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 that
may result from its use. No license is granted by implication or otherwise
under any patent or patent rights of Analog Devices.
a
AD9826
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781/329-4700www.analog.com
Fax: 781/326-8703 © Analog Devices, Inc., 2001
Complete 16-Bit Imaging
Signal Processor
FUNCTIONAL BLOCK DIAGRAM
OEB
DOUT
SCLK
SLOAD
SDATA
ADCCLKCDSCLK2CDSCLK1
OFFSET
VINB
VING
VINR
AVDD AVSS
CML
CAPT
AVDD
AVSS
BANDGAP
REFERENCE
AD9826
DRVDD DRVSS
GAIN
REGISTERS
OFFSET
REGISTERS
3:1
MUX
16:8
MUX
16-BIT
ADC
DIGITAL
CONTROL
INTERFACE
CONFIGURATION
REGISTER
RED
GREEN
BLUE
MUX
REGISTER
8
16
RED
GREEN
BLUE
PGA
6
9
PGA
PGA
CDS
9-BIT
DAC
CDS
9-BIT
DAC
CDS
9-BIT
DAC
INPUT
CLAMP
BIAS
CAPB
FEATURES
16-Bit 15 MSPS A/D Converter
3-Channel 16-Bit Operation up to 15 MSPS
1-Channel 16-Bit Operation up to 12.5 MSPS
2-Channel Mode for Mono Sensors with Odd/Even Outputs
Correlated Double Sampling
1~6ⴛ Programmable Gain
ⴞ300 mV Programmable Offset
Input Clamp Circuitry
Internal Voltage Reference
Multiplexed Byte-Wide Output
Optional Single Byte Output Mode
3-Wire Serial Digital Interface
3 V/5 V Digital I/O Compatibility
28-Lead SSOP Package
Low Power CMOS: 400 mW (Typ)
Power-Down Mode Available
APPLICATIONS
Flatbed Document Scanners
Digital Copier
Multifunction Peripherals
Infrared Imaging Applications
Machine Vision
PRODUCT DESCRIPTION
The AD9826 is a complete analog signal processor for imaging
applications. It features a 3-channel architecture designed to
sample and condition the outputs of trilinear color CCD arrays.
Each channel consists of an input clamp, Correlated Double
Sampler (CDS), offset DAC, and Programmable Gain Amplifier
(PGA), multiplexed to a high-performance 16-bit A/D converter.
The AD9826 can operate at speeds greater than 15 MSPS with
reduced performance.
The CDS amplifiers may be disabled for use with sensors that
do not require CDS, such as Contact Image Sensors (CIS),
CMOS active pixel sensors, and Focal Plane Arrays.
The 16-bit digital output is multiplexed into an 8-bit output word,
which is accessed using two read cycles. There is an optional
single byte output mode. The internal registers are programmed
through a 3-wire serial interface, and provide adjustment of
the gain, offset, and operating mode.
The AD9826 operates from a single 5 V power supply, typically
consumes 400 mW of power, and is packaged in a 28-lead SSOP.
–2–
REV. A
AD9826–SPECIFICA TIONS
ANALOG SPECIFICATIONS
Parameter Min Typ Max Unit
MAXIMUM CONVERSION RATE
3-Channel Mode with CDS 30 MSPS
2-Channel Mode with CDS 30 MSPS
1-Channel Mode with CDS 18 MSPS
ACCURACY (ENTIRE SIGNAL PATH)
ADC Resolution 16 Bits
Integral Nonlinearity (INL) ±16 LSB
Differential Nonlinearity (DNL) ±0.5 LSB
No Missing Codes Guaranteed
ANALOG INPUTS
Input Signal Range (Programmable)
1
2.0/4.0 V p-p
Allowable Reset Transient
1
1.0 V
Input Limits
2
AVSS – 0.3 AVDD + 0.3 V
Input Capacitance 10 pF
Input Bias Current 10 nA
AMPLIFIERS
PGA Gain 1 6 V/V
PGA Gain Resolution
2
64 Steps
PGA Gain Monotonicity Guaranteed
Programmable Offset –300 +300 mV
Programmable Offset Resolution 512 Steps
Programmable Offset Monotonicity Guaranteed
NOISE AND CROSSTALK
Total Output Noise @ PGA Minimum 3.0 LSB rms
Total Output Noise @ PGA Maximum 9.0 LSB rms
Channel-to-Channel Crosstalk
@ 15 MSPS 70 dB
@ 6 MSPS 90 dB
POWER SUPPLY REJECTION
AVDD = 5 V 0.25 V 0.1 % FSR
DIFFERENTIAL VREF (at 25°C)
CAPT–CAPB 2.0 V
TEMPERATURE RANGE
Operating –40 +85 °C
Storage –65 +150 °C
POWER SUPPLIES
AVDD 4.75 5.0 5.25 V
DRVDD 3.0 5.0 5.25 V
OPERATING CURRENT
AVDD 75 mA
DRVDD 5 mA
Power-Down Mode 200 µA
POWER DISSIPATION
3-Channel Mode 400 mW
1-Channel Mode 300 mW
NOTES
1
Linear Input Signal Range is from 0 V to 4 V when the CCD’s reference level is clamped to 4 V by the AD9826’s input clamp.
4V SET BY INPUT CLAMP
(3V OPTION ALSO AVAILABLE)
1V TYP
RESET TRANSIENT
4V p-p MAX INPUT SIGNAL RANGE
GND
2
The PGA Gain is approximately “linear in dB” and follows the equation:
G
ain=
6.0
1+5.0
63 – G
63
where G is the register value.
Specifications subject to change without notice.
(T
MIN
to T
MAX
, AVDD = 5 V, DRVDD = 5 V, CDS Mode, f
ADCCLK
= 15 MHz, f
CDSCLK1
= f
CDSCLK2
= 5 MHz, PGA
Gain = 1, Input range = 4 V p-p, unless otherwise noted.)
