2V
Figure 7. Output Settling Time (Rising Edge),
CL = 200 pF, 5 V Step, INV = LOW
OUTPUT (%)
OUTPUT (%)
0.25
7V
0
20
40
60
80
100
120
140
160
TIME (ns)
TIME (ns)
Figure 8. Output Settling Time (Rising Edge),
CL = 200 pF, 5 V Step, INV = HIGH
Figure 11. Output Settling Time (Falling Edge),
CL = 200 pF, 5 V Step, INV = HIGH
180
0pF, 12V
0pF, 7V
150pF, 12V
47pF, 7V
200pF, 12V
250pF, 12V
0.25%/DIV
100pF, 12V
OUTPUT (%)
0.25%/DIV
OUTPUT (%)
47pF, 12V
100pF, 7V
150pF, 7V
300pF, 12V
200pF, 7V
250pF, 7V
300pF, 7V
–15
0
15
30
45
60
75
90
105
120
135
–15
0
15
30
45
60
75
90
105
120
TIME (ns)
TIME (ns)
Figure 9. Output Settling Time (Rising Edge) vs. CL,
5 V Step, INV = HIGH
Figure 12. Output Settling Time (Falling Edge) vs. CL,
5 V Step, INV = HIGH
Rev. 0 | Page 8 of 24
135
AD8382
SWITCHING STEP RESPONSE (V)
12
SWITCHING STEP RESPONSE (V)
12
2
2
20ns/DIV
TIME (ns)
TIME (ns)
Figure 13. Invert Switching Step Response (Rising Edge),
10 V Step, CL = 200 pF
Figure 16. Invert Switching Step Response (Falling Edge),
10 V Step, CL = 200 pF
SWITCHING STEP RESPONSE (V)
SWITCHING STEP RESPONSE (V)
20ns/DIV
7
2
7
2
20ns/DIV
TIME (ns)
TIME (ns)
Figure 14. Data Switching Step Response (Rising Edge),
5 V Step, CL=200 pF, INV = LOW
Figure 17. Data Switching Step Response (Falling Edge),
5 V Step, CL = 200 pF, INV = LOW
SWITCHING STEP RESPONSE (V)
SWITCHING STEP RESPONSE (V)
20ns/DIV
12
7
12
7
20ns/DIV
20ns/DIV
TIME (ns)
TIME (ns)
Figure 15. Data Switching Step Response (Rising Edge),
5 V Step, CL = 200 pF, INV = HIGH
Figure 18. Data Switching Step Response (Falling Edge),
5 V Step, CL = 200 pF, INV = HIGH
Rev. 0 | Page 9 of 24
2.0
1.5
1.5
1.0
1.0
0.5
0.5
0
–0.5
INL (LSB)
0
–0.5
–1.0
–1.0
–1.5
–1.5
–2.0
0
512
1024
1536
2048
2564
3072
3584
–2.0
4096
0
512
1024
1536
2048
2564
3072
3584
4096
INPUT CODE
INPUT CODE
Figure 19. Differential Nonlinearity (DNL) vs. Code, INV = LOW
Figure 22. Differential Nonlinearity (DNL) vs. Code, INV = HIGH
2.0
2.0
1.5
1.5
1.0
1.0
0.5
0.5
0
0
–0.5
–0.5
–1.0
–1.0
–1.5
–1.5
–2.0
0
512
1024
1536
2048
2564
3072
3584
–2.0
4096
512
1024
1536
2048
2564
3072
3584
4096
INPUT CODE
INPUT CODE
Figure 23. Integral Nonlinearity (INL) vs. Code, INV = HIGH
3.500
5.00
2.625
3.75
1.750
2.50
0.875
1.25
0
0
–0.875
–1.25
–1.750
–2.50
–2.625
–3.75
–3.500
0
Figure 20. Integral Nonlinearity (INL) vs. Code, INV = LOW
VDE (mV)
VCME (mV)
DNL (LSB)
2.0
INL (LSB)
DNL (LSB)
AD8382
0
512
1024
1536
2048
2564
3072
3584
4096
–5.00
0
512
1024
1536
2048
2564
3072
3584
INPUT CODE
INPUT CODE
Figure 21. Common-Mode Error Voltage (VCME) vs. Code
Figure 24. Differential Error Voltage (VDE) vs. Code
Rev. 0 | Page 10 of 24
4096
4
4
3
3
NORMALIZED VDE, VCME (mV)
NORMALIZED VDE, VCME (mV)
AD8382
2
1
VDE
0
VCME
–1
–2
1
VDE
