NSC LM98503

N
LM98503
10-Bit, 18 MSPS Camera Signal Processor
General Description
Features
The LM98503 is a CCD signal processor for digital cameras.
The processor provides a common interface to a number of
different image sensors including CCD, CMOS, and CIS.
Correlated double sampling reduces kTC noise from the image
signal. A fast, temperature stable, 8-bit digitally programmable
gain amplifier enables pixel-rate white-balancing. An auxiliary
input is provided, allowing for the selection of an external signal,
useful for sampling analog video signals. The 10-bit A/D
converter preserves the image quality with excellent noise
performance. The LM98503 also includes the supporting
functions of digital black level clamp and power down, ideally
suited for portable video applications. This low-power processor
is a natural choice for the most demanding imaging systems.
!
!
!
!
!
!
!
!
!
+3 Volt single power supply
Low power CMOS design
4-Wire serial interface
2.5V data output voltage swing
AUX input with analog clamp and programmable gain
Four color gain and offset registers
Digital black level clamp
Small 48-lead LQFP package
Supports interlace and progressive scan CCDs.
Key Specifications
!Maximum Input Level
1.0 Volt peak-peak
Applications
!CDS Sampling Rate
18 MSPS
!
!
!
!
!
!
!
!
!PGA Gain Steps
256 Steps
Digital still cameras
Digital video camcorders
Video conferencing
Security cameras
Plain paper copiers
Flatbed or handheld color scanners
Video processing for X-ray or infrared
Barcode scanners
!PGA Gain Range
0.0 - 32.0 dB
!ADC Resolution
10-Bit
!ADC Sampling Rate
!*Signal-to-Noise Ratio
18 MSPS
68dB @ 0dB Gain, 1.0V Input
!Power Dissipation
AV+ = DV+ = 2.7V
86 mW (typical)
!Operating Temp
* Note:
0oC to 70oC
20 log10 (VIN / RMS Output Noise)
Typical Digital Camera Block Diagram
CCD/CIS Sensor
Auxiliary
Video Input
LM98503
Image Processor
10
Microcontroller
Sensor Driver
© 2000 National Semiconductor Corporation
Timing Generator
Motor Controllers
www.national.com
LM98503 10-Bit, 18 MSPS Camera Signal Processor
October 2000
/CE SCLK
SI DATA
SO DATA
VREFP VREFN BOL
Serial Port Interface
Configuration Registers
Correlated
Double
Sampler
SHD
CLK
Black
Level
Clamp
8
BLKCLP
2
8
Sample/Hold
AUX IN
ACLP
SHP
Timing Control
Pixel-Rate
Offset DAC
8
V IN
Bandgap
Voltage
Reference
Analog Clamp
10-Bit
Analogto-Digital
Converter
2
Pixel-Rate
PGA
VREFT
AOUT+ AOUT-
Figure 1: Chip Block Diagram
DGND I/O
DV+
AUX IN
1
48 47 46 45 44 43 42 41 40 39 38 37
36
AGND
2
35
D9
DV+ I/O
V IN
3
34
D8
AGND
4
33
D7
AV+
5
32
D6
ACLP
31
D5
30
D4
8
29
D3
DGND
9
28
D2
DGND
10
27
D1
DV+
11
26
6
LM98503
RESET
7
48 Pin LQFP
AV+
Figure 2: Pin Out Diagram
2
DGND I/O
SI DATA
SO DATA
SCLK
/CE
VREFB
VREFN
VREFP
BLKCLP
25
13 14 15 16 17 18 19 20 21 22 23 24
BOL
12
SHD
CLK
www.national.com
DGND
AV+
AGND
VREFT
AV+
AGND
AOUT-
AOUT+
AV+
AGND
Pin Out
SHP
LM98503
Block Diagram
D0
DV+ I/O
VREFB
10-Bit
Data
Output
LM98503
Ordering Information
Commercial
(0°C ≤ TA ≤ 70°C)
70
NS Package
LM98503CCVV
LQFP
Typical Application Circuit
42
0.1µF
Serial Control Bus
SI DATA
/CE
22 23 20
SCLK
BLKCLP
21
SO DATA
16
6
15
ACLP
14
SHP
RESET
13
SHD
7
CLK
12
Camera Control
BOL
System Control
VREFP
VREFT
17
0.1µF
10µF
19
0.1µF
VREFB
VREFN
18
0.1µF
1.5K
3V
AV+
AV+
LM98503
AV+
AV+
AV+
AGND
AGND
3V
10µF
0.1µF
11,39 DV+
25,36 DV+ I/O
AGND
5
47
8
40
43
0.1µF 10µF
44
4,48
2,41
24,37 DGND I/O
AOUT+ 46
3
1
D0
35 34 33 32 31 30 29 28 27 26
Digital Video Bus
Figure 3: Typical Application Circuit Diagram
www.national.com
D1
D2
D3
D4
D5
D6
D7
D8
AOUT- 45
D9
AUX IN
DGND
V IN
9,10,38
3
LM98503
Pin Descriptions
Pin
Name
I/O
Typ
Description
1
AUX IN
I
A
Auxiliary analog input.
2
AGND
P
Analog ground return.
3
V IN
A
Analog input. AC-couple input signal through a 0.1µF capacitor.
4
AGND
P
Analog ground return.
5
AV+
P
+3 Volt power supply for the analog circuits. Bypass each supply pin with 0.1µF and
10µF capacitors in parallel.
6
ACLP
I
D
Analog clamp switch.
7
RESET
I
D
Active-high master reset. Float pin when function not being used.
8
AV+
P
+3 Volt power supply for the analog circuits. Bypass each supply pin with 0.1µF and
10µF capacitors in parallel.
9
DGND
P
Digital ground return.
10
DGND
P
Digital ground return.
11
DV+
P
+3 Volt power supply for the digital circuits. Bypass each supply pin with 0.1µF and 10µF
capacitors in parallel.
12
CLK
I
D
18 MHz clock input.
13
SHP
I
D
Correlated double sampler reset voltage clamp override. Programmable active-high or
active-low through serial interface. Connect to +3 Volt digital supply when function not
being used (register values in default condition).
14
SHD
I
D
Correlated double sampler video signal voltage sample override. Programmable activehigh or active-low through serial interface. Connect to +3 Volt digital supply when function not being used (register values in default condition).
15
BOL
I
D
Active-high beginning of line switch input. Hold high during entire line of effective pixels.
Hold low during blanking period.
16
BLKCLP
I
D
Active-high black level clamp switch input. Pulse high during black pixels to set black
pixel level to the value stored in Output Black Level register. (See page 15.)
17
VREFP
IO
A
Top of DAC reference ladder. Normally bypassed with a 0.1µF capacitor. An external
DAC reference voltage may be applied to this pin.
18
VREFN
IO
A
Bottom of DAC reference ladder. Normally bypassed with a 0.1µF capacitor. An external DAC reference voltage may be applied to this pin. Alternately, an external pull-down
resistor may be used to extend the DAC range. (See section 3.0).
