NSC LM9800CCV 8-bit greyscale/24-bit color linear ccd sensor processor Datasheet

LM9800
8-Bit Greyscale/24-Bit Color Linear CCD Sensor
Processor
General Description
Features
The LM9800 is a high performance integrated signal
processor/digitizer for linear CCD image scanners. The
LM9800 performs all the analog processing (correlated
double sampling for black level and offset compensation,
pixel-by-pixel gain adjust, and 8-bit analog-to-digital conversion) necessary to maximize the performance of a wide
range of linear CCD sensors.
The LM9800 can be digitally programmed to work with a
wide variety of CCDs from different manufacturers. An internal configuration register sets CCD and sampling timing to
maximize performance, simplifying the design and manufacturing processes.
The LM9800 can be used with sequential or parallel output
color CCDs. Coarse gain switching brings weak Blue signals
up to Red/Green levels for best performance. For complementary voltage reference see the LM4041.
n 2.5 million pixels/s conversion rate
n Implements Correlated Double Sampling for minimum
noise and offset error
n Pixel-to-pixel gain correction for individual pixels
maximizes dynamic range and resolution, even on
“weak” pixels
n Reference and signal sampling points digitally controlled
in 25ns increments for maximum performance
n Generates all necessary CCD clock signals
n Compatible with a wide range of linear CCDs
n TTL/CMOS input/output compatible
Key Specifications
n Resolution
8 Bits
n Pixel Conversion Rate
+5V ± 5%
n Supply Voltage
Applications
n Color and Greyscale Flatbed and Sheetfed Scanners
n General Purpose CCD Imaging
2.5MHz
n Supply Voltage (Digital I/O)
n Power Dissipation
+3.3V ± 10% or +5V ± 5%
210 mW (typ)
Connection Diagram
DS012498-1
TRI-STATE ® is a registered trademark of National Semiconductor Corporation.
MICROWIRE™ is a trademark of National Semiconductor Corporation.
© 1997 National Semiconductor Corporation
DS012498
www.national.com
LM9800 8-Bit Greyscale/24-Bit Color Linear CCD Sensor Processor
September 1997
Block Diagram
DS012498-3
Ordering Information
Commercial (10˚C ≤ TA ≤ +70˚C)
LM9800CCV
www.national.com
Package
V52A 52 Pin Plastic Leaded Chip Carrier
2
Absolute Maximum Ratings
Operating Ratings
(Notes 1,
2)
(Note 1) (Note 2)
Operating Temperature
Range
LM9800CCV, LM9800CCVF
VA Supply Voltage
VD Supply Voltage
VD(I/O) Supply Voltage
|VA–VD|
VA–VD(I/O)
OS, REF IN Voltage Range
CD0–CD6, MCLK, SYNC,
SDI, SCLK, CS , RD
Voltage Range
Positive Supply Voltage (V+ = VA = VD = VD(I/O))
With Respect to GND = AGND = DGND =
6.5V
DGND(I/O)
Voltage On Any Input or Output Pin
0.3V to V+ +0.3V
± 25 mA
Input Current at any pin (Note 3)
± 50 mA
Package Input Current (Note 3)
Package Dissipation at TA = 25˚C (Note 4)
ESD Susceptibility (Note 5)
Human Body Model
3000V
Soldering Information
Infrared, 10 seconds (Note 6)
300˚C
Storage Temperature
−65˚C to +150˚C
TMIN ≤ TA ≤ TMAX
10˚C ≤ TA ≤ +70˚C
+4.75V to +5.25V
+4.75V to +5.25V
+2.7V to +5.25V
≤100 mW
≥ −100 mV
−0.05V to VA + 0.05V
−0.05V to VD(I/O) + 0.05V
Electrical Characteristics
The following specifications apply for AGND = DGND = DGND(I/O) = 0V, VA = VD = +5.0VDC, VD(I/O) = +5.0V or +3.0VDC,
REF IN = +1.225VDC, fMCLK = 20MHz, Rs = 25Ω. Boldface limits apply for TA = TJ = TMIN to TMAX; all other limits TA = TJ
= 25˚C (Note 8)
Symbol
Parameter
Conditions
Typical
Limits
Units
(Note 9)
(Note 10)
(Limits)
CCD Source Requirements for Full Specified Accuracy and Dynamic Range (Note 12)
VWHITE
Maximum Peak CCD Differential
Signal Range
VGA Bypassed
VGA On (VGAIN = 2.0V)
Correctable Range of CCD Pixel-to-Pixel
VWHITE Variation
VRFT
CCD Reset Feed Through Amplitude
1.1
V (min)
0.5
V (min)
6
dB (max)
2
V (max)
ADC Characteristics
Resolution with No Missing Codes
8
Bits (min)
±1
±1
LSB (max)
Monotonicity
7
bits (min)
Gain Range
0
dB (min)
6
dB (max)
1.0
% (max)
±6
% (max)
ILE
Integral Linearity Error (Note 11)
DNL
Differential Non-Linearity
± 0.5
± 0.5
LSB (max)
PGA Characteristics
Gain Error at any gain
VGA Characteristics
VGA Gain Error vs. Formula
Coarse Offset Trim Characteristics
Offset DAC LSB Size
1
LSB
System Characteristics
Full Channel Offset Error (Adjusted)
Full Channel Gain Error
Offset DAC, Offset Add
Adjusted
PGA Gain = 0 dB
±1
LSB (max)
−5
% (min)
+10
% (max)
20
nA (max)
Reference and Analog Input Characteristics (Note 7)
OS Input Capacitance
OS Input Leakage Current
RREF
REF IN
5
Measured with OS = 2.45VDC
ADC Reference Ladder (REF OUTHI
to REF IN) Impedance
2
1000
Reference Voltage (Note 12)
1.225
3
pF
500
Ω (min)
2000
Ω (max)
1.19
V (min)
1.26
V (max)
www.national.com
DC and Logic Electrical Characteristics
The following specifications apply for AGND = DGND = DGND(I/O) = 0V, VA = VD = +5.0VDC, VD(I/O) = +5.0 or +3.0VDC, REF
IN = +1.225VDC, fMCLK = 20MHz, Rs = 25Ω. Boldface limits apply for TA = TJ = TMIN to TMAX; all other limits TA = TJ =
25˚C.
Symbol
Parameter
Conditions
Typical
Limits
Units
(Note 9)
(Note 10)
(Limits)
2.0
V (min)
CD0–CD6, MCLK, SYNC, SDI, SCLK, CS , RD Digital Input Characteristics
VIN(1)
Logical “1” Input Voltage
VD(I/O) = 5.25V
VIN(0)
Logical “0” Input Voltage
VD(I/O) = 3.6V
VD(I/O) = 4.75V
IIN
Input Leakage Current
VD(I/O) = 2.7V
VIN = VD
0.1
µA
VIN = DGND
−0.1
µA
5
pF
CIN
Input Capacitance
DD0–DD7, EOC , CCLK, SDO Digital Output Characteristics
VOUT(1)
Logical “1” Output Voltage
VD(I/O) = 4.75V, IOUT = −360 µA
VD(I/O) = 4.75V, IOUT = −10 µA
VD(I/O) = 2.7V, IOUT = −360 µA
VOUT(0)
IOUT
Logical “0” Output Voltage
TRI-STATE ® Output Current
(DD0–DD7 only)
COUT
VD(I/O) = 2.7V, IOUT = −10 µA
VD(I/O) = 5.25V, IOUT = 1.6 mA
VD(I/O) = 3.6V, IOUT = 1.6 mA
VOUT = DGND
VOUT = VD
2.0
V (min)
0.8
V (max)
0.7
V (max)
2.4
V (min)
4.4
V (min)
2.1
V (min)
2.5
V (min)
0.4
V (max)
0.4
V (max)
0.1
µA
−0.1
µA
5
pF
TRI-STATE Output Capacitance
φ1, φ2, RS, TR Digital Output Characteristics
VOUT(1)
VOUT(0)
Logical “1” Output Voltage
Logical “0” Output Voltage
VD = 4.75V, IOUT = −360 µA
VD = 4.75V, IOUT = −10 µA
VD = 5.25V, IOUT = 1.6 mA
2.4
V (min)
4.4
V (min)
0.4
V (max)
42
mA (max)
10
mA (max)
3.5
6
mA (max)
1.6
2
mA (max)
Power Supply Characteristics
IA
Analog Supply Current
ID
Digital Supply Current
ID(I/O)
Digital I/O Supply Current
Operating
32
Standby
20
Operating
6
Standby
µA
4
Operating, VD(I/O) = 5.0V
Operating, VD(I/O) = 3.0V
Standby, VD(I/O) = 5.0V or 3.0V
mA
0.5
mA
AC Electrical Characteristics, MCLK Independent
The following specifications apply for AGND = DGND = DGND(I/O) = 0V, VA = VD = VD(I/O) = +5.0VDC, REF IN = +1.225VDC,
fMCLK = 20MHz, tMCLK = 1/fMCLK, tr = tf = 5ns, Rs = 25Ω, CL (databus loading) = 50 pF/pin. Boldface limits apply for TA =
TJ = TMIN to TMAX; all other limits TA = TJ = 25˚C.