0
VCME
–1
–2
–3
–3
–4
2
0
1.0
2.0
3.0
4.0
5.0
–4
5.0
6.0
6.0
7.0
8.0
9.0
10.0
11.0
V1 = V2 (V)
V2 – V1 (V)
Figure 28. Normalized VDE, VCME vs. V1 = V2 at Code 2048
Figure 25. Normalized VDE, VCME vs. (V2 – V1) at Code 2048
10
4
NORMALIZED VDE, VCME (mV)
NORMALIZED VDE (mV)
3
5
0
–5
V1 = V2 = 5V, VFS = 3V
2
1
VCME
0
VDE
–1
–2
–3
–10
4
5
6
7
8
V1 (V) @ V2 = 7V
9
10
11
–4
12
3.75
1.750
2.50
0
–1.25
–1.750
–2.50
–2.625
–3.75
30
40
50
10
11
12
13
14
60
70
80
90
100
CODE 2048
0
–0.875
20
9
1.25
CODE 2048
VDE (mV)
VCME (mV)
2.625
10
8
VREFLO (V)
5.00
0
7
Figure 29. Normalized VDE, VCME vs. VREFLO at Code 2048
3.500
–3.500
6
V2 (V) @ V1 = 7V
Figure 26. Normalized VDE vs. V1 and V2 at Code 2048
0.875
5
–5.00
0
10
20
30
40
50
60
70
80
90
100
TEMPERATURE (°C)
TEMPERATURE (°C)
Figure 27. Common-Mode Error Voltage (VCME) vs. Temperature
Figure 30. Differential Error Voltage (VDE) vs. Temperature
Rev. 0 | Page 11 of 24
AD8382
7.05
7.05
20ns/DIV
V1 = V2 = 7V
7.04
7.03
7.03
7.02
7.02
7.01
7.01
(VID4 – VID5)
VID (V)
VID (V)
V1 = V2 = 7V
7.04
7.00
6.09
(VID4 – VID5)
7.00
6.09
VID0, 1, 2, 3
6.08
6.08
5V
6.07
6.07
6.06
6.06
6.05
6.05
3.3V
DB (0:11)
20ns/DIV
TIME (ns)
TIME (ns)
Figure 31. All-Hostile Crosstalk at CL = 200 pF
Figure 34. Data Switching Transient (Feedthrough) at CL = 200 pF
900
800
800
700
SLEW RATE (V/µs)
SLEW RATE (V/µs)
700
600
500
10% TO 90%
400
600
500
10% TO 90%
400
20% TO 80%
20% TO 80%
300
200
300
0
50
100
150
200
250
200
300
100
150
200
250
LOAD CAPACITANCE (pF)
LOAD CAPACITANCE (pF)
Figure 35. Slew Rate vs. CL (Rising Edge)
–10
–20
–30
PSR (dB)
50
Figure 32. Slew Rate vs. CL (Falling Edge)
0
–40
INV = LOW
–50
INV = HIGH
–60
–70
–80
10
0
100
1k
10k
100k
1M
10M
FREQUENCY (Hz)
Figure 33. AVCC Power Supply Rejection vs. Frequency
Rev. 0 | Page 12 of 24
300
AD8382
FUNCTIONAL DESCRIPTION
The AD8382 is a system building block designed to directly
drive the columns of LCD microdisplays of the type
popularized for use in projection systems. It comprises six
channels of precision, 12-bit digital-to-analog converters loaded
from a single, high speed, 12-bit wide input. Precision current
feedback amplifiers, providing well-damped pulse response and
fast voltage settling into large capacitive loads, buffer the six
outputs. Laser trimming at the wafer level ensures low absolute
output errors and tight channel-to-channel matching. Tight
part-to-part matching in high resolution systems is guaranteed
by the use of external voltage references.
V1, V2 INPUTS—VOLTAGE REFERENCE INPUTS
Two external analog voltage references set the levels of the
outputs. V1 sets the output voltage at Code 4095 while the INV
input is LOW, and V2 sets the output voltage at Code 4095 while
the INV input is held HIGH.