19
VREFB
IO
A
Bottom of ADC reference ladder. Normally bypassed with a 0.1µF capacitor and 10µF
capacitors in parallel. An external ADC reference voltage may be applied to this pin.
20
/CE
I
D
Active-low chip enable for the serial interface.
21
SCLK
I
D
Serial interface clock used to decode the serial input data.
22
SI DATA
I
D
Serial interface input port.
23
SO DATA
O
D
Serial interface output port.
24
DGND I/O
P
Digital output driver ground return.
25
DV+ I/O
P
+3 Volt power supply for the digital output driver circuits. Bypass each supply pin with
0.1µF and 10µF capacitors in parallel.
26
D0
O
D
Digital output. Bit 0 of 9 (LSB) of the digital video output bus.
27
D1
O
D
Digital output. Bit 1 of 9 of the digital video output bus.
28
D2
O
D
Digital output. Bit 2 of 9 of the digital video output bus.
29
D3
O
D
Digital output. Bit 3 of 9 of the digital video output bus.
30
D4
O
D
Digital output. Bit 4 of 9 of the digital video output bus.
31
D5
O
D
Digital output. Bit 5 of 9 of the digital video output bus.
32
D6
O
D
Digital output. Bit 6 of 9 of the digital video output bus.
www.national.com
I
4
Pin
Name
I/O
Typ
Description
30
D4
O
D
Digital output. Bit 4 of 9 of the digital video output bus.
31
D5
O
D
Digital output. Bit 5 of 9 of the digital video output bus.
32
D6
O
D
Digital output. Bit 6 of 9 of the digital video output bus.
33
D7
O
D
Digital output. Bit 7 of 9 of the digital video output bus.
34
D8
O
D
Digital output. Bit 8 of 9 of the digital video output bus.
35
D9
O
D
Digital output. Bit 9 of 9 (MSB) of the digital video output bus.
36
DV+ I/O
P
+3 Volt power supply for the digital output driver circuits. Bypass each supply pin with
0.1µF and 10µF capacitors in parallel.
37
DGND I/O
P
Digital output driver ground return.
38
DGND
P
Digital ground return.
39
DV+
P
+3 Volt power supply for the digital circuits. Bypass each supply pin with 0.1µF and 10µF
capacitors in parallel.
40
AV+
P
+3 Volt power supply for the analog circuits. Bypass each supply pin with 0.1µF and
10µF capacitors in parallel.
41
AGND
P
Analog ground return.
42
VREFT
A
Top of ADC reference ladder. Normally bypassed with a 0.1µF capacitor and 10µF
capacitors in parallel. An external ADC reference voltage may be applied to this pin.
43
AV+
P
+3 Volt power supply for the analog circuits. Bypass each supply pin with 0.1µF and
10µF capacitors in parallel.
44
AGND
P
Analog ground return.
45
AOUT-
O
A
Negative differential analog output from correlated double sampler or PGA (selectable
through the serial interface).
46
AOUT+
O
A
Positive differential analog output from correlated double sampler or PGA (selectable
through the serial interface).
47
AV+
P
+3 Volt power supply for the analog circuits. Bypass each supply pin with 0.1µF and
10µF capacitors in parallel.
48
AGND
P
Analog ground return.
IO
Legend: (I=Input), (O=Output), (IO=Bi-directional), (P=Power), (D=Digital), (A=Analog)
www.national.com
5
LM98503
Pin Descriptions (continued)
LM98503
Absolute Maximum Ratings (Notes 1 & 2)
Operating Ratings (Notes 1 & 2)
Any Positive Supply Voltage
4.2V
Voltage On Any Input or Output Pin
-0.3V to 4.2V
Input Current at any pin (Note 3)
±35mA
Package Input Current (Note 3)
±50mA
Package Dissipation at TA = 25°C (Note 4)
ESD Susceptibility (Note 5)
Human Body Model
2500V
Machine Model
250V
Soldering Temperature
Infrared, 10 seconds (Note 6)
260°C
Storage Temperature
-65°C to 150°C
Operating Temperature Range
All Supply Voltages
V IN Voltage Range
VREFT Voltage Range
VREFB Voltage Range
VREFP Voltage Range
VREFN Voltage Range
All Digital Inputs Voltage Range
-0°C≤TA≤+70°C
+2.85V to +3.15V
0.0V to AV+
2.0V to 2.5V
0.4V to 0.9V
1.3V to 1.9V
1.3V to 1.9V
-0.05V to 3.35V
DC and Logic Level Specifications
The following specifications apply for DV+ = AV+ = DV+ I/O = +3.0V, CL = 10pF, and fCLK = 18MHz unless otherwise noted. Boldface
limits apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7).
Symbol
Parameter
Conditions
Min
(note 9)
Typical
(note 8)
Max
(note 9)
Units
Digital Input Characteristics
VIH
Logical “1” Input Voltage
VIL
Logical “0” Input Voltage
IIH
2.0
V
1.0
V
VIH = DV+, Digital inputs
except Reset
100
nA
VIH = DV+, Reset (internal
pull-down resistor)
400
µA
VIL = DGND
-100
nA
Logical “1” Input Current
IIL
Logical “0” Input Current
Digital Output Characteristics
VOH
Logical “1” Output Voltage
DV+ = 3.15V, Iout = -0.5mA
DV+ = 2.85V, Iout = -0.5mA
VOL
Logical “0” Output Voltage
DV+ = 3.15V, Iout = 1.6mA
DV+ = 2.85V, Iout = 1.6mA
IOS
Output Short Circuit Current
2.5
2.3
V
0.4
0.4
30
V
mA
Power Supply Characteristics
Power Dissipation Specifications
The following specifications apply for DV+ = AV+ = DV+ I/O = +3.0V, CL = 10pF, and fCLK = 18MHz unless otherwise noted. Boldface
limits apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7)
Symbol
Parameter
Conditions
PWR
Average Power Dissipation
AV+ = DV+ = DV+ I/O = 2.7V
AV+ = DV+ = DV+ I/O = 3.0V
AV+ = DV+ = DV+ I/O = 3.3
www.national.com
6
Min
(note 9)
Typical
(note 8)
86
100
116
Max
(note 9)
Units
mW
The following specifications apply for DV+ = AV+ = DV+ I/O = +3.0V, CL = 10pF, and fCLK = 18MHz unless otherwise noted. Boldface
limits apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7)
Conditions
Min
(note 9)
Typical
(note 8)
Max
(note 9)
Units
Symbol
Parameter
VIN
Input Voltage Level
1.0
Vp-p
IIN
Input Leakage Current
5
nA
CIN
Input Capacitance
5
pF
RIN
Input Resistance
10
kΩ
tAD
Aperture Delay
2
ns
tSHP
CLK falling edge to SHP falling
edge
10
tSHD
CLK rising edge to SHD falling
edge
14
PGA Specifications
The following specifications apply for DV+ = AV+ = DV+ I/O = +3.0V, CL = 10pF, and fCLK = 18MHz unless otherwise noted. Boldface
limits apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7)
Symbol
Parameter
Conditions
Min
(note 9)
Gain Resolution
Typical
(note 8)
Max
(note 9)
Units
8
Bits
0.125
dB
Maximum Gain
32.0
dB
Minimum Gain
0.0
dB
±5
%
Step Size
Gain Error @ 20MHz
(Gain / Resolution)
Deviation from best-fit line after endpoint correction
Offset DAC and Black Level Clamp Specifications
The following specifications apply for DV+ = AV+ = DV+ I/O = +3.0V, CL = 10pF, and fCLK = 18MHz unless otherwise noted. Boldface
limits apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7)
Typical
(note 8)
Max
(note 9)
Units
Parameter
RREF
Reference Ladder Resistance
50
kΩ
Resolution
±7
Bits
tBLKCLP
Conditions
Min
(note 9)
Symbol
Offset Adjustment Range
PGA Gain = 1.0
±54
mV
Black Level Clamp Accuracy
PGA Gain = 0.0dB
PGA Gain = 32.0dB
±.5
±.5
LSB
LSB
Black Clamp Switch Pulse Width
www.national.com
20
7
TCLK
LM98503
Correlated Double Sampler Specifications
LM98503
Analog to Digital Converter Specifications
The following specifications apply for DV+ = AV+ = DV+ I/O = +3.0V, CL = 10pF, and fCLK = 18 MHz unless otherwise noted. Boldface
limits apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7)
Symbol
Parameter
RREF
Reference Ladder Resistance
VREFT
Top of Reference Ladder
VREFB
VREFT VREFB
Conditions
Min
(note 9)
Typical
(note 8)
Max
(note 9)
Units
1000
kΩ
VREFT not driven
externally.