Symbol
fMCLK
Parameter
Conditions
Typical
Limits
Units
(Note 9)
(Note 10)
(Limits)
Maximum MCLK Frequency
20
MHz (min)
Minimum MCLK Frequency
5
MHz (max)
MCLK Duty Cycle
30
40
% (min)
70
60
% (max)
ns (min)
tA
SYNC setup of MCLK
5
10
tCDSETUP
Correction Data valid to CLK Setup
14
20
ns (min)
tCDHOLD
Correction Data valid to CLK Hold
−12
0
ns (min)
tD1H, tD0H
RD High to DD0–DD7 TRI-STATE
7
15
ns (max)
www.national.com
4
AC Electrical Characteristics, MCLK Independent
(Continued)
The following specifications apply for AGND = DGND = DGND(I/O) = 0V, VA = VD = VD(I/O) = +5.0VDC, REF IN = +1.225VDC,
fMCLK = 20MHz, tMCLK = 1/fMCLK, tr = tf = 5ns, Rs = 25Ω, CL (databus loading) = 50 pF/pin. Boldface limits apply for TA =
TJ = TMIN to TMAX; all other limits TA = TJ = 25˚C.
Symbol
Parameter
Conditions
Typical
Limits
Units
(Note 9)
(Note 10)
(Limits)
20
40
ns (max)
MHz (max)
tDACC
Access Time Delay from RD Low to
DD0–DD7 Data Valid
fSCLK
Maximum SCLK Frequency
2.5
SCLK Duty Cycle
40
% (min)
60
% (max)
tSDI
SDI Set-Up Time from SCLK
Rising Edge
3
10
ns (min)
tHDI
SDI Hold Time from SCLK
Rising Edge
2
15
ns (min)
tDDO
Delay from SCLK Falling Edge to
SDO Data Valid
35
55
ns (max)
21
55
ns (max)
5
ns (min)
tHDO
SDO Hold Time from SCLK
RL = 3K, CL = 25 pF
Falling Edge
tDELAY
DELAY from SCLK Falling Edge to
CS Rising or Falling Edge
2
5
ns (min)
tSETUP
Set-Up Time of CS Rising or Falling
Edge to SCLK Rising Edge
−2
10
ns (min)
tSACC
Access Time Delay from CS Falling
Edge to SDO Data Valid
20
40
ns (max)
tS1H, tS0H
Delay from CS Rising Edge to SDO
TRI-STATE
RL = 3K, CL = 50 pF
40
100
ns (max)
SDO Rise Time, TRI-STATE to High
RL = 3K, CL = 50 pF
tRDO
SDO Rise Time, Low to High
tFDO
SDO Fall Time, TRI-STATE to Low
RL = 3K, CL = 50 pF
SDO Fall Time, High to Low
5
20
ns
20
ns
20
ns
20
ns
www.national.com
AC Electrical Characteristics, MCLK Dependent
The following specifications apply for AGND = DGND = DGND(I/O) = 0V, VA = VD = VD(I/O) = +5.0 VDC, REF IN = +1.225
VDC, fMCLK = 20MHz, tMCLK = 1/fMCLK, tr = tf = 5ns, Rs = 25Ω, CL (databus loading) = 50 pF/pin. Refer to Table 2 : Configuration Register Parameters for limits labelled C.R. Boldface limits apply for TA = TJ = TMIN to TMAX; all other limits TA =
TJ = 25˚C.
Symbol
Parameter
tSTART
MCLK to first φ1 high
tφ
φ1, φ2 Clock Period
Conditions
Typical
Limits
Units
(Note 9)
(Note 10)
(Limits)
tMCLK
50ns
1
Standard CCD Mode
400ns
8
tMCLK
Even/Odd CCD Mode
800ns
16
tMCLK
tTRWIDTH
Transfer Pulse (TR) Width
C.R.
µs
tGUARD
φ1 to TR, TR to φ1 Guardband
C.R.
ns
tRSWIDTH
Reset Pulse (RS) Width
C.R.
ns
tRS
Falling Edge of φ1 to RS
Standard CCD Mode
C.R.
ns
Either Edge of φ1 to RS
Even/Odd CCD Mode
C.R.
ns
C.R.
ns
50ns
1
tMCLK
100ns
2
tMCLK (min)
2
tMCLK (max)
tS/HREF
tS/HSIG
tS/HWIDTH
Falling Edge of φ1 to Ref. Sample
Standard CCD Mode
Either Edge of φ1 to Ref. Sample
Even/Odd CCD Mode
Falling Edge of φ1 to Sig. Sample
Standard CCD Mode
Either Edge of φ1 to Sig. Sample
Even/Odd CCD Mode
Sample Pulse Width
(Acquisition Time)
tSYNCLOW
SYNC Low Between Lines
tB
SYNC Setup of φ1 to end line
tCCLKWIDTH
CCLK Pulse Width
tDATAVALID
Data Valid Time from EOC Low
tEOCWIDTH
EOC Pulse Width
φ1 and φ2 Frequency
250ns
5
tMCLK
300
ns (min)
250ns
5
tMCLK
Standard CCD Mode
2.5MHz
fMCLK/8
Hz
Even/Odd CCD Mode
1.25MHz
fMCLK/16
Hz
50
%
φ1 and φ2 Duty Cycle
Electrical Characteristics (Notes)
Note 1: Absolute Maximum Ratings indicate limits beyond which damage to the device may occur. Operating Ratings indicate conditions for which the device is
functional, but do not guarantee specific performance limits. For guaranteed specifications and test conditions, see the Electrical Characteristics. The guaranteed
specifications apply only for the test conditions listed. Some performance characteristics may degrade when the device is not operated under the listed test conditions.
Note 2: All voltages are measured with respect to GND = AGND = DGND = DGND(I/O) = 0V, unless otherwise specified.
Note 3: When the input voltage (VIN) at any pin exceeds the power supplies (VIN < GND or VIN > VA or VD), the current at that pin should be limited to 25 mA. The
50 mA maximum package input current rating limits the number of pins that can simultaneously safely exceed the power supplies with an input current of 25 mA to
two.
Note 4: The maximum power dissipation must be derated at elevated temperatures and is dictated by TJmax, θJA and the ambient temperature, TA. The maximum
allowable power dissipation at any temperature is PD = (TJmax–TA)/θJA. TJmax = 150˚C for this device. The typical thermal resistance θJA of this part when board
mounted is 52˚C/W for the V52A PLCC package.
Note 5: Human body model, 100 pF capacitor discharged through a 1.5Ω resistor.
Note 6: See AN450 “Surface Mounting Methods and Their Effect on Product Reliability” or the section titled “Surface Mount” found in any post 1986 National Semiconductor Linear Data Book for other methods of soldering surface mount devices.
Note 7: A Zener diode clamps the OS analog input to AGND as shown below. This input protection, in combination with the external clamp capacitor and the output
impedance of the CCD, prevents damage to the LM9800 from transients during power-up.
DS012498-4
Note 8: To guarantee accuracy, it is required that VA and VD be connected together to the same power supply with separate bypass capacitors at each supply pin.
Note 9: Typicals are at TJ = TA = 25˚C, fMCLK = 20MHz, and represent most likely parametric norm.
Note 10: Tested limits are guaranteed to National’s AOQL (Average Outgoing Quality Level).
www.national.com
6
Electrical Characteristics (Notes)
(Continued)
Note 11: Integral linearity error is defined as the deviation of the analog value, expressed in LSBs, from the straight line that best fits the actual transfer function
of the ADC.
Note 12: VREF is defined as the CCD OS voltage for the reference period following the reset feedthrough pulse. VWHITE is defined as the peak CCD pixel output
voltage for a white (full scale) image with respect to the reference level, VREF. VRFT is defined as the peak positive deviation above VREF of the reset feedthrough
pulse. The maximum correctable range of pixel-to-pixel VWHITE variation is defined as the maximum variation in VWHITE (due to PRNU, light source intensity variation, optics, etc.) that the LM9800 can correct for using its internal PGA.
DS012498-5
Note 13: Reference voltages below 1.19V may decrease SNR. Reference voltages above 1.26V may cause clipping errors inside the LM9800. The
LM4041EIM3-1.2 (SOT-23 package) or the LM4041EIZ-1.2 (TO-92 package) bandgap voltage references are recommended for this application.
Typical Performance Characteristics
φ1, φ2, RS, TR Rise and Fall Times Through a
Series Resistance vs. Load Capacitance
VGA Gain vs Voltage on GAIN Input
DS012498-48
DS012498-49
Pin Descriptions
REF IN
Analog Inputs. These two pins are the
system reference voltage inputs and
should be tied together to a 1.225V
voltage source and bypassed to AGND
with a 0.1 µF monolithic capacitor.
REF OUTHI
Analog Output. This reference voltage is
developed internally by the LM9800, and
is equal to 3 times REF IN. It should be
bypassed to AGND with a 0.1 µF
monolithic capacitor.
REF OUTMID
Analog Output. This reference voltage is
developed internally by the LM9800, and
is equal to 2 times REF IN. It should be
bypassed to AGND using a 0.1 µF
monolithic capacitor. This pin can
source up to 250 µA when used to
power a voltage divider for setting the
voltage at the GAIN input.
ISET
Analog Input. This input is used to set
internal bias currents inside the LM9800.
It should be connected to VA through a
75 kΩ resistor.
CCD Driver Signals
φ1
Digital Output. CCD clock signal, phase
1.
φ2
Digital Output. CCD clock signal, phase
2.
RS
Digital Output. Reset pulse for the CCD.
TR
Digital Output. Transfer pulse for the
CCD.
Analog I/O
OS
Analog Input. This is the OS (Output
Signal) from the CCD. The maximum
peak signal that can be accurately
digitized is equal to the voltage at REF
IN, typically 1.225V.
GAIN
Analog Input. The voltage on this pin
determines the gain of the VGA when
the VGA is enabled. This input should
be bypassed to AGND using a 0.1 µF
monolithic capacitor.
7
www.national.com
Pin Descriptions
VTEST1,
VTEST2
(Continued)
Digital Output I/O
Analog Inputs/Outputs. These pins are
used for testing the device during
manufacture and should be left
unconnected.