–3–
REV. A
AD9826
DIGITAL SPECIFICATIONS
Parameter Symbol Min Typ Max Unit
LOGIC INPUTS
High Level Input Voltage V
IH
2.0 V
Low Level Input Voltage V
IL
0.8 V
High Level Input Current I
IH
10 µA
Low Level Input Current I
IL
10 µA
Input Capacitance C
IN
10 pF
LOGIC OUTPUTS
High Level Output Voltage V
OH
4.5 V
Low Level Output Voltage V
OL
0.1 V
High Level Output Current I
OH
50 µA
Low Level Output Current I
OL
50 µA
LOGIC OUTPUTS (with DRVDD = 3 V)
High Level Output Voltage, (I
OH
= 50 µA) V
OH
2.95 V
Low Level Output Voltage (IOL = 50 µA) V
OL
0.05 V
Specifications subject to change without notice.
TIMING SPECIFICATIONS
Parameter Symbol Min Typ Max Unit
CLOCK PARAMETERS
3-Channel Pixel Rate t
PRA
200 ns
1-Channel Pixel Rate t
PRB
80 ns
ADCCLK Pulsewidth t
ADCLK
30 ns
CDSCLK1 Pulsewidth t
C1
8ns
CDSCLK2 Pulsewidth t
C2
8ns
CDSCLK1 Falling to CDSCLK2 Rising t
C1C2
0ns
ADCCLK Falling to CDSCLK2 Rising t
ADC2
0ns
CDSCLK2 Rising to ADCCLK Rising t
C2ADR
5ns
CDSCLK2 Falling to ADCCLK Falling t
C2ADF
30 ns
CDSCLK2 Falling to CDSCLK1 Rising t
C2C1
5ns
Aperture Delay for CDS Clocks t
AD
2ns
SERIAL INTERFACE
Maximum SCLK Frequency f
SCLK
10 MHz
SLOAD to SCLK Set-Up Time t
LS
10 ns
SCLK to SLOAD Hold Time t
LH
10 ns
SDATA to SCLK Rising Set-Up Time t
DS
10 ns
SCLK Rising to SDATA Hold Time t
DH
10 ns
SCLK Falling to SDATA Valid t
RDV
10 ns
DATA OUTPUTS
Output Delay t
OD
6ns
3-State to Data Valid t
DV
10 ns
Output Enable High to 3-State t
HZ
10 ns
Latency (Pipeline Delay) 3 (Fixed) Cycles
NOTES
It is recommended that CDSCLK falling edges do not occur within the first 10 ns following an ADCCLK edge.
Specifications subject to change without notice.
(T
MIN
to T
MAX
, AVDD = 5 V, DRVDD = 5 V, CDS Mode, f
ADCCLK
= 15 MHz, f
CDSCLK1
= f
CDSCLK2
= 5 MHz,
CL = 10 pF, unless otherwise noted.)
(T
MIN
to T
MAX
, AVDD = 5 V, DRVDD = 5 V, specs are for 16-bit performance.)
AD9826
–4–
REV. A
ORDERING GUIDE
Temperature Package Package
Model Range Description Option
AD9826KRS –40°C to +85°C 5.3 mm SSOP RS-28
THERMAL CHARACTERISTICS
Thermal Resistance
28-Lead 5.3 mm SSOP
θ
JA
= 109°C/W
θ
JC
= 39°C/W
ABSOLUTE MAXIMUM RATINGS*
With
Respect
Parameter To Min Max Unit
VIN, CAPT, CAPB AVSS –0.3 AVDD + 0.3 V
Digital Inputs AVSS –0.3 AVDD + 0.3 V
AVDD AVSS –0.5 +6.5 V
DRVDD DRVSS –0.5 +6.5 V
AVSS DRVSS –0.3 +0.3 V
Digital Outputs DRVSS –0.3 DRVDD + 0.3 V
Junction Temperature 150 °C
Storage Temperature –65 +150 °C
Lead Temperature 300 °C
(10 sec)
*Stresses above those listed under Absolute Maximum Ratings may cause perma-
nent damage to the device. This is a stress rating only; 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.
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 AD9826 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.
WARNING!
ESD SENSITIVE DEVICE
AD9826
–5–
REV. A
PIN FUNCTION DESCRIPTIONS
Pin No. Mnemonic Type Description
1 CDSCLK1 DI CDS Reference Level Sampling Clock
2 CDSCLK2 DI CDS Data Level Sampling Clock
3 ADCCLK DI A/D Converter Sampling Clock
4 OEB DI Output Enable, Active Low
5 DRVDD P Digital Output Driver Supply
6 DRVSS P Digital Output Driver Ground
7 D7 DO Data Output MSB. ADC DB15 High Byte, ADC DB7 Low Byte
8 D6 DO Data Output. ADC DB14 High Byte, ADC DB6 Low Byte
9 D5 DO Data Output. ADC DB13 High Byte, ADC DB5 Low Byte
10 D4 DO Data Output. ADC DB12 High Byte, ADC DB4 Low Byte
11 D3 DO Data Output. ADC DB11 High Byte, ADC DB3 Low Byte
12 D2 DO Data Output. ADC DB10 High Byte, ADC DB2 Low Byte
13 D1 DO Data Output. ADC DB9 High Byte, ADC DB1 Low Byte
14 D0 DO Data Output LSB. ADC DB8 High Byte, ADC DB0 Low Byte
15 SDATA DI/DO Serial Interface Data Input/Output
16 SCLK DI Serial Interface Clock Input
17 SLOAD DI Serial Interface Load Pulse
18, 28 AVDD P 5 V Analog Supply
19, 27 AVSS P Analog Ground
20 CAPB AO ADC Bottom Reference Voltage Decoupling
21 CAPT AO ADC Top Reference Voltage Decoupling
22 VINB AI Analog Input, Blue Channel
23 CML AO Internal Bias Level Decoupling
24 VING AI Analog Input, Green Channel
25 OFFSET AO Clamp Bias Level Decoupling
26 VINR AI Analog Input, Red Channel
TYPE: AI = Analog Input, AO = Analog Output, DI = Digital Input, DO = Digital Output, P = Power.