2.25
V
Bottom of Reference Ladder
VREFB not driven
externally.
0.75
V
Differential Reference Voltage
VREFT and VREFB not
driven externally.
1.5
V
Overrange Output Code
V IN > VREFT
1023
Underrange Output Code
V IN < VREFB
0
AC Electrical Characteristics
The following specifications apply for DV+ = AV+ = DV+ I/O = +3.0V, CL = 10pF, and fCLK = 18 MHz unless otherwise noted. Boldface
limits apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7)
Symbol
Parameter
Conditions
Min
(note 9)
Typical
(note 8)
Max
(note 9)
Units
Symbol
Parameter
Conditions
Min
(note 9)
Typical
(note 8)
Max
(note 9)
Units
1
18
20
MHz
50
55
1000
ns
fCLK
Input Clock Frequency
TCLK
Input Clock Period
tch
Clock High Time
@ CLKmax
27.5
ns
tcl
Clock Low Time
@ CLKmax
22.5
ns
Clock Duty Cycle
@ CLKmax
45/55
min/max
trc, tfc
Clock Input Rise and Fall Time
5
ns
Pipeline Delay (Latency)
7
TCLK
tVALID
Data valid time
40
tOH
Output Data Hold Time
27
tOD
Output Delay Time
20
ns
33
ns
ns
Full Channel Performance Specifications
The following specifications apply for AV+ = DV+ = DV+ I/O = 3.0V, CL = 10pF, and fCLK = 18MHz unless otherwise noted. Boldface
limits apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7)
Typical
(note 8)
Max
(note 9)
Units
Parameter
DNL
Differential Non-Linearity
±0.5
LSB
INL
Integral Non-Linearity
+3.0/-1.5
LSB
* Note:
Note 1:
Note 2:
Note 3:
Conditions
Min
(note 9)
Symbol
20 log10 (VIN / RMS Output Noise)
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.
All voltages are measured with respect to GND = AGND = DGND = 0V, unless otherwise specified.
When the voltage at any pin exceeds the power supplies (VIN < GND or VIN > AV+ or DV+), the current at that pin should
be limited to 35mA. The 50mA maximum package input current rating limits the number of pins that can safely exceed the
power supplies with an input current of 25mA to two.
www.national.com
8
Note 4:
Note 5:
Note 6:
Note 7:
The absolute maximum junction temperature (TJmax) for this device is 150oC. The maximum allowable power dissipation
is dictated by TJmax, the junction-to-ambient thermal resistance (*JA), and the ambient temperature (TA), and can be calculated using the formula PDMAX = (TJmax - TA)/*JA. In the 48-pin LQFP, *JA is 69oC/W, so PDMAX = 1,811mW at 25oC
and 1,159 mW at the maximum operating ambient temperature of 70oC. Note that the power dissipation of this device
under normal operation will typically be about TBDmW. The values for maximum power dissipation listed above will be
reached only when the LM98503 is operated in a severe fault condition.
Human body model is 100pF capacitor discharged through a 1.5kΩ resistor. Machine model is 220pF discharged through
ZERO Ohms.
See AN450, “Surface Mounting Methods and Their Effect on Product Reliability”, or the section entitled “Surface Mount”
found in any post 1986 National Semiconductor Linear Data Book, for other methods of soldering surface mount devices.
The analog inputs are protected as shown below. Input voltage magnitude up to 500mV beyond the supply rails will not
damage this device. However, input errors will be generated If the input goes above AV+ and below AGND.
VDD
Pad
Internal Circuits
IOP
VSS
Note 8:
Note 9:
Typical figures are at TJ = 25oC, and represent most likely parametric norms
Test limits are guaranteed to National's AOQL (Average Outgoing Quality Level).
www.national.com
9
LM98503
Full Channel Performance Specifications (continued)
(Default conditions are fS = 27MHz, TA = 3.0V, and AV+ = DV+ = DV+ I/O = 3.0V, unless otherwise noted.)
Gain variation part to part
Power Dissipation vs. Clock Frequency
40
39
38
fC LK =18 M H z
37
36
fC LK =18M H z
35
34
33
32
31
30
2.7
3
3.3
S up p ly V o ltag e (V o lts)
Grounded Input Noise @ 14MHz Clock Frequency
160000
Grounded Input Noise @ 18MHz Clock Frequency
160000
σ = 0 .8
M in im u m P G A
G a in
140000
140000
120000
120000
100000
100000
80000
60000
80000
60000
40000
40000
20000
20000
0
0
30
31
32
33
34
35
36
30
D ig ita l O u t p u t C o d e ( D e c im a l)
31
32
33
34
35
D ig ita l O u t p u t C o d e ( D e c im a l)
Spectral Response @ 20MHz Clock, Fin= 4.4MHz
www.national.com
σ = 0 .8
M in im u m P G A
G a in
Number of Hits
Number of Hits
LM98503
Typical Performance Characteristics
10
36
LM98503
CDS Sampling Timing
CCD
SHP1
CLAMP2
CLK
tAD
tSHP
1SHP
2
overrides the CLAMP signal’s falling edge for sampling the reset voltage (SHP is active-low by default).
