CIRRUS CS7622

CS7622
CCD Imager Analog Processor
Features
Description
l 13-Bit
The CS7622 is a low-power analog front-end processor
for interline or frame transfer CCD imagers. Main applications include digital still image cameras and video
cameras.
A/D Conversion Using DRX™
Technology
l Backlight Compensation
l Supports Full Scale Analog Input Voltage
Ranges from 300 mV to 1 V in 100 mV
Increments
l High Resolution Output Mode
l Low Resolution (Preview) Output Mode for
LCD Driver
l Integrated Correlated Double Sampler
l Digital Black Level Clamp
l Digital Outputs Selectable for 13, 12, or 10 Bits
l Low Power Consumption
l Power Down Mode
l High Speed Serial Interface
l Supports a Large Variety of Clock Input
Frequencies
l Low power mode option
CCD OUTPUT
CDS/DRX
GAIN
The architecture includes a correlated double sampler,
black level clamp and a 13-bit A/D conversion module
using patented DRX technology.
Chip parameters can be programmed using a high
speed 4-wire asynchronous digital interface.
The chip outputs digitized CCD data in either 13-bit, 12bit or 10-bit format. 10-bit outputs are generated from the
13-bit A/D output by a programmable companding curve.
ORDERING INFORMATION
CS7622-IQ -40 to +85 °C 32-pin TQFP 7x7x1.4m
A/D
CONVERTER
OUTPUT
COMPANDER
CLOCK
BLACK
LEVEL
DATA OUT
CLOCK OUT
REGISTER
BLOCK
CK_FT
CLOCK
GENERATOR
CK_DATA
SERIAL
INTERFACE
SERIAL BUS
Preliminary Product Information
P.O. Box 17847, Austin, Texas 78760
(512) 445 7222 FAX: (512) 445 7581
http://www.cirrus.com
This document contains information for a new product.
Cirrus Logic reserves the right to modify this product without notice.
Copyright  Cirrus Logic, Inc. 1999
(All Rights Reserved)
JUL ‘99
DS322PP1
1
CS7622
TABLE OF CONTENTS
1.0 CHARACTERISTICS/SPECIFICATIONS ........................................................... 4
1.1 DIGITAL CHARACTERISTICS.................................................................... 4
1.2 POWER CONSUMPTION ........................................................................... 4
1.3 RECOMMENDED OPERATING CHARACTERISTICS............................... 4
1.4 ABSOLUTE MAXIMUM RATINGS .............................................................. 4
1.5 ADC (ANALOG-TO-DIGITAL CONVERTER).............................................. 5
1.6 CDS/VGA PARAMETERS........................................................................... 5
1.7 SERIAL INTERFACE TIMING SPECIFICATIONS ...................................... 5
2.0 GENERAL DESCRIPTION .................................................................................. 7
3.0 OPERATION ........................................................................................................ 8
3.1 CDS/VGA (correlated double sampling/variable gain amplification) ........... 8
3.2 Black Level Adjustment ............................................................................ 10
3.3 Gain Adjust Block ..................................................................................... 11
3.4 13-to-10 Bit Compander ........................................................................... 12
3.5 Stand By and Preview Mode ................................................................... 14
3.6 Serial Interface .......................................................................................... 14
3.7 Input Timing for Sampling Clocks ............................................................. 15
4.0 REGISTER DESCRIPTIONS ............................................................................. 17
Reset ........................................................................................................ 18
Power down Control 1 .............................................................................. 18
Operation Control 1 .................................................................................. 19
Operation Control 2 .................................................................................. 20
Black Level Control (8 LSBs) .................................................................... 20
Black Level Control (MSB) ........................................................................ 21
Black Level Control - General ................................................................... 21
Black Level Control - Loop Gain, Clamp Length ....................................... 22
Gain Calibration Offset 1 .......................................................................... 23
Gain Calibration Offset 2 .......................................................................... 23
Gain Calibration Offset 3 .......................................................................... 23
Fixed Gain ................................................................................................ 25
Compander - Black slope, Slopes (MSBs) ............................................... 27
Compander Slope 1 (LSBs) ...................................................................... 27
Compander Slope 2 (LSBs) ...................................................................... 27
Compander Slope 3 (LSBs) ...................................................................... 28
Compander Slope 4 (LSBs) ...................................................................... 28
Compander Offset 1 ................................................................................. 28
Compander Offset 2 (MSBs) .................................................................... 29
Compander Offset 2 (LSBs) ..................................................................... 29
Compander Offset 3 (LSBs) ..................................................................... 29
Compander Offset 4 (LSBs) ..................................................................... 29
Contacting Cirrus Logic Support
For a complete listing of Direct Sales, Distributor, and Sales Representative contacts, visit the Cirrus Logic web site at:
http://www.cirrus.com/corporate/contacts/
Preliminary product information describes products which are in production, but for which full characterization data is not yet available. Advance product information describes products which are in development and subject to development changes. Cirrus Logic, Inc. has made best efforts to ensure that the information
contained in this document is accurate and reliable. However, the information is subject to change without notice and is provided “AS IS” without warranty of
any kind (express or implied). No responsibility is assumed by Cirrus Logic, Inc. for the use of this information, nor for infringements of patents or other rights
of third parties. This document is the property of Cirrus Logic, Inc. and implies no license under patents, copyrights, trademarks, or trade secrets. No part of
this publication may be copied, reproduced, stored in a retrieval system, or transmitted, in any form or by any means (electronic, mechanical, photographic, or
otherwise) without the prior written consent of Cirrus Logic, Inc. Items from any Cirrus Logic website or disk may be printed for use by the user. However, no
part of the printout or electronic files may be copied, reproduced, stored in a retrieval system, or transmitted, in any form or by any means (electronic, mechanical,
photographic, or otherwise) without the prior written consent of Cirrus Logic, Inc.Furthermore, no part of this publication may be used as a basis for manufacture
or sale of any items without the prior written consent of Cirrus Logic, Inc. The names of products of Cirrus Logic, Inc. or other vendors and suppliers appearing
in this document may be trademarks or service marks of their respective owners which may be registered in some jurisdictions. A list of Cirrus Logic, Inc. trademarks and service marks can be found at http://www.cirrus.com.