VREFHI, VREFLO INPUTS—FULL-SCALE
REFERENCE INPUTS
Twice the difference between these analog input voltages sets
the full-scale output voltage VFS.
START SEQUENCE CONTROL—INPUT DATA
LOADING
VFS = 2 × (VREFHI – VREFLO)
A valid STSQ control input initiates a new 6-clock loading
cycle, during which six input data-words are loaded sequentially
into six internal channels. A new loading sequence begins on
the current active CLK edge only when STSQ was held HIGH at
the preceding active CLK edge. Active CLK edge is defined by
the E/O Control.
INV CONTROL—ANALOG OUTPUT INVERSION
EVEN/ODD CONTROL—INPUT DATA LOADING
The AD8382 has two regions of operation, where the video
output voltages are either above reference voltage V2 or below
reference voltage V1. The transfer function defines the video
output voltage as the function of the digital input code as
follows:
To facilitate 12-channel, single data bus systems, the active CLK
edge, at which input data is loaded, is selected with the E/O
control input. Input data is loaded on the rising CLK edges
while the E/O input is held HIGH; input data is loaded on the
falling CLK edges while the E/O input is held LOW.
The analog voltage equivalent of the input code is subtracted
from (V2 + VFS) while INV is held HIGH and added to (V1 –
VFS) while INV is held LOW.
Transfer Function
VIDx(n) = V2 + VFS × (1 – n/4095), for INV = HIGH
VIDx(n) = V1 – VFS × (1 – n/4095), for INV = LOW
RIGHT/LEFT CONTROL—INPUT DATA LOADING
To facilitate image mirroring, the direction of the loading
sequence is set by the R/L control. A new loading sequence
begins at channel 0 and proceeds to Channel 5 when the R/L
control is held LOW. It begins at Channel 5 and proceeds to
channel 0 when the R/L control is held HIGH.
where n = input code
VFS = 2 × (VREFHI – VREFLO)
A number of internal limits define the usable range of the video
output voltages, VIDx, as shown in Figure 36
AVCC
≥ 1.3V
V2 + VFS
XFR CONTROL—DATA TRANSFER TO OUTPUTS
INV = HIGH
0 ≤ VFS ≤ 5.5V
VIDx (V)
Data transfer to the outputs is initiated by the XFR control.
While XFR is held HIGH during a rising CLK edge, data is
simultaneously transferred to all outputs on the immediately
following falling CLK edge.
INTERNAL LIMITS AND
USABLE VOLTAGE RANGES
9V ≤ AVCC ≤ 18V
V2
5V ≤ V2 ≤ (AVCC – 4)
V1
STBY CONTROL—STANDBY MODE
0 ≤ VFS ≤ 5.5V
A HIGH applied to the STBY input debiases the internal
circuitry, dropping the quiescent power dissipation to a few
milliwatts. Upon returning STBY to LOW, normal operation is
restored. Since both analog and digital circuitry is debiased, all
stored data will be lost in standby mode.
5V ≤ V1 ≤ (AVCC – 4)
INV = LOW
V1 – VFS
≥ 1.3V
AGND
0
4095
INPUT CODE
Figure 36. Transfer Function and Usable Voltage Ranges
Rev. 0 | Page 13 of 24
AD8382
Accuracy
VDE, the differential error voltage, measures the difference
between the rms value of the output and the rms value of the
ideal. The defining expression is:
VDE(n) =
AVCC
(V2 + VFS)
VIDx (V)
To best correlate transfer function errors to image artifacts, the
overall accuracy of the AD8382 is defined by three parameters:
VDE, VCME, and ∆V.