The CLAMP signal is an internal signal derived from the CLK input whose falling edge samples the CCD reset voltage by default.
Figure 4: Pixel Rate Reset Voltage Sampling
CCD
SHD1
SAMPLE2
CLK
tAD
tSHD
1SHD
2
overrides the SAMPLE signal’s falling edge for sampling the video signal (SHD is active-low by default).
The SAMPLE signal is an internal signal derived from the CLK input whose falling edge samples the CCD video signal by default.
Figure 5: Pixel Rate Video Signal Sampling
www.national.com
11
LM98503
Horizontal Interval Timing
Effective Pixels
Black Pixels
Blanking Interval
Dummy Black
Pixels
Effective Pixels
CCD
BOL
ACLP
BLKCLP
Figure 6: Typical Horizontal Interval Timing
Digital Output Timing
tAD
n
n+1
n+2
n+7
n+8
CCD
tRC
90%
CLK
10%
10%
tFC
OUTPUT
DATA
n-7
tOD
tOH
n-6
n-5
n-4
n-3
tVALID
Figure 7: Digital Output Data Timing
www.national.com
12
n-2
n-1
n
n+9
LM98503
System Timing
CCD
n
n+1
n+2
n+3
CLAMP
SHP
SAMPLE
SHD
CLK
OUTPUT
DATA
n-8
n-7
n-2
Figure 8: System Timing
www.national.com
13
n-1
n
LM98503
System Overview
signal, before the resulting error is finally applied to the input of
the PGA as an analog offset generated by the DAC. The offset
integration scaling factor is stored in two bits of the software
control register 0, and the values available range from full offset
to offset divided-by-16. In addition, an offset output enable bit is
provided in the software control register 0, which when set,
routes the offset value to the digital output bus rather than the
DAC. Use of the automatic mode involves enabling the black
level offset auto-calibration bit in the software control register 0
through the serial interface. Refer to Figure 10.
1.0 Introduction
The LM98503 is a 10-bit, complete analog-to-digital camera
signal processor for use with CCD imager systems operating
from a single +3 Volt supply. The internal processing is carefully
optimized to maintain the signal-to-noise ratio and excellent
dynamic performance of most popular CCD imagers. The
system block diagram of the LM98503, shown on the cover
page of the datasheet, highlights the main features of the
device: correlated double sampling (CDS), 0-32dB digitally
programmable gain amplifier (PGA), digital black level
correction feedback loop, 8-bit DAC, analog clamp, bandgap
voltage reference, and a 10-bit, 18MHz analog-to-digital
converter.
Analog Offset Register
Automatic
Offset
Calibration
DAC
Black Pixels
Correlated double sampling (CDS) is a key feature in CCD
image processors. The sampling process consists of two
samples being taken for each pixel. The first stores the reset
voltage of the input pixel, and the second sample stores the
video signal amplitude. The two samples are subtracted from
one another, effectively removing the reset error offset of each
pixel. This sampling system operates from 1 to 18MHz. .
CDS
+
Pixel
Averaging
PGA
ADC
10
BLKCLP
Output
Data
Figure 10: Digital Black Level Correction Loop
The manual method is intended for use with processing systems
where the desired black level correction loop is external to the
LM98503. In this mode, up to four available configuration
registers may be used to store predetermined offset values that
will be applied on a pixel-rate basis. During the vertical interval,
new values may be stored in these registers for each horizontal
line.
Reset
Voltage
n+1
Offset
Integration
8
1.1 Correlated Double Sampling
n
Black Level Register
n+2
CCD
1.4 Auxiliary Input
The LM98503 includes a high-level video switch that allows a an
auxiliary video signal to be selected instead of the camera
image. When the auxiliary input is selected, the PGA gain and
DAC offset are fixed to the value in register 0, so the appropriate
gain and offset values should be written to PGA gain and DAC
offset register 0 prior to AUX IN usage.
Video
Signal
Figure 9: Correlated Double Sampling
1.2 Programmable Gain Amplifier
The amplifier has a gain ranging from 0-32dB, and is “linear in
dB” as shown in Figure 15. The PGA is addressed via an 8 bit
word downloaded through the serial interface.
1.5 Analog Clamp
During optical black pixels, a signal appears at the CCD output.
This signal is generally refered to as the “black signal level”.
This signal may be seen by the CDS circuitry as a valid video
signal rather than the actual black level signal; therefore, the
LM98503 provides an analog clamp designed to eliminate this
black signal level. Pulsing the analog input pin, ACLP, causes
the output of the CDS to be sampled by the analog clamp
circuitry. Subsequently, an adjustment is made to the CDS
reference voltages by the analog clamp to effectively eliminate
any signal level present during black pixels.
10-Bit Analog-to-Digital Converter
The selected imager’s analog signal is sampled by the CDS and
amplified to match the input requirements of the 10-bit analog to
digital converter by the PGA. The final step performed by the
system is to convert the selected analog image to digital values
with a 10 bits of resolution. The ADC has differential inputs
which aids in the coping with headroom constraints common to
+3 Volt systems. Data is acquired at the falling edge of the clock
and is available at the digital output pins 7.0 clock cycles plus
tOD later.Internal Timing Generation
All if the necessary clocks for the CDS and ADC operation are
generated internally from the LM98503’s master clock input. The
CDS sampling clocks may be overridden by the user via the
SHP and SHD clock inputs. As depicted in Figure 4 and Figure
1.3 Black Level Clamp
CCD signal processors require a reference level for the proper
handling of input signals; this reference level is commonly
referred to as the black level. The LM98503 is designed to
determine a signal’s black level during the CCD imager’s optical
black pixels.
The LM98503 provides both an analog clamp and a digital black
level correction loop. Pulsing the ACLP pin during optical black
pixels causes the analog clamp circuitry to remove the offset
associated with the input signal. Pulsing the BLKCLP pin during
dummy black pixels at the begining of a horizontal line enables
the digital black level correction loop.
Black level correction may be performed through one of two
available methods- automatic or manual. In automatic mode, the
black level is sampled from the ADC output during black pixels
by setting the BLKCLP input of the LM98503. The ADC black
level output value is then averaged over eight pixels and
subtracted from the desired black level code stored in the black
level configuration register. The result of the subtraction may
then be integrated by a preset scaling factor, effectively
smoothing any sharp transitions present in the black level
www.national.com
5, there are two signals generated internally for CDS sampling
14
to optical black. For example, a user that wants an output level
of 16 for black pixels must write this value into the register
during the horizontal interval. Note that it is not recommended
that a value of 0 be stored as the output black level. Once this
has been accomplished, driving the BLKCLP high for 20 cycles
of CLK activates the digital black clamp loop and the black level
is forced to the value stored in the output black level register, in
the example case the code value of 16. As a result of the
relationship between the DAC input and the ADC output, under
default conditions the largest black level code the LM98503 is
capable of clamping to is 36 codes. As a result, the offset DAC
may be ‘pinned’ at full range when the default setting of 32 is
used. When this occurs, the result may be that the DAC is
unable to correct line to line variations in the black level.
consequently, horizontal lines or ‘banding’ may be observed.