2
DS322PP1
CS7622
Compander X1 (MSBs) ............................................................................ 30
Compander X1 (LSBs) ............................................................................. 30
Compander X2 (MSBs) ............................................................................ 30
Compander X2 (LSBs) ............................................................................. 30
Compander X3 (MSBs) ............................................................................ 31
Compander X3 (LSBs) ............................................................................. 31
Device ID .................................................................................................. 31
Revision Code .......................................................................................... 31
5.0 PIN DESCRIPTIONS ......................................................................................... 32
Supply ...................................................................................................... 32
Ground ..................................................................................................... 32
CMOS Input ............................................................................................. 32
CMOS Analog Input ................................................................................. 33
CMOS 4 mA Output ................................................................................. 33
6.0 PACKAGE DIMENSIONS ................................................................................. 34
LIST OF FIGURES
Figure 1. SEN Timing.........................................................................................................................6
Figure 2. Serial Write Timing..............................................................................................................6
Figure 3. Read Data Timing ...............................................................................................................6
Figure 4. Digital Camera Block Diagram............................................................................................7
Figure 5. CS7622 Block Diagram.......................................................................................................7
Figure 6. Idealized CCD output waveform .........................................................................................8
Figure 7. Transfer function of VGA circuit (assuming full scale level of 1.0 V) ..................................9
Figure 8. Block diagram of CDS/VGA circuit......................................................................................9
Figure 9. Idealized timing diagram of VGA/CDS circuit ...................................................................10
Figure 10.Black level adjustment loop ..............................................................................................11
Figure 11.Transfer function of Vin to Gain Adjust output Block (assuming full scale level of 1.0 V).12
Figure 12.Gain Adjust output Block...................................................................................................12
Figure 13.13-to-10 bit compander.....................................................................................................13
Figure 14.CS7622 output data and clocks ........................................................................................14
Figure 15.Input Timing ......................................................................................................................14
Figure 16.Typical Connection Diagram.............................................................................................16
Figure 17.Transfer Function of Analog Input to Digital Output (assuming full scale level of 1.0 V) ..24
Figure 18.Transfer Function of ADC with Fixed Gain Settings (assuming full scale level of 1.0 V)..26
DS322PP1
3
CS7622
1.0 CHARACTERISTICS/SPECIFICATIONS
DIGITAL CHARACTERISTICS (TA = 25 °C; VDDD = 3.3 V)
Parameter
Logic Inputs
High-Level Input Voltage
Low-Level Input Voltage
Input Leakage Current
Logic Outputs
High-Level Output Source Current @ IOH = 4 mA
Low-Level Output Sink Current @ IOL = 4 mA
3-State Leakage Current
Symbol
Min
Typ
Max
Units
VIH
VDD-0.8
-
-
V
VIL
-
-
0.8
V
IIN
-
-
10
mA
VOH
VDD-0.4
-
VOL
-
-
0.4
mV
IOZ
-
-
10
µA
mV
POWER CONSUMPTION (TA = 25 ° C; VDDA = VDDD = 3.3 V; Output Load = 30 pF; Input Clock = 15MHz)
Parameter
Symbol
Min
Typ
Max
Units
Power Dissipation
Peak Mode
Preview Mode
Stand By Down
PD
PDLR
PDPD
-
214
162
0.0825
-
mW
mW
mW
Analog Power Supply Current
Peak Mode
Preview Mode
Stand By Down
IAN
IALR
IAPD
-
53
37
0.025
-
mA
mA
mA
Peak/Preview Mode
Preview Mode
IDN
IDPD
-
12
0
-
mA
mA
Symbol
Min
Typ
Max
Units
VDDA
VDDD
3.0
2.5
3.3
3.6
3.6
V
V
10
mV
Digital Power Supply Current
RECOMMENDED OPERATING CHARACTERISTICS
Parameter
Power Supply Voltage
GNDA to GNDD Voltage Differential
Analog Full Scale Input Voltage Range
AIN
300 mV
-
1V
Vp-p
-
20 MHz
-
MHz
Input Clock Rate
ABSOLUTE MAXIMUM RATINGS
Parameter
Power Supply Voltage
Symbol
Min
Max
Units
VDDA, VDDD
-0.3
6.0
V
GNDD-0.3
VDDD+0.3
V
GNDA-0.3
VDDA+0.3
V
10
mA
-0
+70
°C
+260
°C
+150
°C
Digital Input Voltage
Analog Input Voltage
Input Current
AIN
(except supply pins)
Ambient Temperature Range
Lead Solder Temperature (10sec duration)
Storage Temperature Range
-65
WARNING: Operation at or beyond these limits may result in permanent damage to the device.
Normal operation is not guaranteed at these extremes.
4
DS322PP1
CS7622
ADC (ANALOG-TO-DIGITAL CONVERTER)
Parameter
Symbol
Full Scale Input Voltage Range
Min
Typ
300 mV
Full Scale Input Voltage Range Resolution
Max
Unit
1V
V0-p
100
mV
ADC resolution
-
10
-
bits
Total Differential Non-Linearity
-
±1
-
LSB
Total Integral Non-Linearity
-
±1
-
LSB
Min
Typ
Max
Unit
1V
V0-p
CDS/VGA PARAMETERS
Parameter
Symbol
Input Voltage Range
300 mV
Total Gain Range
Input Referred Noise (rms)
Maximum Gain Setting
AVGA
-
18
-
dB
VnVGA
-
-
0.2
mV
SERIAL INTERFACE TIMING SPECIFICATIONS
Description
Enable Setup
SDAT Setup
SDAT Hold
Serial Clock Period
Write Data Invalid
Read Data Valid
Clock to Disable
SEN Rise to SEN Fall
(Note 1)
Symbol
t1
t2
t3
t4
t5
t6
t7
t8
Minimum
10
10
10
143
0
0
143
200
Maximum
10
10
-
Unit
ns
ns
ns
ns
ns
ns
ns
ns
Notes: 1. the minimum serial clock period must be longer than two pixel clock periods.
DS322PP1
5
CS7622
SEN
t1
t7
t8
SCLK
SDATI
R/W, ADDR <6.0>
DATA <7.0>
Figure 1. SEN Timing
SEN
t4
SCLK
SDATI
R/W
t2
A6
A5
A6
A3
t3
Figure 2. Serial Write Timing
SCLK
t5
SDATI
SDATO
t6
XX (DON’T CARE)
A0
D7
D6
D5
Figure 3. Read Data Timing
6
DS322PP1
CS7622
2.0 GENERAL DESCRIPTION
The CS7622 forms the heart of a four chip digital
CCD Camera. The four chips include the CCD imager, the CS7622 CCD digitizer, a vertical drive interface chip and a backend DSP chip to further
process the digital data (see Figure 4.)
The patented DRX technology allows the CS7622
to output data with 13-bit dynamic range, and at the
same time reducing the power consumption to a 10bit equivalent A/D converter.
The digitized output is either available in 13-bits,
12-bits or 10-bits. The 10-bit output is created by
companding the 13-bit A/D output to 10-bits. The
companding curve consists of 4 linear segments,
where each slope and each start point is user programmable.
A block diagram of the CS7622 chip is shown in
Figure 5.
CS7622
CDS/ADC
Video Output
Backend
DSP
CCD
Control
LCD Panel
Vertical Drive
Timing Signals
+5 V
+5 V to -5 V
DC-DC converter
Figure 4. Digital Camera Block Diagram
VDD[2]
GND[2]
AIN
Σ
CDS/VGA
A/D
Gain
Adjust
13 to 10-bit
Compander
M
U
X
Black
Level
CK_FT
CK_DATA
Clock
Generator
Reference
Serial Interface
SEN
DOUT[12:0]
(up to 3 may be unused)
CLKO
CLAMP
TEST
RST
REF_CAPP
1 µF
REF_CAPN
BG_RES
10 kΩ
SDATI SDATO SCLK
Figure 5. CS7622 Block Diagram
DS322PP1
7
CS7622
PIXEL PERIOD
VIDEO
SIGNAL
RANGE
RESET LEVEL
FEED THROUGH
LEVEL
DARK
VIDEO LEVEL
MAX. BRIGHTNESS
Figure 6. Idealized CCD output waveform
3.0 OPERATION
3.1 CDS/VGA (correlated double
sampling/variable gain amplification)
An idealized waveform of the CCD output is
shown in Figure 6.