DD0 (LSB)–
DD7 (MSB)
Digital Outputs. Pixel Output Databus.
This data bus outputs the 8-bit digital
output data during line scan.
EOC
Digital Output. This is the End of
Conversion signal from the ADC
indicating that new pixel data is
available.
RD
Digital Input. Taking this input low
places the data stored in the output
latch on the bus. When this input is high
the DD0–DD7 bus is in TRI-STATE.
VA
This is the positive supply pin for the
analog supply. It should be connected to
a voltage source of +5V and bypassed
to AGND with a 0.1 µF monolithic
capacitor in parallel with a 10 µF
tantalum capacitor.
AGND
This is the ground return for the analog
supply.
VD
This is the positive supply pin for the
digital supply. It should be connected to
a voltage source of +5V and bypassed
to DGND with a 0.1 µF monolithic
capacitor.
DGND
This is the ground return for the digital
supply.
VD(I/O)
This is the positive supply pin for the
digital supply for the LM9800’s I/O. It
should be connected to a voltage source
of +3V to +5V and bypassed to
DGND(I/O) with a 0.1 µF monolithic
capacitor. If the supply for this pin is
different than the supply for VA and VD,
it should also be bypassed with a 10 µF
tantalum capacitor.
DGND(I/O)
This is the ground return for the digital
supply for the LM9800’s I/O.
Configuration Register I/O
SDI
Digital Input. Serial Data Input pin.
SDO
Digital Output. Serial Data Output pin.
SCLK
Digital Input. This is the serial data
clock, used to clock data in through SDI
and out through SDO. SCLK is
asynchronous to MCLK. Input data is
latched and output data is changed on
the rising edge of SCLK.
CS
Analog Power
Digital Input. This is the Chip Select
signal for writing to the Configuration
Register through the serial interface.
This input must be low in order to
communicate with the Configuration
Register. This pin is used for serial I/O
only–it has no effect on any other
section of the chip. Note: The SYNC
pin must be high to read or write
from the Configuration register.
Digital Power
General Digital I/O
MCLK
Digital Input. This is the 20 MHz (typical)
master system clock.
SYNC
Digital Input. A low-to-high transition on
this input begins a line scan operation.
The line scan operation terminates when
this input is taken low. A low-to-high
transition on this input will also reset the
serial I/O port to the Configuration
Register. The SYNC pin must be high to
read or write from the Configuration
register.
Digital Coefficient I/O
CD0 (LSB)–
CD6 (MSB)
Digital Inputs. Correction Coefficient
Databus. This is the 7-bit data path for
the gain adjust PGA, used during line
scan.
CCLK
Digital Output. This is the signal that is
used to clock the Gain coefficients into
the LM9800. Data is latched on the
rising edge of CCLK.
www.national.com
8
Timing Diagrams
DS012498-6
FIGURE 1. Line Scan Timing Overview
DS012498-7
FIGURE 2. Pixel Pipeline Timing Overview
9
www.national.com
Timing Diagrams
(Continued)
DS012498-8
FIGURE 3. Timing for Start of Line Scan
DS012498-9
FIGURE 4. Timing for End of Line/Start of Next Line
DS012498-10
FIGURE 5. TR Pulse
DS012498-11
FIGURE 6. RS Pulse Polarity
www.national.com
10
Timing Diagrams
(Continued)
DS012498-12
Note: Clamp signal only active during optical black pixels at beginning of line.
FIGURE 7. CCD Timing
DS012498-13
Note: Clamp signal only active during optical black pixels at beginning of line.
FIGURE 8. CCD Timing (Even/Odd CCDs)
11
www.national.com
Timing Diagrams
(Continued)
DS012498-14
Note: j = value programmed in Optical Black Register.
i = value programmed in Dummy Pixel Register - 1 (for example: dummy pixel register = 1 ≥ i = 0 ≥ no dummy pixels).
FIGURE 9. Dummy Pixel and Optical Black Pixel Timing
DS012498-15
FIGURE 10. Coefficient Data Timing
DS012498-16
FIGURE 11. Output Data Timing
www.national.com
12
Timing Diagrams
(Continued)
DS012498-17
FIGURE 12. Data Timing (Output and Coefficient Data Sharing Same Bus)
DS012498-18
Note: VGA operation (on or off) for “a” pixels, “b” pixels, and “c” pixels is set in the configuration register.
VGA is always off for first “a” pixel immediately following last optical black pixel.
FIGURE 13. VGA Gain Switching
13
www.national.com
Serial Configuration Register Timing Diagrams
Note: SYNC pin must be held HIGH (+5V) to
read from or write to configuration register.
DS012498-19
FIGURE 14. Configuration Register Write Timing with CS Continuously Low (16-bit Word)
DS012498-20
FIGURE 15. Configuration Register Write Timing with CS Continuously Low (Two 8-bit bytes)
DS012498-21
FIGURE 16. Configuration Register Read Timing with CS Continuously Low (16-bit Word)
DS012498-22
FIGURE 17. Configuration Register Read Timing with CS Continuously Low (Two 8-bit bytes)
SDO Falling and Rising Edge
SDO “TRI-STATE” Falling and Rising Edge
DS012498-23
DS012498-24
FIGURE 18. SDO Timing
www.national.com
14
Serial Configuration Register Timing Diagrams
Note: SYNC pin must be held HIGH (+5V) to
read from or write to configuration register. (Continued)
DS012498-25
FIGURE 19. Configuration Register Write Timing using CS , Continuous SCLK (16-bit Word)
DS012498-26
FIGURE 20. Configuration Register Write Timing using CS , Continuous SCLK (Two 8-bit bytes)
DS012498-27
FIGURE 21. Configuration Register Read Timing using CS , Continuous SCLK (16-bit Word)
DS012498-28
FIGURE 22. Configuration Register Read Timing using CS , Continuous SCLK (Two 8-bit bytes)
15
www.national.com
EOC and CCLK Start of Line Timing Diagrams
(See Section 2.4)
RS Even
RS Odd
tEOC1
tEOC2
tCCLK1
tCCLK2
(MCLKs)
(MCLKs)
(MCLKs)
(MCLKs)
SR = 0
SR = 1 or 2
SR = 3 or 4
SR = 5 or 6
SR = 0 or 1
SR = 2 or 3
SR = 4 or 5
SR = 6 or 7
2
5
4
3
3
5
5
3
3
6
6
3
3
7
5
5
SR = 7 or 8
SR = 9 or 10
SR = 8 or 9
SR = 10 or 11
3
8
5
6
3
9
5
7
DS012498-29
RS Even, SR = 0 to 10: tEOC0 = 2 MCLKs, tCCLK0 = 3 MCLKs
RS Odd, SR = 0 to 11: tEOC0 = 2.5 MCLKs, tCCLK0) = 3.5 MCLKs
FIGURE 23. SR = 0 through 10 (RS Even), SR = 0 through 11 (RS Odd)
DS012498-30
RS Even, SR = 11 or 12: tEOC0 = 2 MCLKs, tCCLK0 = 3 MCLKs
RS Odd, SR = 12 or 13: tEOC0 = 2.5 MCLKs, tCCLK0 = 3.5 MCLKs
FIGURE 24. SR = 11 or 12 (RS Even), SR = 12 or 13 (RS Odd)
www.national.com
16
EOC and CCLK Start of Line Timing Diagrams
(Continued)
DS012498-31
RS Even, SR = 13 or 14: tEOC0 = 2 MCLKs, tCCLK0 = 3 MCLKs
RS Odd, SR = 14 or 15: tEOC0 = 2.5 MCLKs, tCCLK0 = 3.5 MCLKs
FIGURE 25. SR = 13 or 14 (RS Even), SR = 14 or 15 (RS Odd)
DS012498-32
FIGURE 26. SR = 15 (RS Even)
17
www.national.com
Configuration Register
tMCLK = 1/fMCLK = 1 MCLK period. Examples given in parenthesis are for fMCLK =
20MHz (tMCLK = 50ns).