See Analog Offset DAC Range Adjustment, Section 3.0 of the
applications information for instructions on how to increase this
range.
referred to as CLAMP and SAMPLE.These signals provide the
rising edge reference for the sampling of the CCD input signal.
The timing of CLAMP and SAMPLE is derived from the clock;
therefore, shifting the clock phasing with respect to the CCD
input signal would also shift the rising (and falling) edges of
CLAMP and SAMPLE. The actual sampling of the CCD’s reset
voltage and video signal is performed on the falling edges of the
CLAMP and SAMPLE signals respectively. The user may modify
the position of the falling edges where the sampling of the CCD
input occurs by driving the SHP and SHD inputs of the
LM98503. The falling edges of SHP and SHD will supersede the
falling edges of CLAMP and SAMPLE respectively and cause
the duration of the sample pulse to shorten accordingly. As
evidenced in Figure 4 and Figure 5, the falling edges of SHP
and SHD should not occur earlier than tSHP or tSHD after the
respective falling [SHP] (or rising [SHD]) edge of CLK.
60
1.6 Serial Interface and Configuration Registers
LM98501, LM98502
There are many options available to the user that may be
programmed via the LM98503’s serial interface. Configuration
values are stored in registers for use by several functions such
as programmable gain, offset, black level, and color filter array.
ADC Output Code
50
The LM98503’s serial interface is used to store values into 16 8bit configuration registers. Upon power-up or external reset, the
configuration registers will contain their respective default
values. Default values place the LM98503 in ‘single channel’
mode, where only one PGA gain and offset are applied to the
input signal.
40
LM98503
Default Black Level
30
20
10
0
0
15
30
45
60
75
90
105 120 135 150 165 180 195 210 225 240 255
Output Black Level Register Value
The master CLK input is required to be running during serial
interface commands. Each command issued through the serial
interface must have a minimum of 13 data bits (see Figure 12
and Figure 13).
Figure 11: ADC Output vs. Black Level Register Value
1.10 Color Filter Array (CFA) Configuration
In order to utilize the LM98503’s programmable pixel-rate gain,
a color filter array (CFA) pattern must be defined. Some
commonly used CFA patterns are as follows:
1.7 PGA Gain Registers
Four PGA gain registers store four possible gain values for the
programmable gain amplifier (PGA). For example, these four
gain values may correspond to four possible colors in a color
filter array.
Bayer Pattern
Line 0
Line 1
Green
Blue
Red
Green
Green
Blue
Red
Green
1.8 Analog Offset
Four analog offset registers store four possible offset values that
correspond to the four PGA gain values. For example, the value
stored in the PGA gain register 0 (address 0h) is used in
conjunction with the offset value stored in the analog offset
register 0 (address 4h). This allows for four possible
combinations of PGA gain and analog offset, one for each color
filter. These registers are read-only when offset auto-calibration
is enabled in the software control register 0. It should be noted
that each offset DAC step (1 LSB) corresponds to a 0.4 LSB
step at the ADC output. Therefore, if an offset of 20 digital codes
is desired at the ADC output, a digital code value of 50 should
be stored in the analog offset register(s). As a result, the
maximum offset seen at the ADC output as a result of digital
code values stored in the analog offset register(s) is ±54 codes.
It is possible to increase the digital output range of the analog
offset DAC, resulting in an increased maximum ADC output
code corresponding to a given DAC input, but at the expense of
DAC step resolution. For more information on increasing the
DAC range, please see “Analog Offset DAC Range Adjustment”
on page 23.
CMYG Pattern
Line 0
Line 1
Magenta
Green
Yellow
Yellow
Green
Magenta
Therefore, two 8-bit words must be written to the CFA line
registers to specify the CFA pattern being used. Also, two 2-bit
numbers must be written to the CFA definition register indicating
the number of pixels per pattern in each line of the defined CFA
pattern. The information contained in the CFA line registers
indicates the registers where the respective PGA gain and offset
values are stored. For example, a system using the Bayer
pattern defined above would first write four PGA gains and their
respective offsets into the four PGA gain and four analog offset
registers. Next, two 8-bit words (one word/CFA line) would be
written to the CFA configuration registers. The 8-bit CFA
configuration words each consist of four 2-bit numbers, each of
which is the address for the gain and offset values of the of the
color that appears in that location in the CFA line. Finally, two 2bit numbers specifying the number of elements in each CFA line
must be written into the CFA definition register. A CFA
configuration will then contain four 2-bit numbers indicating the
1.9 Output Black Level
The output black level register is occupied by an 8-bit word
stored by the user that specifies the output level corresponding
www.national.com
Cyan
Cyan
registers where the gain and offset values are located for a
15
LM98503
System Overview (continued)
LM98503
System Overview (continued)
descriptions for more information regarding the power level
control registers.
configuration will then contain four 2-bit numbers indicating the
registers where the gain and offset values are located for a
maximum of four colors on each CFA line. In addition, the CFA
definition register will contain two 2-bit numbers that designate
the number of elements used in each CFA line for the particular
CFA pattern being applied to the system.
The ADC coarse and fine bank power adjustment bits are
located in the power level control 2 register, bits 7:4. Altering
these bits may significantly affect performance and power
dissipation. Please see ’DNL vs. Power Control Setting @
18MHz Clock Frequency’ and ’Power Dissipation vs. Power
Control Setting’ on page ’13.
Example A contains a CFA pattern that repeats the colors cyan
and magenta on the first line, and repeats the pattern blue,
green, green, blue on the second line. Each 2-bit number in the
CFA line registers refers to a common set of PGA gain and
offset registers for each color. The first line indicates that the
color magenta uses the gain and offset values stored in PGA
gain register 1 (address 1h) and analog offset register 1
(address 5h). Also, the first line indicates that the color cyan
uses the gain and offset values in PGA gain register 2 (address
2h) and analog offset register 2 (address 6h). The second line
indicates gain and offset values for the color blue and the color
green in the same fashion as the first line.
Example A
7
CFA Line 0
CFA Line 1
CFA Definition
XX
11
00
0
XX
00
00
01
00
11
10
11
01
In addition to specifying the gain and offset for each line, it is
also necessary to specify the number of elements contained in
each CFA line’s pixel pattern. The CFA definition register is used
to store this value (number of elements per line). In example A,
the user has stored the 2-bit binary number 01 into the CFA
definition register’s two LSB’s indicating that the pattern in line 0
contains two repeating colors or elements. Also, the 2-bit binary
number 11 has been written into bit 2 and bit 3 of the CFA
definition register indicating that the respective CFA pattern
contains four repeating colors or elements, as the colors blue
and green alternate position in the example pattern.