PIN CONFIGURATION
TOP VIEW
(Not to Scale)
28
27
26
25
24
23
22
21
20
19
18
17
16
15
1
2
3
4
5
6
7
8
9
10
11
12
13
14
AD9826
CDSCLK1
AVD D
CDSCLK2
AVSS
VINR
ADCCLK
OFFSET
OEB
VING
DRVDD
DRVSS
CML
(MSB) D7
VINB
D6
CAPT
D5
CAPB
D4
AVSS
D3
AVD D
SLOAD
D2
SCLK
D1
(LSB) D0
SDATA
AD9826
–6–
REV. A
DEFINITIONS OF SPECIFICATIONS
INTEGRAL NONLINEARITY (INL)
Integral nonlinearity error refers to the deviation of each individual
code from a line drawn from “zero scale” through “positive full
scale.” The point used as “zero scale” occurs 1/2 LSB before the
first code transition. “Positive full scale” is defined as a level
1 1/ 2 LSB beyond the last code transition. The deviation is
measured from the middle of each particular code to the true
straight line.
DIFFERENTIAL NONLINEARITY (DNL)
An ideal ADC exhibits code transitions that are exactly 1 LSB
apart. DNL is the deviation from this ideal value. Thus every
code must have a finite width. No missing codes guaranteed
to 16-bit resolution indicates that all 65536 codes, respectively, must be present over all operating ranges.
OFFSET ERROR
The first ADC code transition should occur at a level 1/2 LSB
above the nominal zero scale voltage. The offset error is the
deviation of the actual first code transition level from the
ideal level.
GAIN ERROR
The last code transition should occur for an analog value
1 1/2 LSB below the nominal full scale voltage. Gain error is
the deviation o f the actual difference between first and last
code transitions and the ideal difference between the first and
last code transitions.
INPUT REFERRED NOISE
The rms output noise is measured using histogram techniques.
The ADC output codes’ standard deviation is calculated in
LSB, and can be converted to an equivalent voltage, using the
relationship 1 LSB = 4 V/65536 = 61 µV. The noise may then
be referred to the input of the AD9826 by dividing by the
PGA gain.
CHANNEL-TO-CHANNEL CROSSTALK
In an ideal 3-channel system, the signal in one channel will not
influence the signal level of another channel. The channel-tochannel crosstalk specification is a measure of the change that
occurs in one channel as the other two channels are varied. In
the AD9826, one channel is grounded and the other two channels are exercised with full scale input signals. The change in the
output codes from the first channel is measured and compared
with the result when all three channels are grounded. The difference is the channel-to-channel crosstalk, stated in LSB.
APERTURE DELAY
The aperture delay is the time delay that occurs from when a
sampling edge is applied to the AD9826 until the actual sample
of the input signal is held. Both CDSCLK1 and CDSCLK2
sample the input signal during the transition from high to low,
so the aperture delay is measured from each clock’s falling edge
to the instant the actual internal sample is taken.
POWER SUPPLY REJECTION
Power supply rejection specifies the maximum full-scale change
that occurs from the initial value when the supplies are varied
over the specified limits.
AD9826
–7–
REV. A
Typical Performance Characteristics–
–20
0
12000
0
–10
10
20
24000 36000 48000 64000
TPC 1. Typical INL Performance at 15 MSPS
–1.0
0 12000
0
–0.5
0.5
1.0
24000 36000 48000
64000
TPC 2. Typical DNL Performance at 15 MSPS
GAIN SETTING
0
0
15
NOISE – LSB RMS
5
10
30 45 63
TPC 3. Output Noise vs. Gain
–1.0
0
200
0
–0.5
0.5
1.0
400 600 800 1000
TPC 4. Typical INL Performance at 30 MSPS
–1.0
0
200
0
–0.5
0.5
1.0
400 600 800 1000
TPC 5. Typical DNL Performance at 30 MSPS
GAIN SETTING
0
0
15
NOISE – LSB RMS
5
10
30 45 63
TPC 6. Input Referred Noise vs. Gain
AD9826
–8–
REV. A
TIMING DIAGRAMS
ANALOG
INPUTS
CDSCLK1
CDSCLK2
ADCCLK
OUTPUT
DATA
D<7:0>
PIXEL n (R,G,B)
PIXEL
(n+1)
PIXEL
(n+2)
t
AD
t
C1
t
AD
t
C2C1
t
C2
t
C2ADF
t
C2ADR
t
ADC2
t
OD
t
ADCLK
t
ADCLK
HIGH
BYTE
LOW
BYTE
HB LB HB LB HB LB HB LB HB HBLB LB
G(n)G(n)R(n)R(n)B(n–1)B(n–1)G(n–1)G(n–1)R(n–1)R(n–1)B(n–2)B(n–2)G(n–2)G(n–2)R(n–2)
t
PRA
t
C1C2
Figure 1. 3-Channel CDS Mode Timing
It is recommended that CDSCLK falling edges do not occur within the first 10 ns following an ADCCLK edge.
ANALOG
INPUTS
CDSCLK1
CDSCLK2
ADCCLK
OUTPUT
DATA
D<7:0>
PIXEL n PIXEL
(n+1)
PIXEL
(n+2)
t
AD
t
C1
t
AD
t
C2C1
t
C2ADR
t
OD
HIGH BYTE LOW BYTE
t
C1C2
LOW BYTE LOW BYTEHIGH BYTE HIGH BYTE
t
PRB
PIXEL (n–4) PIXEL (n–4) PIXEL (n–3) PIXEL (n–3) PIXEL (n–2) PIXEL (n–2)
t
C2ADF
t
ADCLK
t
ADCLK
t
C2
NOTE
IN 1-CHANNEL CDS MODE, THE CDSCLK1 FALLING EDGE AND THE CDSCLK2 RISING EDGE MUST OCCUR WHILE ADCCLK IS “LOW.”