V2
VOUTN(n)
V1
[VOUTN (n) – V 2] − [VOUTP(n) − V 1]
n
− 1 −
× VFS
2
4095
VOUTP(n)
(V1 – VFS)
VCME, the common-mode error voltage, measures one-half the
dc bias of the output. The defining expression is:
AGND
0
n
4095
INPUT CODE
VCME(n) =
1 VOUTN (n) +VOUTP(n) (V 2 +V 1)
–
2
2
2
Figure 37. AD8382 Transfer Function
∆V measures the maximum deviation between the output
voltages. The defining expression is:
∆V (n) = max{∆VN(n), ∆VP(n)}
where ∆VN(n) = max{VOUTN(n)(0–5)} – min{VOUTN(n)(0–5)}
and ∆VP(n) = max{VOUTP(n)(0–5) } – min{VOUTP(n)(0–5)}
Rev. 0 | Page 14 of 24
AD8382
APPLICATIONS
PIXEL
CLK
OPERATING MODES—6-CHANNEL SYSTEMS
–1
0
1
2
3
4
5
6
7
8
9
10
11
12
CLK
CLK
STSQ
STSQ
XFR
STSQ
R/L
XFRF
E/O
EVEN
0
9
10
4
CH 4
CH 5
8
3
CH 3
E/O
ODD
7
2
CH 2
12
6
1
CH 1
–1
5
11
VID0
–6
0
6
VID1
–5
1
7
VID2
–4
2
8
VID3
–3
3
9
VID4
–2
4
10
VID5
–1
5
11
AD8382 EVEN
INTERNAL LATCHES
CH 0
OUTPUTS
OUTPUTS
INTERNAL LATCHES
INPUTS
DB(0:11)
DB(0:11) –3 –2 –1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
INPUTS
Depending on the speed of the LCD microdisplay, 6-channel
systems are compatible with up to XGA resolutions and require
one AD8382 per color. The input/output timing diagram of the
AD8382 in such systems is shown in Figure 38.
12-channel systems are usually those requiring video line
doubling or compatibility with SXGA and higher resolutions.
Depending on the input data rates, two types of 12-channel
systems are in common use.
12-Channel, Even/Odd Systems
Single data bus systems are characterized by an image processor
with a single data bus output. They require two AD8382s per
color.
OUTPUTS
OPERATING MODES—12-CHANNEL SYSTEMS
AD8382 ODD
INTERNAL LATCHES
Figure 38. Timing Diagram in a Typical 6-Channel System, E/O=HIGH,
R/L=LOW
One AD8382 is set to operate in EVEN mode, while the other is
set to operate in ODD mode. Both AD8382s share the same
data bus and CLK. The timing diagram of such a system is
shown in Figure 39.
0
CH 0
12
2
CH 1
14
4
CH 2
16
6
CH 3
18
8
CH 4
20
–2
CH 5
10
22
VID0
–12
0
12
VID1
–10
2
14
VID2
–8
4
16
VID3
–6
6
18
VID4
–4
8
20
VID5
–2
10
22
1
CH 0
15
5
CH 2
17
7
CH 3
CH 4
13
3
CH 1
–3
19
9
21
CH 5
–1
11
23
VID0
–11
1
13
VID1
–9
3
15
VID2
–7
5
17
VID3
–5
7
19
VID4
–3
9
21
VID5
–1
11
23
Figure 39. 12-Channel Even/Odd System Timing Diagram
OPERATING MODES—BEYOND 12 CHANNELS
12-Channel Parallel Systems
Dual data bus systems are characterized by an image processor
with two data bus outputs. They require two AD8382s per color.
Both AD8382s in dual data bus systems can be set
independently. The timing diagram of each AD8382 in such
systems is identical to that of a 6-channel system.
Any number of AD8382s may be cascaded in even/odd pairs or
in parallel to facilitate very high resolution systems.