Once both lines for the pattern have been stored, it is applied
when the beginning of line (BOL) signal is asserted by the user.
One line of the CFA pattern is applied repeatedly until the BOL
signal is reset (at the end of the current line). Once the BOL
signal is set again, the CFA line information is changed from that
defined by the CFA line 0 register to that defined by the CFA line
1 register and the process starts again. For more details of the
timing of the BOL signal, please refer to Figure 6.
1.11 Software Control
There are two software control registers accessible via the serial
interface. The software control registers are divided into
customer (register 0) and advanced (register 1) functions.
Please refer to the register data descriptions for more
information on the software control registers.
1.12 Power Level Control
The LM98503 is equipped with two power trim registers that
may be used to adjust power levels of various circuits internal to
the device. In its default condition, the LM98503 is set for
optimum power and performance, and modifying the values
stored in the power level control registers will affect performance
as a result of the change in power level(s). In applications where
maximum performance is desired, the default values should be
used. Otherwise, power levels may be decreased at the slight
expense of performance. Please refer to the register data
www.national.com
16
LM98503
Register Memory Map
Title
Address
Default Value
PGA Gain 0
0000
0000 0000 (0d)
PGA Gain 1
0001
0000 0000 (0d)
PGA Gain 2
0010
0000 0000 (0d)
PGA Gain 3
0011
0000 0000 (0d)
Analog Offset 0
0100
0000 0000 (0d)
Analog Offset 1
0101
0000 0000 (0d)
Analog Offset 2
0110
0000 0000 (0d)
Analog Offset 3
0111
0000 0000 (0d)
CFA Configuration 0
1000
0000 0000 (0d)
CFA Configuration 1
1001
0000 0000 (0d)
CFA Definition
1010
XXXX 0000 (0d)
Output Black Level
1011
0001 0000 (32d)
Software Control 0
1100
0100 1110 (78d)
Software Control 1
1101
XX00 0X00 (0d)
Power Level Control 0
1110
1010 1010 (170d)
Power Level Control 1
1111
0101 1010 (90d)
Register Data
The following section describes all available registers in the
LM98503 register bank and their functions.
Register Name
Address
Type
Reset Value
2.0 PGA Gain Registers
Register Name
Address
Type
Reset Value
PGA Gain 0
0 Hex
Read/Write
0000 0000 Binary
Bit
Bit Symbol
Description
[7:0]
PGA Gain
0.0dB - 32.0dB in 0.125dB
steps.
Register Name
Address
Type
Reset Value
Bit Symbol
Description
[7:0]
PGA Gain
0.0dB - 32.0dB in 0.125dB
steps.
Address
Type
Reset Value
Bit Symbol
Description
[7:0]
PGA Gain
0.0dB - 32.0dB in 0.125dB
steps.
www.national.com
Bit Symbol
Description
[7:0]
PGA Gain
0.0dB - 32.0dB in 0.125dB
steps.
Register Name
Address
Type:
Reset Value
2
2 Hex
Read/Write
0000 0000 Binary
Bit
Bit
3.0 Analog Offset Registers
PGA Gain 1
1 Hex
Read/Write
0000 0000 Binary.Register NamePGA Gain
Bit
PGA Gain 3
3 Hex
Read/Write
0000 0000 Binary
17
Analog Offset 0
4 Hex
Read/Write
0000 0000 Binary
Bit
Bit Symbol
Description
[7:0]
Signed Analog
Offset
Digital representation of the
analog offset to be applied to
the input of the PGA. See
“Analog Offset” on page 15.
LM98503
Register Data (continued)
Register Name
Address
Type
Reset Value
Register Name
Address
Type
Reset Value
Analog Offset 1
5 Hex
Read/Write
0000 0000 Binary
Bit
Bit Symbol
Description
[7:0]
Signed Analog
Offset
Digital representation of the
analog offset to be applied to
the input of the PGA. See
“Analog Offset” on page 15.
Register Name
Address
Type
Reset Value
Analog Offset 2
6 Hex
Read/Write
0000 0000 Binary
Bit
Bit Symbol
Description
[7:0]
Signed Analog
Offset
Digital representation of the
analog offset to be applied to
the input of the PGA. See
“Analog Offset” on page 15.
Register Name
Address
Type
Reset Value
Bit
Bit Symbol
Description
[7:0]
Signed Analog
Offset
Digital representation of the
analog offset to be applied to
the input of the PGA. See
“Analog Offset” on page 15.
Bit Symbol
Description
[7:6]
Line0:Pixel3
Gain/Offset
2 LSB’s of register addresses
where the gain and offset for pixel
3 of the CFA pattern are stored.
[5:4]
Line0:Pixel2
Gain/Offset
2 LSB’s of register addresses
where the gain and offset for pixel
2 of the CFA pattern are stored.
Line0:Pixel 1
Gain/Offset
2 LSB’s of register addresses
where the gain and offset for pixel
1 of the CFA pattern are stored.
[3:2]
[1:0]
Line0:Pixel0
Gain/Offset
www.national.com
Description
[7:6]
Line1:Pixel3
Gain/Offset
2 LSB’s of register addresses
where the gain and offset for pixel
3 of the CFA pattern are stored.
[5:4]
Line1:Pixel2
Gain/Offset
2 LSB’s of register addresses
where the gain and offset for pixel
2 of the CFA pattern are stored.
[3:2]
Line1:Pixel 1
Gain/Offset
2 LSB’s of register addresses
where the gain and offset for pixel
1 of the CFA pattern are stored.
[1:0]
Line1:Pixel0
Gain/Offset
2 LSB’s of register addresses
where the gain and offset for pixel
0 of the CFA pattern are stored.
Color Filter Array Definition
A Hex
Read/Write
XXXX 0000 Binary
Bit
Bit Symbol
Description
[3:2]
Line 1 Pixels
Number of pixels in CFA pattern
defined in CFA line 1.
[2:1]
Line 0 Pixels
Number of pixels in CFA pattern
defined in CFA line 0.
Register Name
Address
Type
Reset Value
Color Filter Array Configuration 0
8 Hex
Read/Write
0000 0000 Binary
Bit
Bit Symbol
5.0 Output Black Level Register
4.0 Color Filter Array Registers
Register Name
Address
Type
Reset Value
Bit
Register Name
Address
Type
Reset Value
Analog Offset 3
7 Hex
Read/Write
0000 0000 Binary
Color Filter Array Configuration 1
9 Hex
Read/Write
0000 0000 Binary)
Output Black Level
B Hex
Read/Write
0001 0000 Binary
Bit
Bit Symbol
Description
[7:0]
Black Level
0 - 256 output black level digital
code value. (see “Output Black
Level” on page 15)
6.0 Software Control Registers
Register Name
Address
Type
Reset Value
2 LSB’s of register addresses
where the gain and offset for pixel
0 of the CFA pattern are stored.