Figure 2. 1-Channel CDS Mode Timing
AD9826
–9–
REV. A
t
AD
t
AD
t
C2ADR
ANALOG
INPUTS
CDSCLK1
CDSCLK2
ADCCLK
OUTPUT
DATA
D<7:0>
PIXEL n
PIXEL (n+1)
PIXEL (n+2)
t
C1
t
C2C1
t
C2
t
C2ADF
t
ADC2
t
ADCLK
t
ADCLK
HIGH
BYTE
LOW
BYTE
CH1(n–2)
t
C1C2
CH2(n–2) CH1(n–1) CH2(n–1)
LOW
BYTE
LOW
BYTE
LOW
BYTE
LOW
BYTE
HIGH
BYTE
HIGH
BYTE
HIGH
BYTE
HIGH
BYTE
t
PRA
CH1(n)
Figure 3. 2-Channel CDS Mode Timing
t
AD
t
ADCLK
t
ADCLK
t
C2ADR
ANALOG
INPUTS
CDSCLK2
ADCCLK
OUTPUT
DATA
D<7:0>
PIXEL n
PIXEL
(n+1)
t
C2
t
C2ADF
t
ADC2
HIGH
BYTE
LOW
BYTE
LOW
BYTE
LOW
BYTE
LOW
BYTE
LOW
BYTE
HIGH
BYTE
HIGH
BYTE
HIGH
BYTE
HIGH
BYTE
CH1(n–2)
CH2(n–2)
CH1(n–1) CH2(n–1)
CH1(n)
Figure 4. 2-Channel SHA Mode Timing
AD9826
–10–
REV. A
t
AD
PIXEL n (R,G,B)
t
C2
t
ADCLK
t
ADCLK
t
C2ADR
ANALOG
INPUTS
CDSCLK2
ADCCLK
OUTPUT
DATA
D<7:0>
R (n–1)
t
C2AD
t
ADC2
HIGH
BYTE
LOW
BYTE
t
OD
t
PRA
HB LB HB HB HB HB HBLB LB
LB
LB LB
R (n–2)
G (n–2)
G (n–2)
B (n–2) B (n–2) R (n–1)
G (n–1)
G (n–1) B (n–1)
B (n–1) R (n) R (n) G (n)
G (n)
PIXEL (n+1)
Figure 5. 3-Channel SHA Mode Timing
HIGH BYTE LOW BYTE LOW BYTE LOW BYTEHIGH BYTE HIGH BYTE
ANALOG
INPUTS
CDSCLK2
ADCCLK
OUTPUT
DATA
D<7:0>
PIXEL n
t
AD
t
C2ADR
t
OD
t
PRB
PIXEL (n–4) PIXEL (n–4) PIXEL (n–3) PIXEL (n–3) PIXEL (n–2) PIXEL (n–2)
t
C2ADF
t
ADCLK
t
C2
t
ADCLK
t
ADCLK
NOTE
IN 1-CHANNEL SHA MODE, THE CDSCLK2 RISING EDGE MUST OCCUR WHILE ADCCLK IS “LOW.”
Figure 6. 1-Channel SHA Mode Timing
AD9826
–11–
REV. A
ADCCLK
OEB
OUTPUT
DAT A
<D7:D0>
HIGH BYTE
DB15–DB8
LOW BYTE
DB7–DB0
HB
n+1
LB
n+1
LB
n+2
HB
n+3
t
DV
t
HZ
t
OD
PIXEL n PIXEL n
t
OD
Figure 7. Digital Output Data Timing
ADCCLK
OEB
OUTPUT
DAT A
<D7:D0>
HIGH BYTE
DB15–DB8
HB
n+2
HB
n+3
t
DV
t
HZ
t
OD
PIXEL n
PIXEL n+1
HIGH BYTE
DB15–DB8
Figure 8. Single Byte Mode Digital Output Data Timing
t
LH
D8
D7
D6
D5
D4
D3 D2
D1
D0
t
DS
t
LS
t
DH
A0A2
R/Wb
SDATA
A1
SCLK
SLOAD
Figure 9. Serial Write Operation Timing
t
LH
D8
D7
D6
D5
D4
D3 D2
D1
D0
t
RDV
t
LS
A0A2 A1
R/Wb
SDATA
SCLK
SLOAD
Figure 10. Serial Read Operation Timing
AD9826
–12–
REV. A
ANALOG
INPUTS
CDSCLK1
CDSCLK2
RED
PGA
OUT
OUTPUT
DATA
D<7:0>
PIXEL n (R,G,B)
PIXEL (n+1)
HIGH
BYTE
LOW
BYTE
ADCCLK
GREEN
PGA
OUT
BLUE
PGA
OUT
MUX
OUT
HB LB LB LBLB LB LBHB HB HB HB HB
R(n–2) G(n–2) G(n–2) B(n–2)
B(n–2) R(n–1)
R(n–1)
G(n–1) G(n–1)
B(n–1) B(n–1) R(n)
R(n)
G(n) G(n)
BLUE (n–1)GREEN (n–1) GREEN (n)
BLUE (n) GREEN (n+1)
RED (n+1)
GREEN (n–1)
BLUE (n–1)
RED (n–1)
RED (n)
GREEN (n)
BLUE (n)
RED (n+1)
GREEN (n+1)
BLUE (n+1)
RED (n)