TABLE 1. Configuration Register Address Table
A2
A1
A0
D7
D6
D5
D4
D3
D2
D1
D0
Standard
RS
Pulse
Width
Mode or
0
0
0
Even/Odd
RS
Pulse
Polarity
RS Pulse Position
(Minimum Value is 1)
Mode
MODE
0
0
0
1
1
1
1
0
1
1
0
0
1
1
1
0
1
0
1
0
1
www.national.com
RSW1
RSW0
RSPOL
RSPOS3
RSPOS2
Sample Reference Position
RSPOS1
RSPOS0
Sample Signal Position
SR3
SR2
SR1
SR0
φ1
φ2
RS
TR
Enable
Enable
Enable
Enable
φ1EN
φ2EN
RSEN
TREN
SS3
SS2
SS1
TR
Guardband
Polarity
TRW0
TRGRD
TRPOL
BLS2
BLS1
BLS0
BLL1
BLL0
TR Pulse Width
TRW1
SS0
TR–φ1
Dummy Pixels (Minimum Value is 1)
BLS7
BLS6
BLS5
BLS4
BLS3
Optical Black Pixels (Minimum Value is 1)
BLL7
BLL6
BLL5
BLL4
BLL3
BLL2
PGA Gain Coefficient
0
GAIN6
GAIN5
GAIN4
GAIN3
GAIN2
GAIN1
GAIN0
PGA Gain
Test
Power-
Offset
VGA
VGA State
VGA State
VGA State
Source
Mode
down
Add
On/Off
Pixel c
Pixel b
Pixel a
PGASRC
0
PD
OFFADD
VGA
G3
G2
G1
Offset DAC
Offset DAC
Offset DAC
Offset DAC
Sign
MSB
ODSIGN
VOS2
Test Modes
LSB
VOS1
VOS0
18
0
0
0
0
Configuration Register
tMCLK = 1/fMCLK = 1 MCLK period. Examples given in parenthesis are for fMCLK =
20MHz (tMCLK = 50ns). (Continued)
TABLE 2. Configuration Register Parameters
Parameter
MODE
Control Bits
Result
MODE
Standard CCD (φ frequency = fMCLK/8)
Even/Odd CCD (φ frequency = fMCLK/16)
0
1
RS Pulse Width
(tRSWIDTH)
RS Pulse Polarity
RS1
RS0
0
0
1 tMCLK (50ns)
0
1
2 tMCLK (100ns)
1
0
3 tMCLK (150ns)
1
1
4 tMCLK (200ns)
RSPOL
0
RS
1
RS Pulse Position
(tRS)
Note: Minimum
Value is 1
Sample Reference
Position (tS/HREF)
RS
RSPOS3
RSPOS2
RSPOS1
RSPOS0
0
0
0
0
Not Valid
0
0
0
1
0.5tMCLK (25 ns)
0
0
1
0
1.0tMCLK (50ns)
0
0
1
1
1.5tMCLK (75ns)
0
1
0
0
2.0tMCLK (100ns)
0
1
0
1
2.5tMCLK (125ns)
0
1
1
0
3.0tMCLK (150ns)
0
1
1
1
3.5tMCLK (175ns)
1
0
0
0
4.0tMCLK (200ns)
1
0
0
1
4.5tMCLK (225ns)
1
0
1
0
5.0tMCLK (250ns)
1
0
1
1
5.5tMCLK (275ns)
1
1
0
0
6.0tMCLK (300 ns)
1
1
0
1
6.5tMCLK (325 ns)
1
1
1
0
7.0tMCLK (350 ns)
1
1
1
1
7.5tMCLK (375 ns)
SR3
SR2
SR1
SR0
0
0
0
0
0.0tMCLK (0 ns)
0
0
0
1
0.5tMCLK (25 ns)
0
0
1
0
1.0tMCLK (50 ns)
0
0
1
1
1.5tMCLK (75 ns)
0
1
0
0
2.0tMCLK (100 ns)
0
1
0
1
2.5tMCLK (125 ns)
0
1
1
0
3.0tMCLK (150 ns)
0
1
1
1
3.5tMCLK (175 ns)
1
0
0
0
4.0tMCLK (200 ns)
1
0
0
1
4.5tMCLK (225 ns)
1
0
1
0
5.0tMCLK (250 ns)
1
0
1
1
5.5tMCLK (275 ns)
1
1
0
0
6.0tMCLK (300 ns)
1
1
0
1
6.5tMCLK (325 ns)
1
1
1
0
7.0tMCLK (350 ns)
1
1
1
1
7.5tMCLK (375 ns)
19
www.national.com
Configuration Register
tMCLK = 1/fMCLK = 1 MCLK period. Examples given in parenthesis are for fMCLK =
20MHz (tMCLK = 50ns). (Continued)
TABLE 2. Configuration Register Parameters (Continued)
Parameter
Sample Signal
Position (tS/HSIG)
φ1 Enable
φ2 Enable
RS Enable
TR Enable
Control Bits
SS3
SS2
0
0
0
0
0.0tMCLK (0 ns)
0
0
1
0.5tMCLK (25 ns)
0
0
1
0
1.0tMCLK (50 ns)
0
0
1
1
1.5tMCLK (75 ns)
0
1
0
0
2.0tMCLK (100 ns)
0
1
0
1
2.5tMCLK (125 ns)
0
1
1
0
3.0tMCLK (150 ns)
0
1
1
1
3.5tMCLK (175 ns)
1
0
0
0
4.0tMCLK (200 ns)
1
0
0
1
4.5tMCLK (225 ns)
1
0
1
0
5.0tMCLK (250 ns)
1
0
1
1
5.5tMCLK (275 ns)
1
1
0
0
6.0tMCLK (300 ns)
1
1
0
1
6.5tMCLK (325 ns)
1
1
1
0
7.0tMCLK (350 ns)
1
1
1
1
7.5tMCLK (375 ns)
φ1EN
0
φ1 Output Off
1
φ1 Output On
φ2EN
0
φ2 Output Off
1
φ2 Output On
RSEN
0
RS Output Off
1
RS Output On
TREN
TR Output Off
1
TR-φ1
SS0
0
0
TR Pulse Width
(tTRWIDTH)
Result
SS1
TRW1
TR Output On
TRW0
0
0
20 tMCLK (1.0 µs)
0
1
30 tMCLK (1.5 µs)
1
0
40 tMCLK (2.0 µs)
1
1
50 tMCLK (2.5 µs)
TRGRD
Guardband
0
1 tMCLK (50 ns)
(tGUARD)
1
2 tMCLK (100 ns)
TR Polarity
www.national.com
TRPOL
0
TR
1
TR
20
Configuration Register
tMCLK = 1/fMCLK = 1 MCLK period. Examples given in parenthesis are for fMCLK =
20MHz (tMCLK = 50ns). (Continued)
TABLE 2. Configuration Register Parameters (Continued)
Parameter
Dummy Pixels
Note: Minimum
Value is 1. Actual
number of dummy
pixels in CCD should
be one less than
number in this
register.
Optical Black Pixels
Note: Minimum
Value is 1
Internal PGA Gain
Coefficient
PGA Gain
Coefficient
Source
Power Down
Offset Add
Master VGA Control
VGA State for
Pixel “c”
VGA State for
Pixel “b”
Control Bits
Result
BLS7
BLS6
BLS5
BLS4
BLS3
BLS2
BLS1
BLS0
0
0
0
0
0
0
0
0
Not Valid
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
1
•
•
•
1
1
BLL7
0
0
0
•
•
•
1
1
BLL6
0
0
0
•
•
•
1
1
BLL5
0
0
0
•
•
•
1
1
BLL4
0
0
0
•
•
•
1
1
BLL3
0
0
0
•
•
•
1
1
BLL2
0
0
0
•
•
•
1
1
BLL1
0
0
1
•
•
•
0
1
BLL0
0
1
0
•
•
•
253
254
•
•
•
1
1
GAIN6
0
0
•
•
•
1
1
GAIN5
0
0
•
•
•
1
1
GAIN4
0
0
•
•
•
1
1
GAIN3
0
0
•
•
•
1
1
GAIN2
0
0
•
•
•
1
1
GAIN1
0
0
•
•
•
1
1
GAIN0
0
1
•
•
•
0
1
•
•
•
1
1
•
•
•
1
1
•
•
•
1
1
•
•
•
1
1
•
•
•
0
1
•
•
•
•
•
•
1
1
1
1
PGASRC
0
1
PD
0
1
OFF ADD
0
1
VGA
0
1
VGAc
0
1
VGAb
0
1
Not Valid
1
2
•
•
•
254
255
0 [0 dB]
1
•
•
•
126
127 [6 dB]
Internal
External
Operating
Power Down
Offset z0 LSB
Offset z+1 LSB
VGA Disabled
VGA Enabled
VGA bypassed
VGA on
VGA bypassed
VGA on
21
www.national.com
Configuration Register
tMCLK = 1/fMCLK = 1 MCLK period. Examples given in parenthesis are for fMCLK =
20MHz (tMCLK = 50ns). (Continued)
TABLE 2. Configuration Register Parameters (Continued)
Parameter
VGA State for
Pixel “a”
Control Bits
Result
VGAa
0
VGA bypassed
1
Offset DAC
VGA on
Sign
VOS1
VOS1
VOS0
Offset (LSB)
1
0
0
0
0
1
0
0
1
+1
1
0
1
0
+2
1
0
1
1
+3
1
1
0
0
+4
1
1
0
1
+5
1
1
1
0
+6
1
1
1
1
+7
0
0
0
0
0
0
0
0
1
−1
0
0
1
0
−2
0
0
1
1
−3
0
1
0
0
−4
0
1
0
1
−5
0
1
1
0
−6
0
1
1
1
−7
Block Diagram of LM9800-Based System
DS012498-33
Note: Power supplies and bypass capacitors not shown for clarity.
FIGURE 27. LM9800 System Block Diagram
www.national.com
22
Finally, the output of the PGA is digitized by the ADC and
made available on the DD0–DD7 bus. For a detailed explanation of the ADC, see section 4.9.
Three reference voltages are used throughout the signal
path: the externally supplied REF IN (1.225V), and the internally generated REF OUTMID (2.45V) and REF OUTHI
(3.675V).
Applications Information
1.0 THEORY OF OPERATION
The LM9800 removes errors from and digitizes a linear CCD
pixel stream, while providing all the necessary clock signals
to drive the CCD. Offset and gain errors are removed at the
pixel rate, for individual pixels. Offset errors are removed
through correlated double sampling (CDS). Gain errors
(which may come from any combination of PRNU, uneven illumination, cos4 effect, RGB filter mismatch, etc.) are removed through the use of a 7-bit PGA in front of the ADC.
1.2 The CCD Clocking Signals
To maximize the flexibility of the LM9800, the CCD’s φ1, φ2,
RS, and TR pulses are internally generated, with a wide
range of options, compatible with most commercial linear
CCDs. In most cases, these output signals can drive most
CCD clock inputs directly, with only series resistors (for slew
rate control) between the LM9800’s outputs and the CCD
clock inputs.
1.1 The Analog Signal Path
(See Block Diagram)
The analog output signal from the CCD is connected to the
OS Input of the LM9800 through a 0.001µF (typical, see section 4.2, Clamp Capacitor Selection ) DC blocking capacitor.