The CCD output contains reset noise, thermal
noise, and 1/f noise generated in the CCD output
circuit. This degrades the S/N ratio and must be
cancelled. Since the noise during the active video
portion of the CCD signal is assumed to be correlated with the noise during the feed through portion
of the signal, this noise can be cancelled by subtracting the feed through level from the video level.
This operation is called correlated double sampling. The active video signal is the difference between the feed through and video levels. The active
video signal varies according to light conditions. In
order to insure that the full dynamic range of the
ADC is utilized even under low light conditions,
the CCD output is amplified using a VGA. The
gain control is provided by a 2 bit control word
generated by an ADC after stage 1, which has a
gain of 1. Based on the input voltage, a gain of 1x,
2x, 4x, or 8x is subsequently applied to the signal.
The amount of gain is later adjusted in the digital
section. After the VGA, the signal gets digitized by
a 10 bit ADC. The 2 bit ADC output is used in
8
combination with the 10 bit ADC output to produce
a 13 bit output.
Adding more gain before the ADC does not offer
performance improvement because the noise of the
CCD (after gain is applied to it) begins to dominate
over the quantization noise. Any additional gain
should be done in digital since the performance is
the same as when the ADC output has the additional gain applied.
In order to add more flexibility, the full scale input
range is programmable through register 05h. This
setting will determine what input level maps to the
highest ADC output code. Thus depending on the
saturation level of the particular CCD used in the
system, an appropriate full scale input level can be
chosen in the CS7622. The choices of full scale input level are 300 mV to 1 V in 100 mV increments.
In the remainder of this document, all the figures
and discussions assume a full scale level of 1 V is
used.
The transfer function of the VGA portion of the circuit is shown in Figure 7 with full scale level = 1 V.
It is assumed that the CDS has already been performed. If desired, the gain switching functionality
can be disabled and forced to a fixed gain of 8x, 4x,
2x, or 1x. This way any dynamic range enhancement is lost and the digital output is only 10 bits. If
DS322PP1
CS7622
VOUT (V)
1.07
0.5
8X
0.125
00
4X
2X
0.5
0.25
01
1X
1.0
10
11
VIN (V)
ADC OUTPUT
Figure 7. Transfer function of VGA circuit (assuming full scale level of 1.0 V)
Φ2
Φ1
C3
C1
100 KΩ
Φ1
C1
C5
C2
C4
VOUT
VIN
-A1
100 KΩ
Cb
VREF
-A2
Vo1
Vo2
STAGE 1
STAGE 2
ADC
2
-A3
STAGE 3
CONTROLS C3, C5
CONTROLS GAIN ADJUST BLOCK IN DIGITAL
Figure 8. Block diagram of CDS/VGA circuit
a fixed gain of 1x is selected, DOUT[12:3] is used
as the output, a fixed gain of 2x will use
DOUT[11:2], etc. In order to use this mode, the
fixed gain register (14h) should be set and the calibration offset registers (OEh - 10h) should be set to
0.
The CDS/VGA circuit is composed of three stages.
The first stage has a fixed gain of 1, and the second
and third stages have variable gain with a combined
gain range of 1 to 8 (0-18 dB). Figure 8 shows a
block diagram of the CDS/VGA circuit. The total
gain is A = (C2/C3)(C4/C5) which is adjusted by
varying C3 and C5. The capacitor Cb on the front
DS322PP1
of stage 1 is for black level adjustment and will be
discussed in detail later.
This circuit utilizes a two phase non-overlapping
clock to perform the desired CDS function. The
two phase clock also allows the video signal to be
passed to the output while retaining a positive polarity signal. Figure 9 shows a timing diagram of
the two phase clock along with the CCD signal and
output signals of stages one, two and three.
There is an internal mid-scale DC bias level circuit
at the input pin. This allows AC coupling into the
CS7622 with a capacitor and having the input auto-
9
CS7622
matically biased to mid-supply without worrying
about external circuitry to perform this task.
3.2 Black Level Adjustment
In order to maintain a constant reference level for
black pixels, a feedback loop is implemented that
sets the black level value at the output of the ADC
to 64 in the 13 bit digital code. This loop is active
during the optically black pixels which are output
at the beginning and end of a frame as well as during a portion of the horizontal blanking period. The
presence of black pixels in the CCD output is indicated by the CLAMP pulse, which is supplied externally through the CLAMP pin. The black level
can also be written to through the serial port.
In order to acquire a starting value for the black level, the loop will run over the several lines of black
pixels at the beginning of the frame. The block diagram of the loop is shown in Figure 10. The update rate is once per line during active pixel lines as
long as the Clamp pulse is < n+10 cycles. Where n
is the number of pixels accumulated before the
black loop is updated and is programmable through
register 0Dh bits 5:0. If the Clamp pulse is longer
CCD
INPUT
SIGNAL
than n+10 cycles the black loop is updated every
n+10 cycles. For example, during optical back lines
the loop is updated several times at a rate of once
every n+10 cycles.
The open-loop transfer function of the black level
adjustment loop is
K×n
H ( z ) = ------------z–1
blk_gain = 1, 2, 4, or 8
where blk_gain is programmable through a register
and n = # of black pixels during clamp time, which
is also programmable. The value of Kxn will determine the open-loop gain of the system. The settling
time for the loop can be calculated using the following formula:
For offset range=1 (reg 06h, bit 0)
1
1
τ =  – ---------------------------  -----
ln ( 1 – nK ) fu
For offset range =0



 1
1
τ =  – ----------------------------  -----
fu
 ln  1 – nK

------- 


2
V(2)
V(1)
1
K = --------- blk_gain
256
V(3)
ck_ft
ck_data
OUT OF
STAGE 1
OUT OF
STAGE 2
OUT OF
STAGE 3
V(1)
V(2)
V(1)
V(3)
V(2)
V(1)
V(3)
V(2)
Figure 9. Idealized timing diagram of VGA/CDS circuit
10
DS322PP1
CS7622
During fixed gain mode the time constant is a little
different.



 1
1
τ =  – ----------------------------  -----
 fu
 ln  1 – nK
------- 


8
For a fixed gain of 2:



 1
1
τ =  – ----------------------------  -----
 fu
 ln  1 – nK
------- 


4
For a fixed gain of 4:



 1
1
τ =  – ---------------------------  -----
 fu
 ln  1 – nK

------- 


2
For a fixed gain of 8:
1
1
τ =  – --------------------------  -----
ln ( 1 – nK )  fu
3.3 Gain Adjust Block
In order to increase the dynamic range of the ADC,
a variable gain, whose value is determined by the
signal level, is applied to each pixel. This allows
for 13 bits of dynamic range and 10 bits of resolution after accounting for the significance of the
ADC output bits. The gain applied in the analog is
illustrated in the transfer curve in Figure 7. Once
the signal is digitized, the gain adjust block uses the
gain information for a given pixel word and shifts
its bits accordingly. For example, using a full scale
level of 1.0 V, if Vin = 0.3 V, the VGA would
choose a gain of 2X so the ADC input is 0.6 V. The
10-bit output of the ADC (with no black level) is
(0.6/1.0) × 1024 = 614, or “1001100110.” in binary. The gain adjust block will take this value plus
the bits representing the 2x gain and divide the output by two (shift right by 1). The output of the gain
adjust block is then “0100110011.000.” Note that
the decimal point is virtual, having no existence in
silicon. It is representing the fact that we keep 3 extra bits of lower significance in the output. In the
same manner, if Vin = 0.75 V, a gain of 1X would
be chosen and the output of the gain adjust block
In order to achieve no ringing in the settling use,
n
n
---- ≤ 1 for offset range = 1, and ------≤ 1 for offset range
K
2K
= 0.