NOTES
1. THE MUX STATE MACHINE IS INTERNALLY RESET AT THE CDSCLK2 RISING EDGE.
2. EACH PIXEL IS SAMPLED AND AMPLIFIED BY THE PGAs AT CDSCLK2 FALLING EDGE.
3. AFTER CDSCLK2 RISING EDGE, THE NEXT ADCCLK RISING EDGE WILL ALWAYS SELECT RED PGA OUTPUT.
4. THE ADC SAMPLES THE MUX OUTPUT ON ADCCLK FALLING EDGES.
5. THE MUX SWITCHES TO THE NEXT PGA OUTPUT AT ADCCLK RISING EDGES.
Figure 11. Internal Timing Diagram for 3-Channel CDS Mode
AD9826
–13–
REV. A
FUNCTIONAL DESCRIPTION
The AD9826 can be operated in six different modes: 3-Channel
CDS Mode, 3-Channel SHA Mode, 2-Channel CDS Mode,
2-Channel SHA Mode, 1-Channel CDS Mode, and 1-Channel
SHA Mode. Each mode is selected by programming the Configuration Registers through the serial interface. For more detail on
CDS or SHA mode operation, see the Circuit Operation section.
3-Channel CDS Mode
In 3-Channel CDS Mode, the AD9826 simultaneously samples
the Red, Green, and Blue input voltages from the CCD outputs.
The sampling points for each Correlated Double Sampler (CDS)
are controlled by CDSCLK1 and CDSCLK2 (see Figures 11
and 13). CDSCLK1’s falling edge samples the reference level of
the CCD waveform. CDSCLK2’s falling edge samples the data
level of the CCD waveform. Each CDS amplifier outputs the
difference between the CCD’s reference and data levels. Next,
the output voltage of each CDS amplifier is level-shifted by an
Offset DAC. The voltages are then scaled by the three Programmable Gain Amplifiers before being multiplexed through the
16-Bit ADC. The ADC sequentially samples the PGA outputs
on the falling edges of ADCCLK.
The offset and gain values for the Red, Green, and Blue channels are programmed using the serial interface. The order in
which the channels are switched through the multiplexer is
selected by programming the MUX Configuration register.
Timing for this mode is shown in Figure 1. It is recommended
that the falling edge of CDSCLK2 occur before the rising edge
of ADCCLK, although this is not required to satisfy the minimum timing constraints. The rising edge of CDSCLK2 should
not occur before the previous falling edge of ADCCLK, as
shown by t
ADC2
. The output data latency is three clock cycles.
3-Channel SHA Mode
In 3-Channel SHA Mode, the AD9826 simultaneously samples
the Red, Green, and Blue input voltages. The sampling point is
controlled by CDSCLK2. CDSCLK2’s falling edge samples the
input waveforms on each channel. The output voltages from the
three SHAs are modified by the offset DACs and then scaled by
the three PGAs. The outputs of the PGAs are then multiplexed
through the 16-bit ADC. The ADC sequentially samples the
PGA outputs on the falling edges of ADCCLK.
The input signal is sampled with respect to the voltage applied
to the OFFSET pin (see Figure 14). With the OFFSET pin
grounded, a zero volt input corresponds to the ADC’s zero scale
output. The OFFSET pin may also be used as a coarse offset
adjust pin. A voltage applied to this pin will be subtracted from
the voltages applied to the Red, Green, and Blue inputs in the first
amplifier stage of the AD9826. The input clamp is disabled in this
mode. For more information, see the Circuit Operation section.
Timing for this mode is shown in Figure 5. CDSCLK1 should
be grounded in this mode. Although it is not required, it is recommended that the falling edge of CDSCLK2 occur before the
rising edge of ADCCLK. The rising edge of CDSCLK2 should
not occur before the previous falling edge of ADCCLK, as shown
by t
ADC2
. The output data latency is three ADCCLK cycles.
The offset and gain values for the Red, Green, and Blue channels are programmed using the serial interface. The order in
which the channels are switched through the multiplexer is
selected by programming the MUX Configuration register.
2-Channel CDS Mode
The 2-Channel Mode is selected by writing a “1” into two of the
channel select bits of the MUX register (D4–D6). Bit D5 of the
configuration register also needs to be set low to take the part out
of 3-Channel Mode. The channels that will be used is determined
by the contents of Bits D4–D6 of the MUX Configuration Register (see Table III). The combination of inputs that can be
selected are; RG, RB, or GB by writing a “1” into the appropriate bit. The sample order is selected by Bit D7. If D7 is high,
the MUX will sample in the following order: RG or RB or GB
depending on which channels are turned on. If Bit D7 is set low
the mux will sample in the following order: GR or BR or BG
depending on which channels are turned on.
The AD9826 simultaneously samples the selected channels’
input voltages from the CCD outputs. The sampling points
for each Correlated Double Sampler (CDS) are controlled by
CDSCLK1 and CDSCLK2 (see Figure 11). CDSCLK1’s falling edge samples the reference level of the CCD waveform.
CDSCLK2’s falling edge samples the data level of the CCD
waveform. Each CDS amplifier outputs the difference between
the CCD’s reference and data levels. Next, the output voltage of
each CDS amplifier is level-shifted by an Offset DAC. The voltages are then scaled by the two Programmable Gain Amplifiers
before being multiplexed through the 16-bit ADC. The ADC
sequentially samples the PGA outputs on the falling edges of
ADCCLK.
The offset and gain values for the Red, Green, and Blue channels are programmed using the serial interface. The order in
which the channels are switched through the multiplexer is
selected by programming the MUX Configuration Register.
Timing for this mode is shown in Figure 3. The rising edge of
CDSCLK2 should not occur before the previous falling edge of
ADCCLK, as shown by t
ADC2
. The output data latency is three
clock cycles.