During the CCD’s optical black pixel segment at the beginning of every line, this input is clamped to the REF OUTMID
voltage (approximately 2.45V). This DC restore operation
fixes the reference level of the CCD pixel stream at REF
OUTMID.
The signal is then buffered and fed to a voltage-controlled
VGA (variable gain amplifier). The VGA can be used to compensate for peak white CCD outputs less than the 1.225V
full-scale required by the LM9800 for maximum dynamic
range. It can also be used to increase the gain of the Blue
signal in a sequential-output RGB CCD, since the VGA can
be switched in and out of the circuit at the pixel rate. When
used with parallel output CCDs the VGA can be used (in conjunction with an external multiplexer) to fine-tune the amplitude of the red, green, and blue signals. For a detailed explanation of the VGA, see section 4.3, GAIN (VGA) Input.
The output of the VGA goes into the CDS (Correlated Double
Sampling) stage, consisting of two sample/hold amplifiers:
S/H Ref (Reference) and S/H Signal. The Reference Level is
sampled and held by the S/H Ref circuit and the active pixel
data is sampled and held by the S/H Signal circuit. The output of S/H Ref is subtracted from the S/H Signal output and
amplified by 2. The full-scale signal range at this point is approximately 2.45Vp-p. CDS reduces or eliminates many
sources of noise, including reset noise, flicker noise, and
both high and low frequency pixel-to-pixel offset variation.
For more information on the CDS stage, see section 4.5,
Correlated Double Sampler (CDS) .
At this point an offset voltage can be injected by the 4-bit Offset DAC. This voltage is designed to compensate for any
small fixed DC offset introduced by the CDS S/Hs and the x2
amplifier. The LSB size of the DAC is approximately 1 LSB
(10mV).The adjustment range is ± 7 LSBs. For a detailed explanation of the Offset DAC, see section 4.6.
1.3 The Digital Interface
There are three main sections to the digital interface of the
LM9800: a serial interface to the Configuration Register,
where all device programming is done, a 7 bit-wide input
databus for gain correction coefficients with a synchronous
clock output, and an 8-bit output databus for the final pixel
output data with a synchronous EOC output signal and a RD
input. Please note that CS affects only the serial I/O–it has
no effect on the output databus, input coefficient bus, or any
other section of the LM9800.
2.0 DIGITAL INTERFACE
2.1 Reading and Writing to the Configuration Register
Communication with the Configuration Register is done
through a standard MICROWIRE™ serial interface. This interface is compatible with the Motorola SPI standard and is
simple enough to easily be implemented in custom hardware
if needed.
The serial interface timing is shown in Figures 14, 15, 16, 17
and Figures 19, 20, 21, 22. The SYNC pin must be pulled
high to read or write to the configuration register. Taking
SYNC high resets the internal serial counters. Data is sent
serially, LSB first. Input data is latched on the rising edge of
SCLK, and output data changes on the falling edge of SCLK.
CS must be low to enable serial I/O.
If SCLK is only clocked when sending or receiving data from
the LM9800, and held low at all other times, then CS can be
tied low permanently as shown in Figures 14, 15, 16, 17. If
SCLK is continuous, then CS is used to determine the beginning and the end of a serial byte or word (see Figures 19, 20,
21, 22). Note that CS must make its high-to-low and
low-to-high transitions when SCLK is low, otherwise the internal bit counter may receive an erroneous pulse, causing
an error in the write or read operation.
The next stage is the PGA. This is a programmable gain amplifier that changes the gain at the pixel rate to correct for
gain errors due to PRNU, uneven illumination (such as cos4
effect), RGB filter mismatch, etc. The gain adjustment range
is 0 to 6 dB (x1 to x2) with 7 bits of resolution. The gain data
(correction coefficients) is provided on the CD0–CD6 bus.
The gain may also be fixed at any value between 0 dB and
6 dB with the PGA Gain Coefficient configuration register.
For further information on the PGA, see section 4.7.
An approximately 1 LSB (10mV) offset can be added at the
output of the PGA stage if necessary to ensure that the offset
is zero or positive. This eliminates the possibility of a negative offset clipping the darkest output pixels. For more information on the Offset Add Bit, see section 4.8.
Data may be transmitted and received in two 8-bit bytes
(typical with microcontroller interfaces) or one 16-bit word
(for custom serial controllers).
The Configuration Register is programmed by sending a
control byte to the serial port. This byte indicates whether
this is a read or a write operation, and gives the 3-bit address
of the register bank to be read from or written to. If this is a
read operation, the next 8 SCLKs will output the data at the
requested location on the SDO pin. If this is a write operation, the data to be sent to the specified location should be
clocked in on the SDI input during the next 8 SCLKs. Data is
sent and received using the LSB (Least Significant Bit) first
format.
23
www.national.com
Applications Information
2.6 SYNC
This input signals the beginning of a line. When SYNC goes
high, the LM9800 generates a TR pulse, then begins converting pixels until the SYNC line is brought low again. Since
there is no pixel counter in the LM9800, it will work with
CCDs of any length.
SYNC must be high to read from or write to the Configuration
Register. The rising edge of SYNC resets the serial I/O’s internal shift register, so any noise or sporadic signals on the
SCLK input prior to SYNC going high will be ignored.
(Continued)
For maximum system reliability, each configuration register
location can be read back and verified after a write.
If the serial I/O to the configuration register falls out of sync
for any reason, it can be reset by a rising edge on the SYNC
pin input.
2.2 Writing Correction Coefficient Data on the
CD0–CD6 Bus
Correction coefficient data for each pixel is latched on the
rising edge of the CCLK output signal (see Figure 10). Note
that there is a 4 pixel latency between when the coefficient
data is latched and when the output data is available. As Figure 2, Pixel Pipeline Timing Overview shows, coefficient
data for pixel n is latched shortly before the output data for
pixel n-2 becomes available on the output databus
(DD0–DD7). Figures 23, 24, 25, 26 show the timing for coefficient data and output data at the beginning of each line.
Note that there is no way to provide a correction coefficient
for pixel 1, the first pixel in the CCD array. This is usually not
a problem since the first several pixels of most CCDs are not
used. The timing at the beginning of a line is discussed in detail in section 2.4.
3.0 DIGITAL CCD INTERFACE
3.1 Buffering φ1, φ2, RS, and TR
The LM9800 can drive the φ1, φ2, RS, and TR inputs of
many CCDs directly, without the need for external buffers between the LM9800 and the CCD. Most linear CCDs designed for scanner applications require 0V to 5V signal
swings into 20pF to 500pF input loading. Series resistors are
typically inserted between the driver and the CCD to control
slew rate and isolate the driver from the large load capacitances. The values of these resistors are usually given in the
CCD’s datasheet.
4.0 ANALOG INTERFACE
2.3 Reading Output Data on the DD0–DD7 Bus
The corrected digital output data representing each pixel is
available on the DD0–DD7 databus. The data is valid after
the falling edge of the EOC output. The RD input takes the
databus in and out of TRI-STATE. RD can be held low at all
times if there are no other devices needing the bus, or it can
be used to TRI-STATE the bus between pixels, allowing
other devices access to the bus. Figure 12 shows how EOC
can be tied to RD to automatically multiplex between coefficient data and conversion data.
4.1 Voltage Reference
The two REF IN pins should be connected to a 1.225V ± 2%
reference voltage capable of sinking between 2mA and 8mA
of current coming from the 400Ω–900Ω resistor string between REF OUTHI and REF IN. The LM4041-1.2 1.225V
bandgap reference is recommended for this application as
shown in Figure 28. The inexpensive “E” grade meets all the
requirements of the application and is available in a TO-92
(LM4041EIZ-1.2) package as well as a SOT-23 package
(LM4041EIM3-1.2) to minimize board space.
Due to the transient currents generated by the LM9800’s
ADC, PGA, and CDS circuitry, the REF IN pins, the REF
OUTMID pin and the REF OUTHI pin should all be bypassed
to AGND with 0.1µF monolithic capacitors.
2.4 EOC and CCLK at Startup
At the beginning of every line, the LM9800 internally synchronizes the EOC and CCLK signals with the
user-programmed sample periods. The timing of EOC and
CCLK during this adjustment period depends on the settings
in the RS Pulse Position (RS) and Sample Reference Position (SR) registers, as shown in Figures 23, 24, 25, 26.
The numbers inside the CCLK and EOC signals indicate the
width (in MCLK periods) of that signal while it is in that state.
The EOC and CCLK pulse train is synchronized with the position of the Sample Reference pulse. For most RS and SR
combinations (shown in table and timing of Figure 23), increasing SR simply stretches out the second EOC and
CCLK cycle. When SR is greater than 10 (typically with
even/odd CCDs), the timing gets more complex, and should
be considered when building any systems that rely on the
EOC and/or CCLK signals for counting, or expect them to
never be low simultaneously.
2.5 MCLK
This is the master clock input that controls the LM9800. The
pixel conversion rate is fixed at 1/8 of this frequency. Many of
the timing parameters are also relative to the frequency of
this clock.
www.national.com
DS012498-34
FIGURE 28. Voltage Reference Generation
4.2 Clamp Capacitor Selection
The output signal of most CCDs rides on a large DC offset
(typically 8V to 10V) which is incompatible with the LM9800’s
5V operation. To eliminate this offset without resorting to additional higher voltage components, the output of the CCD is
AC coupled to the LM9800 through a DC blocking capacitor,
CCLAMP (the CCD’s DOS output is not used). The value of
this capacitor is determined by the leakage current of the
24
Applications Information
the CCD’s output impedance, and the desired accuracy of
the final clamp voltage and provides the maximum clamp capacitor value:
(Continued)
LM9800’s OS input and the output impedance of the CCD.