The 9 MSBs of the black level accumulator can be
read or written through a register. If written, the
LSBs are set to zero. The black level is set to “8” in
a 10-bit digital output representation. In a 13-bit
representation, it is set to “64.” The power-up default value in the accumulator is at mid level.
‘64’
VIN
10
Σ
CDS/VGA
ADC
-
+
FROM SERIAL INTERFACE
BLK LVL LOOP
GAIN REG
CLIP
7
K
+
FU = UPDATE FREQUENCY
FP = PIXEL FREQUENCY
Z-1
Z-1
FP
FU
DAC
+
MUX
For a fixed gain of 1:
Also note that the black level adjust loop can be
disabled. In addition, the black level can be programmed through the serial port.
BINARY
TO
THERM
9
Figure 10. Black level adjustment loop
DS322PP1
11
CS7622
would be “1100000000000.” The transfer function
of the Vin/gain adjust out is shown in Figure 11.
A block diagram of the gain adjust block is shown
in Figure 12.
Since the analog gain changes do not match the
digital shifts exactly, there is a potential to have
non-monotonic digital output. In order to remove
this problem, calibration is performed. During calibration, offset values are found that will be used to
counteract the errors caused by the analog gain
mismatch. Using these offset values, the final output is a monotonic continuous 13-bit value.
3.4 13-to-10 Bit Compander
While a 13 bit output may be useful in some applications, others may require the standard 10 bit output. To accommodate this and yet still retain the
advantages of the increased dynamic range, a 13to-10 (or 13-to-12) bit compander is included. By
using the picture content as a guide, the user can select which curve will lead to the best overall dynamic range in the picture. The Companding
module takes 13-bit data as input, and outputs either 10-bit companded data, 12-bit MSB-clipped
data or it lets the original 13-bit data pass through.
By programming the compander in the way that is
shown in Figure 13, it is possible to compensate for
DIG ADJUST OUT (13 BITS)
8192
4096
8X
4X
2X
1X
2048
1024
0
0.25
0.125
00
01
0.5
1.0
10
11
VIN (V)
ADC OUTPUT
Figure 11. Transfer function of Vin to Gain Adjust output Block (assuming full scale level of 1.0 V)
ADC OUTPUT
10
VGA_ADC OUTPUT
2
GAIN ADJUST
13
TO DIGITAL GAIN
SHIFT BY 0,1,2, OR 3
Figure 12. Gain Adjust output Block
12
DS322PP1
CS7622
backlighting conditions. Details in dark areas stay
visible, even in very complex lighting conditions.
These three modes can be selected through 2 register bits in operational control.
Bits_out register bits
0x
10
11
Output mode
10 bits companded
13 bits
12 bits (clipped)
ble offset value, offset1. This may be set to 0 if desired. This option will lose the “blacker-thanblack” pixel information, but allow for slightly
more dynamic range. Note: If using the linear mode
(option 1), offset1 must be set to 8.
Registers x1 through x3 should be programmed
with the x coordinates of each one of the three
knees.
Table 1.
In the 12-bit clipped mode, any input above 4095
gets clipped to 4095. In the 10-bit companded
mode, the input gets companded through a four
segment, three knees, fully programmable curve.
Registers slope1 through slope4 should be programmed with 256 multiplied by the calculated
slopes.
Finally, the offsets can be programmed following
the formulas below:
To program the curve, the placement of the three
knees in the companding curve must be determined. The next step is to determine the slope of
the four segments created by the three knees (slope
for each segment is defined as delta y / delta x). Finally, offsets must be calculated to keep the companding curve continuous.
y1 = slope1/256 × (x1-64) + offset1
A fourth knee exists in the curve, which represents
the black level value. There are two options for the
10-bit black value. In case one, a linear mapping is
employed such that “blacker-than-black” pixel information is kept, with black (code 64 in the 13 bit
data) being defined as code 8 in the 10 bit domain.
The second option clips all pixel values less than
black (code 64 in the 13 bit data) to a programma-
offset4 = y3 - (x3 × slope4 / 256)
y2 = slope2/256 × (x2-x1) + y1
y3 = slope3/256 × (x3-x2) + y2
offset2 = y1 - (x1 × slope2 / 256)
offset3 = y2 - (x2 × slope3 / 256)
(use integer division and discard the remainder)
When using the 10 bit companded output, be aware
of the non-linearity of the output data. If linear output is needed to perform Auto White Balance
(AWB) or Automatic Gain Control (AGC), a linear
curve can be implemented to gather statistics. This
can be achieved by writing 8191 to x1 (set register
CODE_OUT
1023
(x3,y3)
OFFSET4
SLOPE4
(x2,y2)
SLOPE3
OFFSET3
(x1,y1)
SLOPE2
OFFSET2
OFFSET1
SLOPE1
64
X1
X2
X3
8191
CODE_IN
Figure 13. 13-to-10 bit compander
DS322PP1
13
CS7622
1Fh to 1fh and set register 20h to ffh) and setting
slope1 to 32 (set register 15h to 00010xxxb and set
register 16h to 20h). Once the statistics have been
gathered, all four registers should be returned to
their previous values before taking the actual picture.
The output of the compander is available at the pins
DOUT<9:0> and it makes transitions either at the
falling or rising edges of the pixel rate clock CLKO, controlled by a register bit. The Falling edge
option is shown in Figure 14.
strongly recommend that the chip should be kept in
Stand By mode when not in use in order to save
power. When in preview mode, a user may wish to
cut down the resolution of the ADC output to 6 bits
in order to reduce the power consumption of the
CS7622. In this mode, the current is reduced by
20 mA. With the DRX (Dynamic Range eXtension) circuitry, 3 bits of dynamic range are added to
the 6-bit ADC output producing a 9-bit output. The
pins DOUT[12:4] are used to output the digitized
data in preview or Stand By mode.
3.5 Stand By and Preview Mode
3.6 Serial Interface
In order to enter power down mode a value of 07h
must be written to register 01h. This will power
down all the analog sections. Stopping the input
clocks will power down the digital. To power up
again, the input clocks must be turned on first then
a value of 00h needs to be written to register 01h.
The user must wait at least 500µs for the internal
analog references to settle to their appropriate values before normal operation is resumed. It is
The serial interface is designed to allow high speed
input to control the chip’s registers. The specifications on this interface are as follows:
Asserting the enable pin, SEN, enables the serial
interface to perform data transfers. Data present on
the SDATI pin is latched into the CS7622 on each
rising edge of the serial clock, SCLK. Data output
on SDATO from the CS7622 is clocked out on the
rising edge of SCLK.