Rev. 0 | Page 15 of 24
AD8382
IMAGE PROCESSOR
STSQ1
CLK
CLK
CLK
DB(0:11)
CLK
XFR
R/L
Pixel CLK ÷2
÷6 COUNTER
CLK
XFR
R/L
STSQ1
INV1
E/O1
H. REVERSE
CLK
AD8382
DB1(0:11)
STSQ2
STSQ
INV
E/O
VREFHI
VREFLO
V2
V1
STSQ2
INV2
E/O2
÷6 COUNTER
HSTART
VID0
Ch 0
VID1
Ch 2
VID2
Ch 4
VID3
Ch 6
VID4
Ch 8
VID5
Ch 10
12 – CHANNEL
LCD
AD8382
DB(0:11)
HSYNC
CLK
XFR
R/L
INV1
INV2
VSYNC
STSQ
INV
E/O
VREFHI
VREFLO
V2
V1
REFERENCES
VREFHI
VREFLO
V2
V1
VID0
CH 1
VID1
CH 3
VID2
CH 5
VID3
CH 7
VID4
CH 9
VID5
CH 11
Figure 40. Single Data Bus 12-Channel Even/Odd System Block Diagram
IMAGE PROCESSOR
D(0:9) Odd
D(0:9) Even
H. REVERSE
÷6 COUNTER
HSTART
DB(0:11)
CLK
XFR
R/L
Pixel CLK ÷2
CLK
AD8382
DB1(0:11)
“1”
CLK
XFR
R/L
STSQ
INV1
E/O
STSQ
INV
E/O
VREFHI
VREFLO
V2
V1
INV2
VID0
Ch 0
VID1
Ch 2
VID2
Ch 4
VID3
Ch 6
VID4
Ch 8
VID5
Ch 10
12 – CHANNEL
LCD
AD8382
D(0:9) Even
DB2(0:11)
D(0:9) Odd
HSYNC
VSYNC
DB(0:11)
CLK
XFR
R/L
INV1
INV2
STSQ
INV
E/O
VREFHI
VREFLO
V2
V1
REFERENCES
VREFHI
VREFLO
V2
V1
Figure 41. Dual Data Bus 12-Channel Even/Odd System Block Diagram
Rev. 0 | Page 16 of 24
VID0
CH 1
VID1
CH 3
VID2
CH 5
VID3
CH 7
VID4
CH 9
VID5
CH 11
AD8382
APPLICATIONS CIRCUIT
VBIAS Generation—V1, V2 Input Pin
Functionality
In order to avoid image flicker, a symmetrical ac voltage is
required and a bias voltage of approximately 1 V minimum
must be maintained across the pixels of HTPS LCDs. The
AD8382 provides two methods of maintaining this bias voltage.
The circuit in Figure 41 ensures VBIAS symmetry to within 1
mV with a minimum component count. Bypass capacitors are
omitted for clarity.
AVCC = 15.5V
VZ = 5.1V
INTERNAL BIAS VOLTAGE GENERATION
Standard systems that internally generate the bias voltage
reserve the uppermost code range for the bias voltage and use
the remaining code range to encode the video for gamma
correction. A high degree of ac symmetry is guaranteed by the
AD8382 in these systems.
1
AD8382
3
–IN
V+
V2 = 8V
5
V2
2 VCOM
AD8132
VCOM = 7V
8
R2 = 1kΩ
+IN
V–
4
V1 = 6V
V1
6
R1 = 6kΩ
The V1 and V2 inputs in these systems are tied together and are
normally connected to VCOM, as shown in Figure 42.
DVCC = 3.3V
Figure 43. External VBIAS Generator with the AD8132
VFS = 5V
AD8382
VCOM
V2
V1
VCOM
VBIAS = 1V
VBIAS = 1V
VFS = 4V
3280 4095
V2
VFS = 5V
VBIAS = 1V
VCOM
RESERVED
CODE
RANGE
VBIAS = 1V
V1
4095
VFS = 4V
Figure 42. V1, V2 Connection and Transfer Function in a Typical Standard
System
Figure 44. The AD8382 Transfer Function in a Typical High Accuracy System
EXTERNAL BIAS VOLTAGE GENERATION
In systems that require improved brightness resolution and
higher accuracy, the V1 and V2 inputs, connected to external
voltage references, provide necessary bias voltage, VBIAS, while
allowing the full code range to be used for gamma correction.
V1 = VCOM − VBIAS
V2 = VCOM + VBIAS
7.50
6.25
5.00
(V2 + V1)/2 – VCOM (mV)
To ensure a symmetrical ac voltage at the AD8382’s outputs,
VBIAS must remain constant for both states of INV. Thus, V1
and V2 are defined as:
8.75
3.75
TA = 25°C
TA = 85°C
2.50
1.25
0.00
–1.25
–2.50
–3.75
–5.00
–6.25
–7.50
–8.75
5.7
6.2
6.7
7.2
7.7
8.2
8.7
9.2
9.7
10.2
10.7
V+ – V– (V)
Figure 45. Typical Asymmetry at the Outputs of the AD8132 vs. Power Supply
for the Application Circuit
The AD8132 typically produces a symmetrical output at 85°C
when its supply, (V+) – (V–), is 7.2 V (Figure 45).
Rev. 0 | Page 17 of 24
AD8382
Power Supply Sequencing
As indicated in the Absolute Maximum Ratings, voltage at any
input pin cannot exceed its supply voltage by more than 0.5 V.
To ensure compliance with these ratings, the following powerup and power-down sequencing is recommended.