The leakage through the OS input determines how quickly
the capacitor value will drift from the clamp value of REF
OUTMID, which then determines how many pixels can be
processed before the droop causes errors in the conversion
( ± 0.1V is the recommended limit). The output impedance of
the CCD determines how quickly the capacitor can be
charged to the clamp value during the black reference period
at the beginning of every line.
Where n = the number of optical black pixels, tDARK is the
amount of time (per pixel) that the clamp is on, ROUT is the
output impedance of the CCD, and accuracy is the ratio of
the worst-case initial capacitor voltage to the desired final
capacitor voltage. For example, if a CCD has 18 black reference pixels, the output impedance of the CCD is 1000Ω, the
LM9800 is configured to clamp for 300ns, the worst case initial voltage across the capaoitor is 10V, and the desired voltage after clamping is 0.05V (accuracy = 10/0.05 = 200),
then:
DS012498-53
FIGURE 29. OS Clamp Capacitor and Internal Clamp
The minimum clamp capacitor value is determined by the
maximum droop the LM9800 can tolerate while converting
one CCD line. The following equation takes the maximum
leakage current into the OS input, the maximum allowable
droop (100mV), the number of pixels on the CCD, and the
pixel conversion rate (fMCLK/8) and provides the minimum
clamp capacitor value:
The final value for CCLAMP should be equal to or slightly less
than CCLAMP MAX, but no less than CCLAMP MIN.
The LM9800 has been designed to work with most of the
single output CCDs in use, but it is possible that some CCDs
could have a combination of high output impedance and low
optical black pixel count that would cause CCLAMP MIN to be
greater than CCLAMP MAX. In this case the LM9800 can be
“short cycled” as shown in Figure 30, by bringing the sync
pin low and then high again shortly after clamping the black
reference pixels. This starts the line over, effectively increasing the number of black reference pixels, giving the capacitor
more time to charge up through the output impedance of the
CCD. This “short cycling” can be repeated as many times as
necessary in order to guarantee that the capacitor is charged
up to the required accuracy. Short cycling, if needed at all, is
only necessary on power up and at the beginning of new
scans where the OS coupling capacitor may have drooped
since the previous scan. If the SYNC input is continuously
being cycled at the line rate, short cycling is usually not
necessary.
For example, if the OS input leakage current is 20nA worstcase, the CCD has 2200 active pixels, the conversion rate is
2.5MHz (fMCLK = 20MHz), and the max droop desired is
0.05V, the minimum clamp capacitor value is:
The maximum size of the clamp capacitor is determined by
the amount of time available to charge it to the desired value
during the optical black portion of the CCD output. The internal clamp is on for each pixel from the rising edge of the S/H
ref pulse to the falling edge of the S/H signal pulse (see Figures 7, 8 ). This time can be calculated using the values
stored in the Sample Signal and Sample Reference configuration registers and the MCLK frequency. For normal CCDs:
And for even/odd CCDs:
DS012498-35
Where SS is the value in the Sample Signal Position register
(0–15), SR is the value in the Sample Reference Position
register (0–15), fMCLK is the MCLK frequency, and tDARK is
the amount of time (per pixel) that the clamp is on.
The following equation takes the number of optical black pixels, the amount of time (per pixel) that the clamp is closed,
FIGURE 30. Example of Short Cycling
25
www.national.com
Applications Information
tude than the Red and Green signals. The LM9800 can compensate for this by allowing the VGA to be turned on and off
at the pixel rate, allowing each pixel in a three pixel triad to
have one of two different gains (see Figure 13, VGA Gain
Switching). When the VGA is bypassed, the gain through the
VGA stage is 1. When the VGA is activated, the gain through
the VGA stage is determined by the external voltage on the
GAIN input. The Configuration Register controls whether or
not gain switching is implemented, and which pixels are
routed through the VGA.
Gain switching is controlled by the VGA On/Off bit and the
three Pixel Gain bits in the Configuration Register. If the VGA
On/Off bit is set to 0, then the VGA is bypassed (a gain of 1)
for all pixels. If the VGA On/Off bit is set to 1, then the gain
for each pixel inside every three pixel “triad” is set by the
VGA State for Pixel “n” bits. For example, if the CCD output data is RGBRGBRGB, then the VGA state bits in the
configuration register should be 001 (Pixel “a” = 0, Pixel “b”
= 0, Pixel “c” = 1). The VGA would then be bypassed for the
Red and Green pixels (for an effective gain of 1), and the
Blue pixels would be routed through the VGA for amplification (as determined by the voltage on the GAIN input). Note
that the VGA is always off (for an effective gain of 1) for the
first pixel immediately following the last optical black pixel,
regardless of the setting of the gain bit.
Other color configurations are described in section 5.0.
(Continued)
4.3 GAIN (VGA) Input
The LM9800 has a VGA (Variable Gain Amplifier) that can be
used to increase the amplitude of the CCD signal prior to
sampling, correction, and digitization. In a greyscale system,
the VGA can provide gain for all the pixels. The VGA is activated or bypassed by setting or resetting the VGA bit in the
configuration register. For greyscale systems using the VGA,
the VGA bit and all 3 Pixel Gain bits (VGAa, VGAb, and
VGAc) should be set to “1”.
The gain of the VGA is determined by the voltage on the
GAIN input as given by the equation:
Where VGAIN is the voltage at the GAIN input of the LM9800.
A detailed graph of this function is provided in the Typical
Performance Characteristics section of the datasheet.
This equation is accurate for the VGAIN voltage range of 0.5V
to 2.0V, corresponding to a gain range of x1.5 to x2.7 (typical).
The voltage at the GAIN input may be supplied by a resistive
divider between REF OUTHI and AGND or by a voltage output DAC. The divider provides a low cost fixed gain suitable
for many applications. The DAC allows fine, closed-loop adjustment of the gain which can eliminate system-to-system
CCD gain and light source intensity variations to maximize
dynamic range. The DAC0854 has the ideal output voltage
range for this application and requires no additional external
components.
4.5 Correlated Double Sampler (CDS)
Figure 33 shows the output stage of a typical CCD and the
resulting output waveform:
DS012498-39
DS012498-37
Note : 10 kΩ ≤ R1 + R2 ≤ 50 kΩ
FIGURE 31. GAIN Control with Voltage Divider
DS012498-40
FIGURE 33. CDS
Capacitor C1 converts the electrons coming from the CCD’s
shift register to an analog voltage. The source follower output stage (Q2) buffers this voltage before it leaves the CCD.
Q1 resets the voltage across capacitor C1 in between every
pixel at intervals 2 and 5. When Q1 is on, the output signal
(OS) is at its maximum. After Q1 turns off (period 3), the OS
level represents the residual voltage across C1 (VRESIDUAL).
VRESIDUAL includes charge injection from Q1, thermal noise
from the ON resistance of Q1, and other sources of error.
When the shift register clock (φ1) makes a low to high transition (period 4), the electrons from the next pixel flow into
C1. The charge across C1 now contains the voltage proportional to the number of electrons plus VRESIDUAL, an error
DS012498-38
FIGURE 32. GAIN Control with DAC
If the VGA is not going to be used, the GAIN input should be
tied to AGND.
4.4 Color Gain Switching
The VGA can also be used in color LM9800-based systems
with sequential (serial) color CCD outputs (RGBRGBRGB. . .). Sequential RGB color CCDs typically have a
Blue signal that is substantially lower (30%–70%) in ampliwww.national.com
26
Applications Information
5.0 COLOR
There are several ways to apply the LM9800 in a color system:
(Continued)
term. If OS is sampled at the end of period 3 and that voltage
is subtracted from the OS at the end of period 4, the
VRESIDUAL term is canceled and the noise on the signal is reduced. ([VSIGNAL+VRESIDUAL]−VRESIDUAL = VSIGNAL). This is
the principal of Correlated Double Sampling.
5.1 Sequential RGB Output CCD
The LM9800 implements CDS with two switched-capacitor
S/H amplifiers. The S/Hs acquire a signal within a 50 ns window which can be placed anywhere in the pixel period with
25 ns precision. See Figures 7, 8 for more detailed timing information.
4.6 Offset DAC
The offset DAC is used to compensate for DC offsets due to
the correlated double sampling stage. The offset can be corrected in 15 steps of 1 LSB size between −7 LSB and +7
LSB. Note that the DAC comes betore the PGA, so any offset errors at this stage are multiplied by the gain of the PGA.
The calibration procedure described in section 6.0 demonstrates how to use the DAC to eliminate offset errors before
scanning begins.
Note that this DAC is programmed during LM9800
calibration/configuration and is not meant to compensate for
pixel-to-pixel CCD offset errors. (CDS cancels the pixel-rate
offset errors.)
DS012498-41
FIGURE 34. Sequential RGB Output CCD
The solution shown in Figure 34 provides a 2.5Mpixels/sec
(830k RGB pixels/sec) pixel rate using a single LM9800 and
no additional external components. It requires a sequential
RGB output CCD, and provides inexpensive color at the expense of some resolution and registration.
5.2 Parallel Output CCD, One LM9800
4.7 Programmable Gain Amplifier (PGA)
The PGA provides 7 bits of pixeI-to-pixel gain correction over
a 0 dB to 6 dB (x1 to x2) range. After the input signal is
sampled and held by the CDS stage, it is amplified by the
gain indicated by the data (“PGA Code”) on the CD0–CD6
databus using the formula:
4.8 Offset Add Bit
In addition to the Offset DAC, there is a bit in the configuration register which, when set, adds a positive 1 LSB offset at
the output of the PGA. This offset ensures that any offset between the output of the PGA and the ADC is positive, so that
no dark level information is lost due to negative offsets. The
calibration procedure described in section 6.0 demonstrates
how to set this bit.