CLKO
DOUT<9:0>
Figure 14. CS7622 output data and clocks
T1
T4
CCD
INPUT
SIGNAL
CK_FT
T3
T2
CK_DT
Figure 15. Input Timing
14
DS322PP1
CS7622
The CS7622 receives only the first 16 rising edges
of the SCLK while SEN is low and then ignores
any remaining SCLK and SDATI information. If
SEN goes high before 16 SCLK pulses have been
received, the CS7622 aborts the serial transfer.
The first bit is the R/W bit. R/W = 1 identifies the
transfer as a read. If (0), the transfer is a write. The
next seven bits define the address. For write transfers, the second byte of the 16-bit packet contains
the data byte. For read transfers, the CS7622 outputs the read data on SDATO after accepting the
address. Address and data are transferred MSB
first. When not reading out data, the SDATO pin is
not driven by the chip (Hi-Z state).
The timing diagrams and specifications are shown
in “Serial Interface Timing Specifications” on
page 5 and Figures 1, 2, and 3 on page 5.
3.7 Input Timing for Sampling Clocks
The input clocks CK_FT and CK_DT are used to
set up the sampling times and also to generate the
internal digital clock. These clocks need to be running when processing pixels from the CCD, writing
to the chip registers, or performing calibration (See
DS322PP1
Register Description of Operation Control 2 reg
05h bit 0 for the details of performaing a calibration). The timing of these clocks is important to ensure optimum settling times and sampling the
correct value. CK_FT and CK_DT need to be nonoverlapping pulses made as wide as possible to
give long settling times. The falling edge of
CK_FT should be close to the end of feedthrough
while the falling edge of CK_DT should be close to
the end of the data section of the CCD signal. See
figure 15. Typical timing is given in table 2.
Timing Parameter
Typical Operating
Values
T1, T4
2 ns
T2, T3
5 ns
Table 2.
Longer non-overlapping values for T1 and T4 will
increase the recovery time, thus requiring a slower
clock rate.
15
CS7622
VCC
13 28
VAA
Sampling
Signals
CK_FT
CK_DT
19 CK_FT
20 CK_DATA
13
DOUT[0:12]
CLKO
21
to Mic
CS7622
17
RESET
NC
16
9
RST
DIAG
TEST
REF_CAPP
REF_CAPN
18
from
Microcontroller
BG_RES
CLAMP
15
1 µF
14
10
10 kΩ ±1%
5
SCLK
7
SDATI
6
SDATO
8
SEN
from
CCD
11
AIN
GND
12 29
Figure 16. Typical Connection Diagram
16
DS322PP1
CS7622
4.0 REGISTER DESCRIPTIONS
Register (hex)
00h
01h
02h - 03h
04h
05h
06h-0Ah
0Bh
0Ch
0Dh
0Eh
0Fh
10h
11h - 13h
14h
15h
16h
17h
18h
19h
1Ah
1Bh
1Ch
1Dh
1Eh
1Fh
20h
21h
22h
23h
24h
25h
26h
Register Function
Software Reset
Power Down Control 1
Reserved
Operation Control 1
Operation Control 2
Reserved
Black Level Control - Accumulator (LSB)
Black Level Control - Accumulator (MSB)
Black Level Control - Loop Gain, Clamp Length
Gain Calibration - Offset 1
Gain Calibration - Offset 2
Gain Calibration - Offset 3
Reserved
Gain Calibration - Fixed Gains
Compander - Black slope, Slopes (MSBs)
Compander - Slope1 (LSBs)
Compande - Slope2 (LSBs)
Compander - Slope3 (LSBs)
Compander - Slope4 (LSBs)
Compander - Offset1
Compander - Offsets (MSBs)
Compander - Offset2 (LSBs)
Compander - Offset3 (LSBs)
Compander - Offset4 (LSBs)
Compander - X1 (MSBs)
Compander - X1 (LSBs)
Compander - X2 (MSBs)
Compander - X2 (LSBs)
Compander - X3 (MSBs)
Compander - X3 (LSBs)
Device ID
Rev Code
Access
W
R/W
R/W
R/W
R/W
Default value (hex)
00h
00h
00h
0Ah
04h
R/W
R/W
R/W
R/W
R/W
R/W
00h
01h
2Ah
00h
00h
00h
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R
R
00h
10h
B2h
60h
20h
07h
08h
0Bh
BFh
05h
20h
03h
20h
05h
18h
0Bh
58h
CCh
00h
Table 3. Register Description
DS322PP1
17
CS7622
Reset
Default = 00h; Read/Write (address 00h)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
RESERVED
0
Bit
Mnemonic
7:1
-
0
sft_rst
0
0
0
0
sft_rst
0
0
0
0
Function
reserved
Software Reset: When this bit is written with a ‘1’, all of the digital circuitry and
the registers will reset to their default values. It automatically clears after 4 pixel
clock periods. The clocks remain running during the reset period.
Power down Control 1
Default = 00h; Read/Write (address 01h)
Bit Number
Bit Name
Default
18
7
6
5
4
3
RESERVED
0
0
0
0
0
2
1
0
pd_vga
pd_adc
pd_ref
0
0
0
Bit
Mnemonic
Function
7:3
-
2
pd_vga
DRX Front End Power Down: When written with a ‘1’, the DRX front end circuitry powers down.
1
pd_adc
ADC Power Down: When written with a ‘1’, the Analog-to-Digital converter circuitry powers down.
0
pd_ref
Voltage Reference Power Down: When written with a ‘1’, the Analog-to-Digital
converter circuitry powers down.
reserved
DS322PP1
CS7622
Operation Control 1
Default = 0Ah; Read/Write (address 04h)
Bit Number
Bit Name
Default
7
Bit
Mnemonic
7:6
-
5
dout_edge
4
3:2
1
0
DS322PP1
6
RESERVED
0
5
4
3
2
1
0
dout_edge
low_res
bits_out1
bits_out0
blk_dis
off_range
0
0
1
0
1
0
0
Function
reserved
This register is used to set when dout changes values. Relative to CLKO
0 - dout output changes on the falling edge of CLKO
1 - dout output changes on the rising edge of CLKO
low_res
Preview Mode: This mode can be used to cut the current consumption of the
chip by 20 mA. The output of the ADC will have 6 bits of resolution in this mode,
and the output of the chip will have 9 bits after using the DRX circuitry. It is intended to be used when driving an LCD display or any other time when a lower
resolution picture is acceptable.
bits_out1-0
Number of Data Bits Out: The range of the output data can be determined by
these bits. The data internal to the chip has a 13-bit range. The output can be
this full range, half this range (12 bits), or an eighth of this range (10 bits). If 12bit data is selected, the top half of the 13-bit range is saturated to the maximum
12-bit code. If 10-bit data is selected, the compander curve which is user programmable is employed to map the 13-bit data to the 10-bit output.
0 - 10 bits output; 1 - 10 bits output
2 - 13 bits output; 3 - 12 bits output
blk_dis
Black Level Loop Disabled: If the user chooses to adjust the black level himself through register access, he may disable the internal black level loop. This
loop usually updates the black level to what it calculates to be the correct level.
If disabled, the offset used will be determined from the value written in the black
level accumulator register.