During power-up, initial application of nonzero voltages to any
input pin must be delayed until supply voltage ramps up to at
least the highest maximum operational input voltage. During
power-down, the voltage at any input pin must reach zero
during a period not exceeding the power supply’s hold-up time.
Failure to comply with the Absolute Maximum Ratings may
result in functional failure or damage to the internal ESD
diodes. Damaged ESD diodes may cause temporary parametric
failures, which may result in image artifacts. Damaged ESD
diodes cannot provide full ESD protection, reducing reliability.
The AD8382 package is designed to provide superior thermal
characteristics, partly through the exposed die paddle on the
bottom surface of the package. In order to take full advantage of
this feature, the exposed paddle must be in direct thermal
contact with the PCB, which then serves as a heat sink.
Table 5. AD8382 Power Dissipation
CLOAD
(pF)
PQUIESCENT
(W)
200
250
300
0.74
0.74
0.74
VSWING = 5 V
PDYNAMIC
PTOTAL
(W)
(W)
0.93
1.67
1.16
1.90
1.39
2.13
VSWING = 4 V
PDYNAMIC
PTOTAL
(W)
(W)
0.74
1.48
0.93
1.67
1.11
1.85
Power ON
Power OFF
A thermally effective PCB must incorporate a thermal pad and
a thermal via structure. The thermal pad provides a solderable
contact surface on the top surface of the PCB. The thermal via
structure provides a thermal path to the inner and bottom
layers of the PCB to remove heat.
1. Apply power to supplies.
1. Remove power from I/Os.
THERMAL PAD DESIGN
2. Apply power to other I/Os.
2. Remove power from
supplies.
Thermal performance of the AD8382 varies logarithmically
with the contact area between the exposed thermal paddle and
the thermal pad on the top layer of the PCB.
PCB Design for Optimized Thermal
Performance
The total maximum power dissipation of the AD8382 is partly
load dependent. In a 6-channel, 65 MHz, 60 Hz XGA system,
the total maximum power dissipation is 1.14 W at an LCD input
capacitance of 200 pF.
At a clock rate of 120 Ms/s, the total maximum power
dissipation can exceed 2 W, as shown below for a black-to-white
video output voltage swing of 4 V and 5 V.
Although the maximum safe operating junction temperature is
higher, the AD8382 is 100% tested at a junction temperature of
125°C. Consequently, the maximum guaranteed operating
junction temperature is 125°C. To limit the maximum junction
temperature at or below the guaranteed maximum, the package,
in conjunction with the PCB, must effectively conduct heat
away from the junction.
In order to minimize thermal performance degradation of
production PCBs, the contact area between the thermal pad and
the PCB should be maximized. Therefore, the size of the
thermal pad should match the exposed paddle size of 5.25 mm
× 5.25 mm. In addition, a second thermal pad of the same size
should be placed on the bottom side of the PCB. At least one
thermal pad should be in direct thermal (and electrical) contact
with the AVCC plane.
THERMAL VIA STRUCTURE DESIGN
Effective heat transfer from the top to the inner and bottom
layers of the PCB requires thermal vias incorporated into the
thermal pad design. Thermal performance increases
logarithmically with the number of vias. θJA reaches its specified
value at approximately 16 vias, provided the AD8382 is on a
standard JEDEC PCB. θJA approaches its optimum value as the
slope of such a curve approaches zero, at above 36 vias. Near
optimum thermal performance of production PCBs is attained
when the number of thermal vias is at least 36.
Rev. 0 | Page 18 of 24
AD8382
AD8382 PCB DESIGN RECOMMENDATIONS
SOLDER MASKING
7 mm
To minimize the formation of solder voids due to solder flowing
into the via holes (solder wicking), via diameter should be small.
Solder masking of the via holes on the top layer of the PCB
plugs the via holes, inhibiting solder flow into the holes. To
optimize the thermal pad coverage, the solder mask diameter
should be no more than 0.1 mm larger than the via diameter.
Land pattern Dimensions
Pad Size:
Pad Pitch:
7 mm
0.5 mm × 0.25 mm
0.5 mm
Solder Mask—Top Layer
Thermal Pad Size:
Pads: Set by customer’s PCB
Design Rules
5.25 mm × 5.25 mm
Thermal via structure:
LAND PATTERN – TOP LAYER
Figure 46. Land Pattern—Top Layer
0.25 mm diameter.
Vias on 0.5mm grid.