4.9 ADC
The ADC converts the normalized analog output signal to an
8-bit digital code. The EOC output goes from high to low to
indicate that a new conversion is ready. ADC data can be
latched by external memory on the rising edge of EOC . The
RD input takes the ADC’s output buffer in and out of
TRI-STATE. RD may be tied to EOC in many applications,
putting the data on the bus only when EOC is low, and allowing other data on the bus (such as CD0–CD6 correction
data) at other times. In this way the output data and correction coefficient data can share the same databus (see Figure
12).
DS012498-42
FIGURE 35. Parallel Output CCD, One LM9800
Figure 35 gives an example of how to use a single LM9800
with a triple-output RGB CCD. In this case an entire line of
red is digitized, followed by an entire line of green, then blue.
This solution provides the same 2.5Mpixels/sec (for an effective 830k RGB pixels/sec after de-interleaving) pixel rate as
the previous solution but uses a higher performance color
CCD. The multiplexers select the color to be digitized and
the resistor tap voltage corresponding to the desired gain for
that color. The resistor ladder and its multiplexer can be replaced with a voltage output DAC (such as the DAC0854) for
precision digital control of the gain for each color. Almost any
multiplexer can be used since the multiplexer switches at the
line rate, not the pixel rate, and goes into a high impedance
input. The 74HC4052 is a good choice.
4.10 ISET Input
This input is used to set internal bias currents inside the
LM9800. It should be tied to VA through a 75kΩ resistor.
27
www.national.com
Applications Information
VADC2 =
2(VOS1 + VDAC1) + VOS2 + VDAC2
(PGA gain = x2)
Solving for VOS1 and VOS2:
VOS1 = (VADC2–VADC1)–VDAC1
VOS2 = (2VADC1–VADC2)–VDAC2
These equations were used to produce this procedure for
cancelling the LM9800’s offset errors:
(Continued)
5.3 Parallel Output CCD, Three LM9800s
1.
2.
Turn off the VGA.
Set the Offset DAC to +7 LSB. (This is done to ensure
the total offset is positive and therefore measurable by
the ADC.)
3. Set the Offset Add bit to 0.
4. Set the PGA Gain to x1 (PGA code = 0).
5. Digitize a black line.
6.
7.
8.
9.
DS012498-43
FIGURE 36. Parallel Output CCD, Three LM9800s
Figure 36 uses three LM9800s to achieve a 7.5Mpixel/sec
(2.5M RGB pixels/sec) pixel rate. The three LM9800s are
synchronized by applying the same MCLK and SYNC signals to all three devices. One LM9800 provides the clock signals required for the CCD. Since the coefficient data for all
three LM9800s will be latched simultaneously on the rising
edge of CCLK, the correction coefficient bus should either be
at least 21 bits wide (7 correction coefficient bits by 3
LM9800s) or run at a 7.5MHz rate and latched into a buffer
between the correction coefficient databus and each
LM9800. Similarly, the output data for all three LM9800s will
be available simultaneously at the 3 output databusses.
10.
11.
Calculate the average (in ADC LSBs) of all the valid pixels in the black line and store that number as “B1”
(equivalent to VADC1).
Set the PGA Gain to x2 (PGA code = 127).
Digitize a black line.
Calculate the average (in ADC LSBs) of all the valid pixels in the black line and store that number as “B2”
(equivalent to VADC2).
Program the Offset DAC using the formula:
Offset DAC code = 7+B1–B2.
If 2B1 > B2, then set the Offset Add bit to 0. If 2B1 < B2,
set the Offset Add bit to 1.
6.2 Coarse Gain Calibration
The LM9800’s PGA corrects for up to 6 dB of varIation in the
CCD output signal’s white level intensity. That 6 dB range
has to be centered inside the 6 dB window of correction.
6.0 CALIBRATION
To calibrate a LM9800-based system, follow these steps:
6.1 Offset Calibration
This procedure corrects for offsets generated inside the
LM9800. Because the LM9800 uses CDS to eliminate the
pixel-to-pixel offset errors of the CCD, very good results can
be obtained even if this procedure is not implemented. In this
case the Offset DAC and Offset Add bit are simply set to 0.
DS012498-44
FIGURE 38. CCD Input Signal In Range
With the VGA off and the PGA set to a gain of 0 dB, the
LM9800 expects the maximum CCD white level output to be
below 1.2V (VREF IN) corresponding to strong pixels, and the
minimum white level output voltage to be above 0.6V (VREF
IN/2), corresponding to weaker CCD pixels or pixels further
from the light source. If the variation for a white input is inside
this range, as shown in Figure 38, the PGA can correct for it
and linearize it.
DS012498-54
If the maximum white level voltage (the voltage from the
strongest pixels) is greater than 1.2V (Figure 39), the
LM9800 will be unable to linearize the CCD’s output.
FIGURE 37. Offset Calibration
To use the Offset DAC and Offset Add bit for offset correction, the offset errors must first be determined. This is done
be measuring the voltage at the PGA output, using the ADC.
If this voltage is known with a PGA gain of x1 (0 dB) and x2
(6 dB), then the offset errors (VOS1 and VOS2) can be determined from the following two equations:
VADC1 =
1(VOS1 + VDAC1) + VOS2 + VDAC2
(PGA gain = x1)
www.national.com
28
Applications Information
close to 6dB, should do this procedure for every system, either once during manufacture or as part of the calibration
routine the scanner does on power-up or before a scan.
1. Set the PGA gain to 0dB.
(Continued)
2.
3.
4.
5.
DS012498-45
6.
FIGURE 39. CCD Input Signal Too Strong
In this case it is necessary to decrease the maximum amplitude of the CCD output. Typically this is done by reducing the
light source intensity, or decreasing the light integration time
of the CCD.
If the minimum white level output voltage (the voltage from
the weakest pixels) is less than 0.6V, then a similar situation
occurs, as shown in Figure 40.
7.
8.
DS012498-46
FIGURE 40. CCD Input Signal Too Weak
In this case the amplitude for some of the weak pixels is too
low, and cannot be corrected for by the 6 dB PGA alone. The
amplitude of the CCD’s output signal must be increased.
This can be accomplished by increasing the intensity of the
light source, increasing the integration time of the CCD, or
using the LM9800’s VGA feature to add gain to the signal
prior to the PGA.
If the minimum white level is less than 0.6V and the maximum white level is greater than 1.2V, then the LM9800 cannot correct for this amount of pixel-to-pixel variation — the
variation is greater than 6dB (Figure 41). In this case the system should be examined and steps taken to bring the variation within a 6dB window. Typically this can be done by using
more even illumination, higher quality CCDs, or better optics.
9.
If using the VGA, turn it on and set the gain voltage to
provide the approximate gain required.
If not using the VGA, turn it off.
Scan a reference line corresponding to the maximum
white input the scanner should ever have to digitize.
Find the minimum and the maximum (in ADC LSBs) of
the valid pixels in the reference line.
The reference line should correspond to a desired output code. For example, if the reference is 100% white,
the desired output code might be 240 LSBs. If the reference was 80% white (light grey), then the desired output
code might be (240)(0.8) = 192. If the maximum code
obtained after scanning the reference line is less than
the desired output code, and the minimum output code
is greater than half the desired output code, the signal is
in range. Go to section 6.3.
If the maximum code obtained after scanning the reference line is greater than the desired output code, and
the minimum output code is less than half the desired
output code, then there is more than 6dB of variation on
the input signal and the LM9800 will not be able to correct for the variations. The pixel-to-pixel variation must
be reduced to less than 6dB before the LM9800 can correct for variations.
If the maximum code obtained after scanning the reference line is greater than the desired output code for that
reference, the CCD output is too large. Reduce the light
intensity, shorten the integration time, or (if using the
VGA) reduce the voltage on the GAIN input. Go back to
Step 4.
If the minimum code obtained after scanning the reference line is smaller than half the desired output code for
that reference, the CCD output is too weak. Increase the
light intensity, lengthen the integration time, or (if using
the VGA) increase the voltage on the GAIN input. Go
back to Step 4.
6.3 PGA Correction Coefficients (Shading Calibration)
Once the input signal has been centered inside the range the
LM9800 can correct for, correction coefficients must be generated for each pixel to compensate for the gain error of that
pixel.
1. Set the PGA gain to 0dB.
2.
If using the VGA, turn it on and set the gain voltage to
provide the required gain (determined by the procedure
in Section 6.2).
3.
If not using the VGA, turn it off (set the VGA On/Off bit to
“0”).
Scan a reference line corresponding to all white or light
grey and store it in memory.
Calculate the required gain correction coefficients for
each pixel using the formula:
4.
5.
DS012498-47
FIGURE 41. CCD Input Signal Range Too Wide
The coarse gain calibration procedure follows. In scanner
systems with low unit-to-unit variations in light intensity and
CCD efficiency, this calibration can be done once and fixed
for all the systems. Other systems, where the light intensity
varies from unit-to-unit or the pixel-to-pixel variation is very
Where Uncorrected Coden is the ADC output code for pixel
n with the PGA gain = 0dB, Desired Code is the number that
corresponds to the desired output from the ADC with the
given reference line input, and Correction Coefficientn is
29
www.national.com
Applications Information
usually given in the CCD’s datasheet. If the datasheet’s requirement is given as a particular rise/fall time, the resistor
can be chosen using the graph of φ1, φ2, RS and TR Rise
Times Through A Series Resistance vs. Load Capacitance graph in the Typical Performance Characteristics
(Continued)
the gain correction number that is sent to the CD0–CD6 correction databus to provide gain correction for pixel n when
digitizing a line with the LM9800’s PGA gain correction operating.