0 - internal black level loop is enabled
1 - black level loop is disabled
off_range
Offset Range: The black level loop is used to cancel any offsets from the CCD
and chip circuitry. If the offsets are small, the user has the option to decrease
the offset cancellation range for the added advantage of increasing the resolution of the offset cancellation.
0 - smaller offset cancellation range used (~50 mV)
1 - larger offset cancellation range used (~100 mV)
19
CS7622
Operation Control 2
Default = 04h; Read/Write (address 05h)
Bit Number
Bit Name
Default
7
6
0
0
5
4
RESERVED
Bit
Mnemonic
7:4
-
3:1
0
0
0
3
2
1
0
fs_lvl2
fs_lvl1
fs_lvl0
gain_cal
0
1
0
0
Function
reserved
fs_lvl2-0
Full Scale Level: This is used to set the full scale input range of the CS7622.
Since CCDs have various saturation levels, it is advantageous to set the full
scale input range of the CS7622 to match the saturation level of the CCD used.
The table below shows the full scale level choices. (See Table 4)
gain_cal
Gain Calibration: A calibration of the gain stages is required to insure a monotonic digital output. If the user wishes to initiate a calibration, he may do so by
setting this bit to ‘1’, which will invoke a gain calibration sequence immediately.
This bit automatically clears itself after a calibration has been initiated. During
the calibration sequence the output will not contain valid data. The input clocks
must be running throughout the whole calibration sequence which lasts for
~760 clocks.
fs_lvl
Full Scale Voltage
000
0.3 V
001
0.4 V
010
0.5 V
011
0.6 V
100
0.7 V
101
0.8 V
110
0.9 V
111
1.0 V
Table 4.
Black Level Control (8 LSBs)
Default = 00h; Read/Write (address 0Bh)
Bit Number
Bit Name
Default
20
7
6
5
4
3
2
1
0
accumulator accumulator accumulator accumulator accumulator accumulator accumulator accumulator
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
Bit
Mnemonic
7:0
accumulator7-0
Function
Black Level Accumulator: See the description of register OCh.
DS322PP1
CS7622
Black Level Control (MSB)
Default = 01h; Read/Write (address 0Ch)
Bit Number
Bit Name
Default
7
6
5
0
0
0
3
2
1
0
0
0
RESERVED
Bit
Mnemonic
7:1
-
0
4
0
0
accumulator8
1
Function
Reserved
accumulator8
Black Level Accumulator: is a 9 bit number representing an amount of offset
added to the input of the CDS circuit. The black level loop alters the black level
accumulator value to make the output of the ADC settle to code 64 during black
pixels. If desired the black loop may be disabled and written to manually to add
any desired amount of offset. There is a total of ~100 mV of offset range if the
offset range register setting is set to “1” or ~50 mV when this register setting is
set to “0”. This offset range is used to correct for CCD offsets plus internal offsets generated in the analog path of this chip. The offset range before subtracting the internal offsets is as shown in the table below with the worst case
internal offsets being ±17 mV.(See Table 5)
Offset Range
(Reg 06h bit 0)
1
0
Max Offset
Blk Acc=511
~30 mV
~11 mV
Min Offset
Blk Acc=0
~-72 mV
~-40 mV
Accumulator
LSB Size
~0.2 mV
~0.1 mV
Table 5.
Black Level Control - General
The black loop is a feedback system that causes the ADC output to settle to 64 during the register defined
black pixels. This has the purpose of removing any CCD and system offsets and defining 64 as the known
black level. The loop has an exponential settling response and the time constant of this loop is effected
by the black loop gain and the number of black pixels to accumulate before updating the black accumulator. See Figure 10 for a block diagram of the black level loop.
DS322PP1
21
CS7622
Black Level Control - Loop Gain, Clamp Length
Default = 2Ah; Read/Write (address 0Dh)
Bit Number
Bit Name
Default
Bit
7:6
5:0
7
6
5
4
3
2
1
0
blk_gain1
blk_gain0
blk_clp_15
blk_clp_14
blk_clp_13
blk_clp_12
blk_clp_11
blk_clp_10
0
0
1
0
1
0
1
0
Mnemonic
Function
blk_gain1-0
Black Loop Gain Factor: can be set to 1x,2x,4x,or 8x and is simply a multiplying constant to effect the weight of each black pixel before it is accumulated.
00 - defines a gain of 1x
01 - defines a gain of 2x
10 - defines a gain of 4x
11 - defines a gain of 8x
blk_clp_15-10
Black Loop Clamp Length: The black clamp length effects the loop time constant and also acts to average out noise in the black level. The larger this value
the more pixels that are summed before the loop is updated which causes
greater averaging and a smaller settling time constant.
The table below shows the black loop time constant for various settings of Offset Range (register 04h, bit 0) and Fixed Gain Settings (register 14h, bits 5-3).
(See Table 5)
Fixed Gain
(Register 16h)
not fixed
x1
x2
x4
x8
Offset Range = 1
Offset Range = 0
-1/(ln(1-nK))(1/fu)
-1/(ln(1-nK/8))(1/fu)
-1/(ln(1-nK/4))(1/fu)
-1/(ln(1-nK/2))(1/fu)
-1/(ln(1-nK))(1/fu)
-1/(ln(1-nK/2))(1/fu)
-1/(ln(1-nK/16))(1/fu)
-1/(ln(1-nK/8))(1/fu)
-1/(ln(1-nK/4))(1/fu)
-1/(ln(1-nK/2))(1/fu)
Table 6.
Where:
K = 1/256*blk_gain
n = Black loop clamp length = blk_clp_l[5:0]
fu = update rate
22
DS322PP1
CS7622
Gain Calibration Offset 1
Default = 00h; Read only (address 0Eh)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
0
gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset
17
16
15
14
13
12
11
10
0
0
0
0
0
0
0
0
Bit
Mnemonic
Function
7:0
gain_offset17-10
offset added to 4x gain segment, values are in 2’s complement. See details in
register 10h.
Gain Calibration Offset 2
Default = 00h; Read only (address 0Fh)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
0
gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset
27
26
25
24
23
22
21
20
0
0
0
0
0
0
0
0
Bit
Mnemonic
Function
7:0
gain_offset27-20
offset added to 2x gain segment, values are in 2’s complement. See details in
register 10h.
Gain Calibration Offset 3
Default = 00h; Read only (address 10h)
Bit Number
Bit Name
Default
Bit
7:0
DS322PP1
7
6
5
4
3
2
1
0
gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset gain_offset
37
36
35
34
33
32
31
30
0
0
0
0
0
0
0
0
Mnemonic
Function
gain_offset37-30
Offset added to 1x gain segment. Values are in 2’s complement.