2-Channel SHA Mode
The 2-Channel Mode is selected by writing a “1” into two of the
channel select bits of the MUX Register (D4–D6). Bit D5 of the
configuration register also needs to be set low to take the part
out of 3-Channel Mode. The channels that will be used is determined by the contents of Bits D4–D6 of the MUX Configuration
Register (see Table III ). The combination of inputs that can be
selected are; RG, RB, or GB by writing a “1” into the appropriate bit. The sample order is selected by Bit D7. If D7 is high,
the mux will sample in the following order: RG or RB or GB,
depending on which channels are turned on. If Bit D7 is set low,
the mux will sample in the following order: GR or BR or BG,
depending on which channels are turned on.
In 2-Channel SHA Mode, the AD9826 simultaneously samples
the selected channels’ input voltages. The sampling point is
controlled by CDSCLK2. CDSCLK2’s falling edge samples the
input waveforms on each channel. The output voltages from the
two SHAs are modified by the offset DACs and then scaled by
the two PGAs. The outputs of the PGAs are then multiplexed
through the 16-bit ADC. The ADC sequentially samples the PGA
outputs on the falling edges of ADCCLK.
The input signal is sampled with respect to the voltage applied
to the OFFSET pin (see Figure 14). With the OFFSET pin
grounded, a zero volt input corresponds to the ADC’s zero scale
output. The OFFSET pin may also be used as a coarse offset
AD9826
–14–
REV. A
adjust pin. A voltage applied to this pin will be subtracted from
the voltages applied to the Red, Green, and Blue inputs in the first
amplifier stage of the AD9826. The input clamp is disabled in this
mode. For more information, see the Circuit Operation section.
Timing for this mode is shown in Figure 4. CDSCLK1 should
be grounded in this mode. The rising edge of CDSCLK2 should
not occur before the previous falling edge of ADCCLK, as shown
by t
ADC2
. The output data latency is three ADCCLK cycles. The
offset and gain values for the Red, Green, and Blue channels are
programmed using the serial interface. The order in which the
channels are switched through the multiplexer is selected by
programming the MUX Configuration Register.
1-Channel CDS Mode
This mode operates the same way as the 3-Channel CDS mode.
The difference is that the multiplexer remains fixed in this mode,
so only the channel specified in the MUX Configuration Register is processed.
Timing for this mode is shown in Figure 2.
1-Channel SHA Mode
This mode operates the same way as 3-Channel SHA mode,
except that the multiplexer remains stationary. Only the channel
specified in the MUX Configuration Register is processed.
Timing for this mode is shown in Figure 6. CDSCLK1 should
be grounded in this mode of operation.
Configuration Register
The Configuration Register controls the AD9826’s operating
mode and bias levels. Bits D8 and D1 should always be set low.
Table II. Configuration Register Settings
D
8D7 D6 D5 D4 D3 D2 D1 D0
Set Input Range Internal VREF 3CH Mode CDS Operation Input Clamp Bias Power-Down Set Output Mode
to
1 = 4 V* 1 = Enabled* 1 = On* 1 = CDS Mode* 1 = 4 V* 1 = On
to
0 = 2 Byte*
0
0 = 2 V 0 = Disabled 0 = Off 0 = SHA Mode 0 = 3 V 0 = Off (Normal)*
0
1 = 1 Byte
*Power-on default value.
Bit D7 controls the input range of the AD9826. Setting D7 high
sets the input range to 4 V while setting Bit D7 low sets the
input range to 2 V. Bit D6 controls the internal voltage reference. If the AD9826’s internal voltage reference is used, then
this bit is set high. Setting Bit D6 low will disable the internal
voltage reference, allowing an external voltage reference to be
used. Setting Bit D5 high will configure the AD9826 for 3channel operation. If D5 is set low, the part will be in either
2CH or 1CH mode based on the settings in the MUX Configuration Register (See Table III and the MUX Configuration
Register description). Setting Bit D4 high will enable the CDS
mode of operation, and setting this bit low will enable the SHA
mode of operation. Bit D3 sets the dc bias level of the AD9826’s
input clamp.
This bit should always be set high for the 4 V clamp bias, unless
a CCD with a reset feedthrough transient exceeding 2 V is used.
If the 3 V clamp bias level is used, then the peak-to-peak i nput
signal range to the AD9826 is reduced to 3 V maximum. Bit D2
controls the power-down mode. Setting Bit D2 high will place
the AD9826 into a very low-power “sleep” mode. All register
contents are retained while the AD9826 is in the powered-down
state. Bit D0 controls the output mode of the AD9826. Setting
Bit D0 high will enable a single byte output mode where only
the 8 MSBs of the 16 b ADC will be output on each rising edge
of ADCCLK (see Figure 8). If Bit D0 is set low, then the 16b
ADC output is multiplexed into two bytes. The MSByte is
output on ADCCLK rising edge and the LSByte is output on
ADCCLK falling edge.
Table I. Internal Register Map
Register Address Data Bits
Name A2 A1 A0 D8 D7 D6 D5 D4 D3 D2 D1 D0
Configuration 0 0 0 0 Input Rng VREF 3CH Mode CDS On Clamp Pwr Dn 0 1 Byte Out
MUX Config 0 0 1 0 RGB/BGR Red Green Blue 0 0 0 0
Red PGA 0 1 0 0 0 0 MSB LSB
Green PGA 0 1 1 0 0 0 MSB LSB
Blue PGA 1 0 0 0 0 0 MSB LSB
Red Offset 1 0 1 MSB LSB
Green Offset 1 1 0 MSB LSB
Blue Offset 1 1 1 MSB LSB
AD9826
–15–
REV. A
MUX Configuration Register
The MUX Configuration Register controls the sampling channel order and the 2-Channel Mode configuration in the AD9826.
Bits D8 and D3–D0 should always be set low. Bit D7 is used
when operating in 3-Channel or 2-Channel Mode. Setting Bit
D7 high will sequence the MUX to sample the Red channel
first, then the Green channel, and then the Blue channel. When
in 3-channel mode, the CDSCLK2 pulse always resets the MUX
to sample the Red channel first (see Figure 11). When Bit D7 is
set low, the channel order is reversed to Blue first, Green second, and Red third. The CDSCLK2 pulse will always reset the
MUX to sample the Blue channel first. Bits D6, D5, and D4 are
used when operating in 1 or 2-Channel Mode. Bit D6 is set high
to sample the Red channel. Bit D5 is set high to sample the
Green channel. Bit D4 is set high to sample the Blue channel.