18
Software Control 0 (Customer)
C Hex
Read/Write
0100 1110 Binary
Bit
Bit Symbol
Description
[7]
Offset
Output
Enable
Directs the offset error calculated
by the digital black level correction
loop to the 10 digital output data
pins.
[6]
Serial
Output
Enable
Enables the serial interface output
for reading register values.
Register Name
Address
Type
Reset Value
Software Control 1
D Hex
Read/Write
XX00 0X00 Binary
Bit
Bit Symbol
Description
[5:4]
Analog
Output
Select
Routes the selected internal
analog signal to the differential
analog output pins AOUT+ and
AOUT-.
00 CDS
01 PGA Stage 1
10 PGA Stage 2
11 PGA Stage 3
[3]
[1]
[0]
[5:4]
[3]
Bit
Bit Symbol
Description
[5:4]
PGA
CommonMode Input
Bias
Adjusts the power level of the
PGA common-mode input. The
power level is relative to the value
of the binary number stored.
[3:2]
CDS
Amplifier
Bias
Adjusts the power level of the
CDS amplifier. The power level is
relative to the value of the binary
number stored.
[1:0]
CDS
CommonMode Input
Bias
Adjusts the power level of the
CDS common-mode input. The
power level is relative to the value
of the binary number stored.
Analog
Output
Enable
Enables the differential analog
output pins AOUT+ and AOUT-. If
the analog outputs are disabled,
the pins should not be loaded.
ADC
Reference
Select
Reference biasing selection for
the analog-to-digital converter.
0 Passive biasing (resistors)
1 Active biasing (bandgap)
Bit
Bit Symbol
Description
[7:6]
ADC Coarse
Bank Bias
Reference biasing selection for
the digital-to-analog converter.
0 Active biasing (bandgap)
1 Passive biasing (resistors)
Adjusts the power level of the
ADC coarse bank. The power
level is relative to the value of the
binary number stored.
[5:4]
ADC Fine
Bank Bias
Adjusts the power level of the
ADC fine bank. The power level is
relative to the value of the binary
number stored.
[3:2]
PGA Stage
3 Amplifier
Bias
Adjusts the power level of the
PGA stage 3 amplifier. The power
level is relative to the value of the
binary number stored.
[1:0]
PGA Stage
2 Amplifier
Bias
Adjusts the power level of the
PGA stage 2 amplifier. The power
level is relative to the value of the
binary number stored.
DAC
Reference
Select
Offset
Integration
Offset integration factor selection:
00 No Scaling
01 Divide-by-4
10 Divide-by-8
11 Divide-by-16
Offset AutoCalibration
Enable
Enables the digital black level
correction loop.
Analog offset
registers are read-only when
offset auto-calibration is enabled.
[2]
SHP/SHD
Active-HIGH
Enable
Inverts the SHP and SHD inputs
causing the CDS to sample on the
rising edges of SHP and SHD.
Sampling is performed on the
falling edges of SHP and SHD
when the signals are active-low.
[1]
CDS Enable
(AUX-In
disable)
Instructs the CDS to sample the
CCD input. Otherwise, the AUX
In input is sampled in sample-andhold mode.
[0]
Analog
Power Down
Cuts power to the on-chip analog
circuitry including the CDS, PGA,
ADC, and bandgap references.
Register Name
Address
Type
Reset Value
Register Name
Address
Type
Reset Value
Power Level Control 0
E Hex
Read/Write
1010 1010 Binary
Bit
Bit Symbol
Description
[7:6]
PGA Stage
1 Amplifier
Bias
Adjusts the power level of the
PGA stage 1 amplifier. The power
level is relative to the value of the
binary number stored.
www.national.com
19
Power Level Control 1
F Hex
Read/Write
0101 1010 Binary
LM98503
Register Data (continued)
LM98503
Serial Interface Timing Specifications
The following specifications apply for all supply pins = +3.0V, CL = 10pF, and fCLK = 18MHz unless otherwise noted. Boldface limits
apply for TA = TMIN to TMAX: all other limits TA = 25oC (Note 7)
Symbol
Parameter
tMIN
SCLK Period
Min
note 9
Conditions
Typical
note 8
Max
note 9
Units
55
SCLK Duty Cycle
ns
50/50
SCLK Rise / Fall Time
60/40
%
4
ns
tCESC
SCLK Start Time After CE Low
10
ns
tDCE
Input Data to CE Rising Edge
13
tMIN
Serial Interface Timing
tCEW
tIH
tSCCE
tIS
SCLK
/CE
SI DATA
XX
1 A3 A2 A1 A0
XX
SO DATA1
D7 D6 D5 D4 D3 D2 D1 D0
tDCE
1 Serial
1 A3 A2 A1 A0
tOD
output enable must be set in software control 0 for SO DATA output. Please see register data section for more information.
Figure 12: Serial Interface Read Command Timing
tCEW
tIH
tCESC
tIS
tMIN
SCLK
/CE
SI DATA
XX
0 A3 A2 A1 A0 D7 D6 D5 D4 D3 D2 D1 D0
SO DATA1
XX
0 A3 A2 A1 A0
D7 D6 D5 D4 D3 D2 D1 D0
tDCE
1
Serial output enable must be set in software control 0 for SO DATA output. Please see register data section for more information.
Figure 13: Serial Interface Write Command Timing
www.national.com
20
LM98503
Serial Interface Read/Write Description
1.0 Writing to the Serial Registers
To write to the serial registers, the timing diagram shown in
Figure 13 must be met. First CE is toggled low. SI DATA may
then be clocked in, and the data present on the rising edge of
the serial clock is loaded in. The data continues to be clocked in,
until CE toggles high. The data present in the register is the last
13 bits of data sent before CE toggled high. Therefore, for
example, if 20 bits were sent when CE was toggled low, the first
7 bits were discarded, and the data loaded are the remaining 13
bits. As seen in Figure 13, these 13 bits are composed of 8 data
bits, 4 address bits, and 1 read/write bit. When writing to the
registers, the read/write bit must be low. When CE toggles high,
the register is written to, and the LM98503 now functions with
this new data.
2.0 Reading the Serial Registers
To read the serial registers, the timing diagram shown in Figure
12 must be met. When CE is toggled low, and data is loaded as
described above in the writing sequence. When CE toggles
high, the new 13 bit word is considered, except this time the
read/write bit should be a 1 indicating a read. The 8 data bits are
not considered, but act only as place holders. When CE toggles
low again, the data that resides at the address considered in the
previous read or write routine, begins clocking out SO DATA.