These registers are used to report some of the calibration settings. After calibration is performed the gain offset registers are automatically updated with values needed for the DRX circuitry to operate correctly. These registers should
not be written to since this will remove the proper settings found during calibration. The gain offset values are used to add an offset to the output of the ADC
when using different analog gain settings (See equations below). The purpose
of this is to produce a continuous transition between the different gain settings
so that the final 13 bit output is monotonic and has no undesired artifacts. (See
Figure 17)
{ADC_out
if in the 8x gain segment}
dout[12:0] =
{ADC_out*2+Offset1
if in the 4x gain segment}
{ADC_out*4+Offset2*2
if in the 2x gain segment}
{ADC_out*8+Offset3*4
if in the 1x gain segment}
23
CS7622
ADC OUT
1024
512
8X
4X
2X
1X
64
4096
USE OFFSET3
USE OFFSET1
8192
USE OFFSET2
INPUT
2048
1024
64
1.0
INPUT
Figure 17. Transfer Function of Analog Input to Digital Output (assuming full scale level of 1.0 V)
24
DS322PP1
CS7622
Fixed Gain
Default = 00h; Read/Write (address 14h)
Bit Number
Bit Name
Default
7
RESERVED
0
Bit
Mnemonic
7:6, 2:0
-
5:3
DS322PP1
6
fixed_gain2-0
5
4
3
2
fixed_gain2 fixed_gain1 fixed_gain0
0
0
0
0
1
0
RESERVED
0
0
0
Function
Reserved
Fixed Gain: This is used to turn off the DRX functionality and apply a fixed gain
to the input before reaching the ADC. A setting of 000 is used for normal operation this will yield the largest dynamic range by switching the front end gain relative to the amplitude of the input signal. The settings 001, 010, 011, and 100
are for fixed gains of 1x, 2x, 4x, and 8x respectively. Figure 18 shows the transfer function of the output of the ADC for a given input with the various fixed gain
settings.
25
CS7622
FIXED GAIN = 000
ADC OUTPUT
1024
8X
4X
2X
1X
INPUT
0.125
0.25
ADC OUTPUT
0.5
1.0
FIXED GAIN = 001
1024
1X
1.0
ADC OUTPUT
INPUT (V)
FIXED GAIN = 010
1024
2X
INPUT (V)
0.5
FIXED GAIN = 011
ADC OUTPUT
1.0
1024
4X
INPUT (V)
ADC OUTPUT
0.25
0.5
FIXED GAIN = 100
1.0
0.25
1.0
1024
8X
INPUT (V)
0.125
0.5
Figure 18. Transfer Function of ADC with Fixed Gain Settings (assuming full scale level of 1.0 V)
26
DS322PP1
CS7622
Compander - Black slope, Slopes (MSBs)
Default = 10h; Read/Write (address 15h)
Bit Number
Bit Name
Default
7
6
5
RESERVED
0
Bit
Mnemonic
7:5
-
0
0
4
3
2
1
0
comp_linear
slope18
slope28
slope38
slope48
1
0
0
0
0
Function
Reserved
Compander Black Level Slope: 0 - The values of “0” to “64” in a 13 bit representation are set to “offset1” in a 10 bit representation. Offset1 can be set in
register 33h.
1 - In this case the black level is mapped linearly from 13 bit values to 10 bit
values. “64” is mapped into “8”. All the other values between “0” and “64” are
divided by 8 in order to get the 10 bit representation. (See Figure 13)
4
comp_linear
3
slope18
Compander Slope 1: MSB of slope of first segment of companding curve.
(See Figure 13)
2
slope28
Compander Slope 2: MSB of slope of second segment of companding curve.
(See Figure 13)
1
slope38
Compander Slope 3: MSB of slope of third segment of companding curve.
(See Figure 13)
0
slope48
Compander Slope 4: MSB of slope of fourth segment of companding curve.
(See Figure 13)
Compander Slope 1 (LSBs)
Default = B2h; Read/Write (address 16h)
Bit Number
Bit Name
Default
7
6
3
2
1
0
slope16
5
slope15
4
slope17
slope14
slope13
slope12
slope11
slope10
1
0
1
1
0
0
1
0
Bit
Mnemonic
7:0
slope17-10
Function
Compander - Slope1: Slope of first segment (slope1[8:0]) of companding
curve. Max value is 1.996. The LSB step size is 0.0039. (See Figure 13)
Compander Slope 2 (LSBs)
Default = 60h; Read/Write (address 17h)
Bit Number
Bit Name
Default
7
6
3
2
1
0
slope26
5
slope25
4
slope27
slope24
slope23
slope22
slope21
slope20
0
1
1
0
0
0
0
0
Bit
Mnemonic
Function
7:0
slope27-20
Compander - Slope2: Slope of second segment (slope2[8:0]) of companding
curve. Max value is 1.996. The LSB step size is 0.0039. (See Figure 13)
DS322PP1
27
CS7622
Compander Slope 3 (LSBs)
Default = 20h; Read/Write (address 18h)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
0
slope37
slope36
slope35
slope34
slope33
slope32
slope31
slope30
0
0
2
0
0
0
0
0
Bit
Mnemonic
7:0
slope37-30
Function
Compander - Slope3: Slope of third segment (slope3[8:0]) of companding
curve. Max value is 1.996. The LSB step size is 0.0039. (See Figure 13)
Compander Slope 4 (LSBs)
Default = 07h; Read/Write (address 19h)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
0
slope47
slope46
slope45
slope44
slope43
slope42
slope41
slope40
0
0
0
0
0
1
1
1
Bit
Mnemonic
7:0
slope47-40
Function
Compander - Slope4: Slope of fourth segment (slope4[8:0]) of companding
curve. Max value is 1.996. The LSB step size is 0.0039. (See Figure 13)
Compander Offset 1
Default = 08h; Read/Write (address 1Ah)
Bit Number
Bit Name
Default
28
7
6
3
2
1
0
offset16
5
offset15
4
offset17
offset14
offset13
offset12
offset11
offset10
0
0
0
0
1
0
0
0
Bit
Mnemonic
7:0
offset17-10
Function
Compander - Offset1: Black level value of companding curve if not in linear
mapping mode (comp_linear = 0). (See Figure 13)
DS322PP1
CS7622
Compander Offset 2 (MSBs)
Default = 0Bh; Read/Write (address 1Bh)
Bit Number
Bit Name
Default
7
6
RESERVED
0
5
4
3
2
1
0
offset29
offset28
offset39
offset38
offset49
offset48
0
0
1
0
1
1
0
Bit
Mnemonic
Function
7:6
-
5:4
offset29-28
MSBs of offset of second segment of companding curve. (See Figure 13)
3:2
offset39-38
MSBs of offset of third segment of companding curve. (See Figure 13)
1:0
offset49-48
MSBs of offset of fourth segment of companding curve. (See Figure 13)
Reserved
Compander Offset 2 (LSBs)