The MUX will remain stationary during 1-channel mode. TwoChannel Mode is selected by setting two of the channel select
Bits (D4–D6) high. The MUX samples the channels in the
order selected by Bit D7.
PGA Gain Registers
There are three PGA registers for individually programming the
gain in the Red, Green, and Blue channels. Bits D8, D7, and
D6 in each register must be set low, and Bits D5 through D0
control the gain range from 1× to 6× in 64 increments. See
Figure 17 for a graph of the PGA gain versus PGA register
code. The coding for the PGA registers is straight binary, with
an all “zeros” word corresponding to the minimum gain setting
(1×) and an all “ones” word corresponding to the maximum
gain setting (6×).
Offset Registers
There are three Offset Registers for individually programming
the offset in the Red, Green, and Blue channels. Bits D8 through
D0 control the offset range from –300 mV to +300 mV in 512
increments. The coding for the Offset Registers is Sign Magnitude, with D8 as the sign bit. Table V shows the offset range
as a function of the Bits D8 through D0.
Table III. MUX Configuration Register Settings
D
8 D7 D6 D5 D4 D3 D2 D1 D0
Set MUX Order Channel Select Channel Select Channel Select Set Set Set Set
to
1 = R-G-B* 1 = RED* 1 = GREEN 1 = BLUE
to to to to
0
0 = B-G-R 0 = Off 0 = Off* 0 = Off*
00 00
*Power-on default value.
Table IV. PGA Gain Register Settings
D8 D7 D6 D5 D4 D3 D2 D1 D0 Gain (V/V) Gain (dB)
Set to 0 Set to 0 Set to 0 MSB LSB
000000000* 1.0 0.0
000000001 1.013 0.12
• • •
• • •
• • •
000111111 5.56 14.9
000111111 6.0 15.56
*Power-on default value.
Table V. Offset Register Settings
D8 D7 D6 D5 D4 D3 D2 D1 D0 Offset (mV)
MSB LSB
000000000* 0
000000001 +1.2
• •
• •
• •
011111111 +300
100000000 0
100000001 –1.2
• •
• •
• •
111111111 –300
*Power-on default value.
AD9826
–16–
REV. A
CIRCUIT OPERATION
Analog Inputs—CDS Mode Operation
Figure 12 shows the analog input configuration for the CDS
mode of operation. Figure 13 shows the internal timing for the
sampling switches. The CCD reference level is sampled when
CDSCLK1 transitions from high to low, opening S1. The CCD
data level is sampled when CDSCLK2 transitions from high to
low, opening S2. S3 is then closed, generating a differential
output voltage representing the difference between the two
sampled levels.
The input clamp is controlled by CDSCLK1. When CDSCLK1
is high, S4 closes and the internal bias voltage is connected to
the analog input. The bias voltage charges the external 0.1 µF
input capacitor, level-shifting the CCD signal into the AD9826’s
input common-mode range. The time constant of the input
clamp is determined by the internal 5 kΩ resistance and the
external 0.1 µF input capacitance.
CCD
SIGNAL
VINR
AD9826
0.1F
OFFSET
0.1F
1F
+
S1
4pF
S3
S2
5K
S4
4V
1.7k⍀
3V
2.2k⍀
4pF
6.9k⍀
CML
INPUT CLAMP LEVEL
IS SELECTED IN THE
CONFIGURATION
REGISTER
CML
Figure 12. CDS-Mode Input Configuration (All Three
Channels Are Identical)
External Input Coupling Capacitors
The recommended value for the input coupling capacitors is
0.1 µF. While it is possible to use a smaller capacitor, this larger
value is chosen for several reasons:
Crosstalk
The input coupling capacitor creates a capacitive divider with
any parasitic capacitance between PCB traces and on chip traces.
C
IN
should be large relative to these parasitic capacitances in
order to minimize this effect. For example, with a 100 pF input
capacitance and just a few hundred f F of parasitic capacitance
on the PCB and/or the IC the imaging system could expect
to have hundreds of LSBs of crosstalk at the 16 b level. Using
a large capacitor value = 0.1 µF will minimize any errors due
to crosstalk.
Signal Attenuation
The input coupling capacitor creates a capacitive divider with a
CMOS integrated circuit’s input capacitance, attenuating the
CCD signal level. C
IN
should be large relative to the IC’s 10 pF
input capacitance in order to minimize this effect.
Linearity
Some of the input capacitance of a CMOS IC is junction capacitance, which varies nonlinearly with applied voltage. If the input
coupling capacitor is too small, then the attenuation of the CCD
signal will vary nonlinearly with signal level. This will degrade
the system linearity performance.
Sampling Errors
The internal 4 pF sample capacitors have a “memory” of the
previously sampled pixel. There is a charge redistribution error
between C
IN
and the internal sample capacitors for larger pixel-
to-pixel voltage swings. As the value of C
IN
is reduced, the
resulting error in the sampled voltage will increase. With a C
IN
value of 0.1 µF, the charge redistribution error will be less than
1 LSB for a full-scale pixel-to-pixel voltage swing.
CDSCLK1
CDSCLK2
Q3
(INTERNAL)
S3 OPEN
S2 OPEN
S1, S4 OPEN
S1, S4 CLOSED
S2 CLOSED
S3 CLOSED
S1, S4 CLOSED
S2 CLOSED
S3 CLOSED
Figure 13. CDS-Mode Internal Switch Timing
AD9826
–17–
REV. A
Analog Inputs—
SHA Mode Operation
Figure 14 shows the analog input configuration for the SHA
mode of operation. Figure 15 shows the internal timing for the
sampling switches. The input signal is sampled when CDSCLK2
transitions from high to low, opening S1. The voltage on the
OFFSET pin is also sampled on the falling edge of CDSCLK2,
when S2 opens. S3 is then closed, generating a differential output voltage representing the difference between the sampled
input voltage and the OFFSET voltage. The input clamp is
disabled during SHA mode operation.