All the Correction Coefficients must be stored and sent to the
LM9800 through the CD0–CD6 databus for every line
scanned.
7.0 POWER SUPPLY CONSIDERATIONS
7.1 General
The LM9800 should be powered by a single +5V source (unless 3V-compatible digital I/O is required — see section 7.2).
The analog supplies (VA) and the digital supplies (VD and
VD(I/O)) are brought out individually to allow separate bypassing for each supply input. They should not be powered
by two or more different supplies.
In systems with separate analog and digital +5V supplies, all
the supply pins of the LM9800 should be powered by the
cleaner analog +5V supply. Each supply input should be bypassed to its respective ground with a 0.1µF capacitor located as close as possible to the supply input pin. A single
10µF tantalum should be placed near the VA supply pin to
provide low frequency bypassing.
To minimize noise, keep the LM9800 and all analog components as far as possible from noise generators, such as
switching power supplies and high frequency digital busses.
If possible, isolate all the analog components and signals
(OS, GAIN, reference inputs and outputs, VA, AGND, ISET)
on an analog ground plane, separate from the digital ground
plane. The two ground planes should be tied together at a
single point, preferably the point where the power supply enters the PCB.
DS012498-50
FIGURE 42. CCD Interface Example
section. Given the required rise time and the input capacitance of the input being driven, the resistor value can be estimated from the graph.
Table 3 shows the Configuration Register parameters recommended for use as a starting point for most even/odd
CCDs. The Mode is set to Even/Odd, RS Pulse Width is set
to its minimum value, and RS polarity is positive. The timing,
shown in Figure 43, is determined by the RS, SR, and SS
registers. The RS pulse position (RS) is set to 10, dividing
the pixel period so that the signal portion is available for the
first 5 MCLKs following a φ1 clock edge and the black reference portion appears during the last 2 MCLKs (following the
1 MCLK wide reset pulse). Sample Reference (SR) is set to
14, so it samples the black reference just before the next φ1
clock edge. Sample Signal (SS) is set to 8, so it samples the
black reference just before the next reset pulse. These values can be adjusted to account for differences in CCDs,
CCD data delays, settling time, etc., but this is often not
necessary.
7.2 3V Compatible Digital I/O
If 3V digital I/O operation is desired, the VD(I/O) pin may be
powered by a separate 3V ± 10% or 3.3V ± 10% supply. In
this case all the digital I/O pins (CD0–CD6, CCLK, MCLK,
DD0–DD7, EOC , RD , SYNC, CS , SCLK, SDO, and SDI)
will be 3V compatible. (The CCD clock signals (φ1, φ2, RS,
and TR) remain 5V outputs, powered by VD.) In this case the
VD(I/O) input should be bypassed to DGND(I/O) with a parallel
combination of a 0.1µF capacitor and a 10µF tantalum capacitor.
7.3 Power Down Mode
Setting the Power Down bit to a “1” puts the device in a low
power standby mode. The CCD outputs (φ1, φ2, RS, and TR)
are pulled low and the analog sections are turned off to conserve power. The digital logic will continue to operate if
MCLK continues and SYNC is held high, so for minimum
power dissipation MCLK should be stopped when the
LM9800 enters the Power Down mode. Recovery from
Power Down typically takes 50µs (the time required for the
reference voltages to settle to 0.5 LSB accuracy).
8.0 TYPICAL APPLICATION
Figure 42 shows the interface between the LM9800 and a
typical even/odd output CCD, the TCD1250. The interface
for most other CCDs will be similar, the only differences being the values for the clamp capacitor and the values for the
series resistors, if needed. The clamp capacitor value is determined as shown in section 4.2. The resistor values are
www.national.com
DS012498-51
FIGURE 43. Typical Even/Odd Timing
30
Applications Information
9.0 HINTS AND COMMON SYSTEM DESIGN
PROBLEMS
(Continued)
All 4 digital outputs (φ1, φ2, RS, and TR) are enabled. The
TR pulse width is set to the minimum, 20 MCLKs, as is the
guardband between φ1 and TR. Either of these settings can
be increased if necessary.
The TR polarity is positive, as is the RS polarity. Some CCDs
may require one or both of these signals to be inverted, in
which case the corresponding bit can be set to a “1”. If there
is an inverting buffer between the LM9800 and the CCD,
these bits can be also used to correct the output polarity at
the CCD. Note that if φ1 and φ2 are inverted, then φ2 should
be used as φ1 at the CCD, and φ1 should be used as φ2 at
the CCD (Figure 44).
9.1 Reading and Writing to the Configuration Register
The Configuration Register sends and receives data LSB
(Least Significant Byte) first. Some microcontrollers send out
data MSB (Most Significant Byte) first.
The SYNC pin must be high to send/receive data to/from the
Configuration Register.
9.2 Examine the CCLK and EOC Timing
As explained earlier, this timing depends on the position of
the Sample Reference (SR) pulse, and for some values of
SR it is not what you would expect. Pay close attention to
Timing Diagrams (Figures 23, 24, 25, 26).
9.3 Setting the Dummy and Optical Black Pixel
Registers
The minimum value in the Dummy Pixels register is 1. Note
that the value in this register should be equal to 1 plus the
actual number of dummy pixels in the CCD. For example, if
the CCD being used with the LM9800 has 12 dummy pixels,
this register should be set to 13. The minimum number in the
Optical Black Pixels register is 1.
DS012498-52
FIGURE 44. φ1 and φ2 After Inversion
The number of dummy pixels and optical black reference
pixels are given in the CCD’s datasheet. The dummy pixel
register should be programmed with the number of dummy
pixels in the CCD + 1 (for example, if the CCD has 16
dummy pixels then the register should contain 17). The optical black reference register should be programmed with the
number of optical black pixels in the CCD.
The PGA gain coefficient register and PGA Gain Source bit
are used during calibration (see section 6.0). The Power
Down bit should be set to 0 for normal operation. The Offset
Add bit is also programmed durIng calibration.
The VGA settings should be all zeros (0 0 0 0) if the VGA is
not going to be used, or all ones (1 1 1 1) if the VGA will be
used to provide additional input signal gain. If using a sequential color CCD, see section 4.4).
The Offset DAC bits are programmed during calibration (section 6.0). The Test Mode bits should always be set to “0”.
31
www.national.com
Applications Information
(Continued)
TABLE 3. Configuration Register Example Data
A2
A1
A0
D7
D6
D5
D4
D3
D2
D1
D0
Standard
RS
Pulse
Width
Mode or
0
0
0
Even/Odd
RS
Pulse
Polarity
RS Pulse Position
(Minimum Value is 1)
Mode
1
0
0
0
1
1
0
0
1
1
1
0
0
1
0
1
1
1
1
1
0
1
www.national.com
0
0
0
1
0
Sample Reference Position
1
0
Sample Signal Position
1
1
1
0
φ1
φ2
RS
TR
Enable
Enable
Enable
Enable
1
1
1
1
1
0
TR Pulse Width
0
0
0
TR–φ1
TR
Guardband
Polarity
0
0
0
Dummy Pixels (Minimum Value is 1)
Set to 1 + Number of Dummy Pixels Given in CCD Datasheet
Optical Black Pixels (Minimum Value is 1)
Set to Number of Optical Black Pixels Given in CCD Datasheet
PGA Gain Coefficient
0
GAIN6
GAIN5
GAIN4
GAIN3
GAIN2
GAIN1
GAIN0
PGA Gain
Test
Power-
Offset
VGA
VGA State
VGA State
VGA State
Source
Mode
down
Add
On/Off
Pixel c
Pixel b
Pixel a
0
0
0
0
1
1
1
1
PGASRC
0
0
OFFADD
Offset DAC
Offset DAC
Offset DAC
Offset DAC
Sign
MSB
ODSIGN
VOS2
Test Modes
LSB
VOS1
VOS0
32
0
0
0
0
33
LM9800 8-Bit Greyscale/24-Bit Color Linear CCD Sensor Processor
Physical Dimensions
inches (millimeters) unless otherwise noted
52-Lead Plastic Leaded Chip Carrier (PLCC)
Order Number LM9800CCV
NS Package Number V52A
LIFE SUPPORT POLICY
NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES OR SYSTEMS WITHOUT THE EXPRESS WRITTEN APPROVAL OF THE PRESIDENT OF NATIONAL SEMICONDUCTOR CORPORATION. As used herein:
2. A critical component in any component of a life support
1. Life support devices or systems are devices or sysdevice or system whose failure to perform can be reatems which, (a) are intended for surgical implant into
sonably expected to cause the failure of the life support
the body, or (b) support or sustain life, and whose faildevice or system, or to affect its safety or effectiveness.
ure to perform when properly used in accordance
with instructions for use provided in the labeling, can
be reasonably expected to result in a significant injury
to the user.
National Semiconductor
Corporation
Americas
Tel: 1-800-272-9959
Fax: 1-800-737-7018
Email: [email protected]
www.national.com
National Semiconductor
Europe
Fax: +49 (0) 1 80-530 85 86
Email: [email protected]
Deutsch Tel: +49 (0) 1 80-530 85 85
English Tel: +49 (0) 1 80-532 78 32
Français Tel: +49 (0) 1 80-532 93 58
Italiano Tel: +49 (0) 1 80-534 16 80
National Semiconductor
Asia Pacific Customer
Response Group
Tel: 65-2544466
Fax: 65-2504466
Email: [email protected]
National Semiconductor
Japan Ltd.
Tel: 81-3-5620-6175
Fax: 81-3-5620-6179
National does not assume any responsibility for use of any circuitry described, no circuit patent licenses are implied and National reserves the right at any time without notice to change said circuitry and specifications.
Similar pages