Thermal Vias: 0.25 mm dia.
circular mask, centered on the
vias.
THERMAL PAD AND VIA CONNECTIONS
Solder Mask—Bottom
Layer
Thermal Pads are connected to AVCC.
Set by customer’s PCB Design
Rules.
For PCBs with the AVCC
plane located on one of
the outer layers, direct
connection of at least one
thermal pad to the AVCC
plane is recommended.
For PCBs with the AVCC
plane located on one of
the internal layers, direct
connection of all thermal
vias to the AVCC plane is
recommended.
LAND PATTERN – BOTTOM LAYER
Figure 47. Land Pattern—Bottom Layer
The use of thermal spokes is not recommended when
connecting the thermal pads or via structure to the AVCC
plane.
Rev. 0 | Page 19 of 24
SOLDER MASK – TOP LAYER
Figure 48. Solder Mask—Top Layer
AD8382
Layout Considerations
POWER SUPPLY BYPASSING
The AD8382 is a mixed-signal, high speed, high accuracy
device. In order to fully realize its specifications, it is essential to
use a properly designed printed circuit board.
All power supply and reference pins of the AD8382 must be
properly bypassed to the analog ground plane for optimum
performance.
LAYOUT AND GROUNDING
The analog outputs and the digital inputs of the AD8382 are on
opposite sides of the package. Keep these sections separated to
minimize crosstalk and coupling of digital inputs into the
analog outputs.
All signal trace lengths should be made as short and direct as
possible to prevent signal degradation due to parasitic effects.
Note that digital signals should not cross and should not be
routed near analog signals.
It is imperative to provide a solid analog ground plane under
and around the AD8382. All ground pins of the part should be
connected directly to this ground plane with no extra signal
path length. This includes DGND, AGNDBIAS, AGND5,
AGND3,4, AGND1,2, AGND0, and AGNDDAC. The return
traces for any of the signals should be routed close to the
ground pin for that section to prevent stray signals from
coupling into other ground pins.
All analog supply pins may be connected directly to an analog
supply plane located as close to the part as possible. A 0.1 µF
chip capacitor should be placed as close to each analog supply
pin as possible and connected directly between each analog
supply pin and the analog ground plane.
A minimum 47 µF tantalum capacitor should be placed near the
analog supply plane and connected directly between the supply
and analog ground planes.
A minimum 10 µF tantalum capacitor should be placed near the
digital supply pin and connected directly to the analog ground
plane. A 0.1 µF chip capacitor should be connected between the
digital supply pin and the analog ground.
VREFHI, VREFLO, V2, V1 REFERENCE
DISTRIBUTION
To ensure well-matched video outputs, all AD8382s must
operate from equal reference voltages.
Each reference voltage should be distributed to each AD8382
directly from the source of the reference voltage with
approximately equal trace lengths.
A 0.1 µF chip capacitor should be placed as close to each
reference input pin as possible and directly connected between
the reference input pin and the analog ground plane.
Rev. 0 | Page 20 of 24
AD8382
OUTLINE DIMENSIONS
7.00
BSC SQ
0.60 MAX
0.60 MAX
37
36
PIN 1
INDICATOR
6.75
BSC SQ
TOP
VIEW
0.20
REF
25
24
PIN 1
INDICATOR
1
5.25
5.10 SQ
4.95
12
13
5.50
REF
0.80 MAX
0.65 NOM
12° MAX
48
BOTTOM
VIEW
0.50
0.40
0.30
1.00
0.90
0.80
0.30
0.23
0.18
0.05 MAX
0.02 NOM
0.50 BSC
SEATING
PLANE
COPLANARITY
0.08
COMPLIANT TO JEDEC STANDARDS MO-220-VKKD-2
Figure 49. 48-Lead Frame Chip Scale Package [LFCSP] (CP-48)—Dimensions shown in millimeters
ESD 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 AD8382 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.
Ordering Guide
Table 6.
Model
Temperature Range
Package Description
Package Option
AD8382ACP
0°C to 85°C
48-Lead LFCSP
CP-48
Rev. 0 | Page 21 of 24
AD8382
Rev. 0 | Page 22 of 24
AD8382
Rev. 0 | Page 23 of 24
AD8382
© 2003 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective companies.
Printed in the U.S.A.
C03371-0-2/03(0)
Rev. 0 | Page 24 of 24