INPUT
SIGNAL
CML
CML
VINR
AD9826
OFFSET
S1
4pF
S3
S2
4pF
OPTIONAL DC
OFFSET (OR
CONNECT
TO GND)
VING
VINB
Figure 14. SHA-Mode Input Configuration (All Three
Channels Are Identical)
CDSCLK2
Q3
(INTERNAL)
S3 OPEN
S1, S2 OPEN
S1, S2 CLOSED S1, S2 CLOSED
S3 CLOSED
S3 CLOSED
Figure 15. SHA-Mode Internal Switch Timing
Figure 16 shows how the OFFSET pin may be used in a CIS
application for coarse offset adjustment. Many CIS signals have
dc offsets ranging from several hundred millivolts to more than
1 V. By connecting the appropriate dc voltage to the OFFSET
pin, the CIS signal will be restored to “zero.” After the large dc
offset is removed, the signal can be scaled using the PGA to
maximize the ADC’s dynamic range.
SHA
SHA
SHA
VINR
VING
VINB
OFFSET
RED
GREEN
BLUE
VRED FROM
CIS MODULE
AVDD
R1
R2
DC OFFSET
REDOFFSET
GREENOFFSET
BLUEOFFSET
AD9826
0.1F
Figure 16. SHA-Mode Used with External DC Offset
AD9826
–18–
REV. A
TOP VIEW
(Not to Scale)
28
27
26
25
24
23
22
21
20
19
18
17
16
15
1
2
3
4
5
6
7
8
9
10
11
12
13
14
AD9826
CDSCLK1
CDSCLK2
ADCCLK
OEB
DRVDD
DRVSS
(MSB) D7
D6
D5
D4
D3
D2
D1
(LSB)D0
AVD D
AVSS
VINR
OFFSET
VING
CML
VINB
CAPT
CAPB
AVSS
AVD D
SLOAD
SCLK
SDATA
0.1F
0.1F
0.1F
0.1F
10F
0.1F
0.1F
0.1F
RED INPUT
GREEN INPUT
BLUE INPUT
CLOCK
INPUTS
0.1F
DAT A
INPUTS
5V/3V
5V
0.1F
1.0F
0.1F
0.1F
5V
SERIAL
INTERFACE
Figure 18. Recommended Circuit Configuration, 3-Channel CDS Mode
Programmable Gain Amplifiers
The AD9826 uses one Programmable Gain Amplifier (PGA) for
each channel. Each PGA has a gain range from 1× (0 dB) to
6.0× (15.56 dB), adjustable in 64 steps. Figure 17 shows the
PGA gain as a function of the PGA register code. Although the
gain curve is approximately “linear in dB,” the gain in V/V varies nonlinearly with register code, following the equation:
GainG=
+
60
150
63
63
.
.
–
where G is the decimal value of the gain register contents, and
varies from 0 to 63.
PGA REGISTER VALUE – Decimal
0
0
12
GAIN – dB
4
12
24 36 63
GAIN – V/V
GAIN – dB
GAIN – V/V
48 60
8
16
1.00
2.25
4.75
3.50
6.00
Figure 17. PGA Gain Transfer Function
APPLICATIONS INFORMATION
Circuit and Layout Recommendations
The recommended circuit configuration for 3-Channel CDS
Mode operation is shown in Figure 18. The recommended
input coupling capacitor value is 0.1 µF (see Circuit Operation
section for more details). A single ground plane is recommended
for the AD9826. A separate power supply may be used for
DRVDD, the digital driver supply, but this supply pin should
still be decoupled to the same ground plane as the rest of the
AD9826. The loading of the digital outputs should be minimized, either by using short traces to the digital ASIC, or by
using external digital buffers. To minimize the effect of digital
transients during major output code transitions, the falling edge
of CDSCLK2 should occur coincident with or before the
rising edge of ADCCLK (see Figures 1 through 6 for timing).
All 0.1 µF decoupling capacitors should be located as close as
possible to the AD9826 pins. When operating in 1CH or 2CH
Mode, the unused analog inputs should be grounded.
For 3-Channel SHA Mode, all of the above considerations also
apply, except that the analog input signals are directly connected
to the AD9826 without the use of coupling capacitors. The analog
input signals must already be dc-biased between 0 V and 4 V.
Also, the OFFSET pin should be grounded if the inputs to the
AD9826 are to be referenced to ground, or a dc offset voltage
should be applied to the OFFSET pin in the case where a coarse
offset needs to be removed from the inputs. (See Figure 16 and
the Circuit Operation section for more details.)
AD9826
–19–
REV. A
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
28-Lead 5.3 mm SSOP
(RS-28)
0.009 (0.229)
0.005 (0.127)
0.03 (0.762)
0.022 (0.558)
8°
0°
0.008 (0.203)
0.002 (0.050)
0.07 (1.79)
0.066 (1.67)
0.078 (1.98)
0.068 (1.73)
0.015 (0.38)
0.010 (0.25)
SEATING
PLANE
0.0256
(0.65)
BSC
0.311 (7.9)
0.301 (7.64)
0.212 (5.38)
0.205 (5.21)
28 15
141
0.407 (10.34)
0.397 (10.08)
PIN 1
Revision History
Location Page
Data Sheet changed from REV. 0 to REV. A.
Edits to ORDERING GUIDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Edits to Figure 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Edits to Figure 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
–20–
C02367–0–10/01(A)
PRINTED IN U.S.A.
Loading...