Default = BFh; Read/Write (address 1Ch)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
0
offset27
offset26
offset25
offset24
offset23
offset22
offset21
offset20
1
0
1
1
1
1
1
1
Bit
Mnemonic
Function
7:0
offset27-20
Offset of second segment (offset2[9:0]) of companding curve. (See Figure 13)
Compander Offset 3 (LSBs)
Default = 05h; Read/Write (address 1Dh)
Bit Number
Bit Name
Default
7
6
3
2
1
0
offset36
5
offset35
4
offset37
offset34
offset33
offset32
offset31
offset30
0
0
0
0
0
1
0
1
Bit
Mnemonic
7:0
offset37-30
Function
Offset of third segment (offset3[9:0]) of companding curve. (See Figure 13)
Compander Offset 4 (LSBs)
Default = 20h; Read/Write (address 1Eh)
Bit Number
Bit Name
Default
7
6
offset47
0
Bit
Mnemonic
7:0
offset47-40
DS322PP1
4
3
2
1
0
offset46
5
offset45
offset44
offset43
offset42
offset41
offset40
0
1
0
0
0
0
0
Function
Offset of fourth segment (offset4[9:0]) of companding curve. (See Figure 13)
29
CS7622
Compander X1 (MSBs)
Default = 03h; Read/Write (address 1Fh)
Bit Number
Bit Name
Default
7
6
5
RESERVED
0
Bit
Mnemonic
7:5
-
4:0
x112-x18
0
0
4
3
2
1
0
x112
x111
x110
x19
x18
0
0
0
1
1
Function
Reserved
End value of first segment of companding curve (MSBs). (See Figure 13)
Compander X1 (LSBs)
Default = 20h; Read/Write (address 20h)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
0
x17
x16
x15
x14
x13
x12
x11
x10
0
0
1
0
0
0
0
0
Bit
Mnemonic
7:0
x17x10
Function
End value of first segment (x1[12:0]) of companding curve (LSBs). (See
Figure 13)
Compander X2 (MSBs)
Default = 05h; Read/Write (address 21h)
Bit Number
Bit Name
Default
7
6
5
RESERVED
0
Bit
Mnemonic
7:5
-
4:0
x212-x28
0
0
4
3
2
1
0
x212
x211
x210
x29
x28
0
0
1
0
1
Function
Reserved
End value of second segment of companding curve (MSBs). (See Figure 13)
Compander X2 (LSBs)
Default = 18h; Read/Write (address 22h)
Bit Number
Bit Name
Default
30
7
6
5
4
3
2
1
0
x27
x26
x25
x24
x23
x22
x21
x20
0
0
0
1
1
0
0
0
Bit
Mnemonic
7:0
x27-x20
Function
End value of second segment (x2[12:0]) of companding curve (LSBs). (See
Figure 13)
DS322PP1
CS7622
Compander X3 (MSBs)
Default = 0Bh; Read/Write (address 23h)
Bit Number
Bit Name
Default
7
6
5
RESERVED
0
Bit
Mnemonic
7:5
-
4:0
x312-x38
0
0
4
3
2
1
0
x312
x311
x310
x39
x38
0
1
0
1
1
Function
Reserved
End value of third segment of companding curve (MSBs). (See Figure 13)
Compander X3 (LSBs)
Default = 58h; Read/Write (address 24h)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
0
x37
x36
x35
x34
x33
x32
x31
x30
0
1
0
1
1
0
0
0
Bit
Mnemonic
7:0
x37-x30
Function
End value of third segment (x3[12:0]) of companding curve (LSBs). (See
Figure 13)
Device ID
Default = CCh; Read only (address 25h)
Bit Number
Bit Name
Default
7
6
5
4
3
2
1
0
device_ID7 device_ID6 device_ID5 device_ID4 device_ID3 device_ID2 device_ID1 device_ID0
1
Bit
Mnemonic
7:0
device_ID7-0
0
0
0
1
1
0
0
Function
This read-only register is the unique ID for the CS7622.
Revision Code
Default = 00h; Read only (address 26h)
Bit Number
Bit Name
Default
7
rev_code7
6
5
4
3
2
1
0
rev_code6
rev_code5
rev_code4
rev_code3
rev_code2
rev_code1
rev_code0
0
0
0
0
0
0
0
0
Bit
Mnemonic
7:0
rev_code7-0
DS322PP1
Function
This read-only register is the revision code for the CS7622.
31
CS7622
5.0 PIN DESCRIPTIONS
GNDD
DOUT6
VDDD
DOUT5
DOUT7
DOUT4
DOUT8
DOUT3
DOUT9
DOUT10
DOUT11
24
2
23
3
DOUT12
4
SCLK
5
SDATO
6
SDATI
SEN
TEST
DOUT2
32 31 30 29 28 27 26 25
1
CS7622
32-pin TQFP
Top View
7
8
22
21
CLKO
20
CK_DATA
19
CK_FT
18
9
10 11 12 13 14 15 16
DOUT1
DOUT0
17
CLAMP
RST
DIAG
BG_RES
REF_CAPP
AIN
GNDA
REF_CAPN
VDDA
Supply
VDDA - Supply for analog
Pin 13
3.3 V analog supply.
VDDD - Supply for digital
Pin 28
3.3 V or 2.7 V digital supply.
Ground
GNDA - Ground for analog
Pin 12
GNDA is supplied by VDDA.
GNDD - Ground for digital
Pin 27
Supplied by VDDD.
CMOS Input
CLAMP - Black level clamp signal
Pin 18
Provided by the external timing generator. Supplied by VDDD.
CK_FT - Clock in feed-through
Pin 19
Sampling clock for feed-through level.
CK_DATA - Clock in data
Pin 20
Sampling clock for data level. Supplied by VDDD
REF_CAPN - Reference capacitor- negative terminal
Pin 14
32
Supplied by VDDA. A 1 µF ceramic capacitor should be connected between
REF_CAPN and REF_CAPP.
DS322PP1
CS7622
REF_CAPP - Reference capacitor- positive terminal
Pin 15
Supplied by VDDA. A 1 µF ceramic capacitor should be connected between
REF_CAPN and REF_CAPP.
RST - Reset pin, negative true
Pin 17
May be connected to external power-on-reset-circuit. Supplied by VDDD.
SCLK - Serial bus clock signal
Pin 5
Supplied by VDDD.
SDATI - Serial bus data input signal
Pin 7
Supplied by VDDD.
SEN - Serial bus enable signal-chip select (active low)
Pin 8
Supplied by VDDD.
TEST - Test enable pin
Pin 9
Supplied by VDDD.
CMOS Analog Input
AIN - Video data input from CCD
Pin 11
Supplied by VDDA.
BG_RES - Band-gap resistor
Pin 10
Supplied by VDDA. A 10 kΩ resistor should be connected between BG_RES
and GNDA.
CMOS 4 mA Output
CLKO - Clock = output
Pin 21
Signal on this pin can either be the pixel clock output or data_valid signal output. Supplied by VDDD.
DOUT[0:12] - Digitized CCD data output
Pins 22-32, and 1-4
DOUT0 is LSB. Supplied by VDDD.
SDATO - Serial bus data output signal
Pin 6
DS322PP1
Supplied by VDDD.
33
CS7622
6.0 PACKAGE DIMENSIONS
32L TQFP PACKAGE DRAWING
E
E1
D D1
1
e
B
∝
A
A1
L
INCHES
MIN
MAX
--0.063
0.002
0.006
0.012
0.018
0.343
0.366
0.272
0.280
0.343
0.366
0.272
0.280
0.028
0.035
0.018
0.030
∝
0.000°
7.000°
* Nominal pin pitch is 0.50 mm
DIM
A
A1
B
D
D1
E
E1
e*
L
MILLIMETERS
MIN
MAX
--1.60
0.05
0.15
0.30
0.45
8.70
9.30
6.90
7.10
8.70
9.30
6.90
7.10
0.70
0.90
0.45
0.75
0.00°
7.00°
Controlling dimension is mm.
JEDEC Designation: MS026
34
DS322PP1
• Notes •