The data streams out MSB-LSB as shown in Figure 12. Whether
a read or a write was invoked in the previous sequence, the SO
DATA will clock out the contents of the address considered in
the previous sequence.
www.national.com
21
40
Max. Gain = 39.0
35
30
PGA Gain (V/V)
25
20
15
10
5
0
0
15
30
45
60
75
90
105
120
135
150
165
180
195
210
225
240
255
Figure 14: PGA Gain (Linear Scale) vs. PGA Gain Code
35
Max. Gain = 32.0 dB
30
25
PGA Gain (dB)
LM98503
PGA Gain Plots
20
15
10
5
0
0
15
30
45
60
75
90
105
120
135
150
165
180
PGA Gain Code
Figure 15: PGA Gain (Logarithmic Scale) vs. PGA Gain Code
www.national.com
22
195
210
225
240
255
of ±128 LSB to be applied at the ADC output rather than the
default maximum and minimum offsets of ±64 LSB, resulting in a
0.8 LSB DAC step to 1 LSB ADC output step relationship.
3.0 Analog-to-Digital Converter Reference Bypassing
Figure 16 shows a simple reference bypassing scheme with
minimal components. The VREFT and VREFB pins should each
be bypassed to analog ground with 10µF tantalum as well as
0.1µF ceramic capacitors. In a case where the internally
generated reference voltages are not sufficient, the user may
supply external voltages to the reference pins. However, the
reference pin VREFT should be within the range of 2.0 to 2.5
Volts. Similarly, VREFB should be driven in the range of 0.4 to 0.9
Volts. Any device used to drive the reference pins should be
able to source adequate current into the VREFT and sink
adequate current from the VREFB pin when the reference resistor
ladder is at its minimum resistivity of 850Ω.
Power Supply Considerations
The LM98503 may draw a sufficient amount of current to corrupt
improperly bypassed power supplies. A 10µF to 50µF capacitor
should be placed within 1cm of the analog power (AV+) pins of
the device in parallel with a 0.1µF ceramic chip capacitor placed
as close to the device as layout permits. Leadless chip
capacitors are preferred because they have a low lead
inductance. As is the case with virtually all high-speed
semiconductors, the LM98503 should be assumed to have little
power supply rejection; therefore, a noise-free analog power
source is required.
The reference voltage at the top of the resistor ladder (VREFT)
may be as low as 1.2 Volts above the voltage at the bottom of
the resistor ladder (VREFB) and may be as high as 1.8 Volts
above. VREFB may be as low as 0.4 Volts and as high as 0.9
Volts above ground. However, noise effects will be minimized
and accurate conversions insured when the total reference
voltage is approximately 2.25 Volts and offset from ground by
0.75 Volts.
The analog and digital power supplies of the LM98503 should
be sourced from the same supply voltage, but the supply pins
should be well isolated from one another. Isolating the supplies
prevents digital noise from coupling back into the analog supply
pins. A choke (ferrite bead) is recommended to be placed
between the analog and digital power supply pins as well as a
ceramic chip capacitor placed as close as possible to the analog
supply pin(s) of the device. Additionally, it is not recommended
that the LM98503’s digital supply be used for any other digital
circuitry on the circuit board. All other digital devices should be
powered from a separate digital supply well isolated from both
the analog and digital supplies of the LM98503.
AV+
960Ω
6.0 The LM98503 Clock
VREFT
Although the LM98503 is tested and its performance guaranteed
with a 18MHz clock, it typically will function with clock
frequencies ranging from 1MHz to 20MHz (see the LM98501 for
conversion speeds greater than 20MHz). Performance is best if
the clock rise and fall times are less than 5ns and the clock trace
is terminated near the clock input pin with a series RC network
consisting of a 100Ω resistor and a 47pF capacitor.
0.1µF
10µF
AGND
0.1µF
1kΩ
VREFB
7.0 Layout and Grounding Techniques
The proper routing of all signals and pertinent grounding
techniques are essential to insure the best signal-to-noise ratio
and dynamic performance possible. Separate analog and digital
ground planes ease meeting the datasheet limits. The analog
ground plane should be low impedance and free from noise of
other components of the system. All bypass capacitors should
be located as close to the pin as possible and connected to the
appropriate ground plane with short traces (<1cm). The analog
input should be isolated from noisy signal traces to avoid
coupling of spurious signals into the input.
610Ω
AGND
AGND
LM98503
Figure 16: Reference Bypassing
4.0 Analog Offset DAC Reference Bypassing
The analog offset DAC reference pins, VREFP and VREFN,
should be capacitively bypassed in the same fashion as the
ADC reference pins VREFT and VREFB (See “Analog-to-Digital
Converter Reference Bypassing”).
Figure 17 provides an example of a suitable layout, including
power supply routing, ground place separation, and bypass
capacitor placement. All input amplifiers, filters, and reference
components should be placed on or over the analog ground
plane. All digital circuitry and I/O lines should be placed over
and grounded via the digital ground plane. Digital and analog
signal lines should never run parallel to each other in close
proximity with each other. These signals should only cross when
absolutely necessary and then only at 90° angles.
5.0 Analog Offset DAC Range Adjustment
The analog offset DAC has an input range of ±127 LSB (see
“Register Data” on page 17). Each DAC LSB corresponds to
approximately 0.4 LSB at the ADC output. Therefore, the offset
DAC is limited to providing offset values less than or equal to
±64 LSB at the ADC output. In some applications, this range of
output may not be sufficient. It is possible to increase the range
of the DAC by adjusting the DAC reference range. The DAC
reference range may be adjusted by lowering the voltage at the
lower DAC reference voltage pin, VREFN, via use of a pull-down
resistor from VREFN to AGND. A resistor value of 1.50kΩ will
increase the DAC range by a factor of 2.0x, allowing for offsets
www.national.com
23
LM98503
Applications Information
LM98503
Applications Information (continued)
Analog Ground Plane
Digital Ground Plane
Bypass capacitors located close to
reference pins.
All analog circuitry mounted atop analog
ground plane and all digital circuitry mounted
atop digital ground plane.
C
C
LM98503
R
C
Short clock trace does not cross other
signal traces.
Narrow, single point connection between
analog and digital ground planes
C C C
Figure 17: Recommended Layout Pattern
thekeep the reference voltage(s) stable. If the reference resistor
predictable output codes.
8.0 Dynamic Performance
The LM98503 is AC tested and its dynamic performance is
guaranteed. The clock source driving the CLK input must be free
of jitter. For best AC performance, the clock source should be
isolated from other system digital circuitry with a clock tree
buffer(s). Meeting noise specifications depends largely upon
keeping digital noise out of the analog input of the LM98503.
9.0 Common Application Pitfalls
Driving the inputs (analog or digital) beyond the power
supply potential. For proper operation, all input potentials
should not be greater than 300mV above that of the power
supply. It is not uncommon for high speed digital circuits (e.g.
74F and 74AC devices) to exhibit undershoot that falls to a
potential greater than 1.0 Volt below the ground potential and
overshoot that rises to a potential greater than 1.0 Volt above
the power supply potential. A resistor of 50Ω to 100Ω in series
with the offending digital input will, in most cases, eliminate this
problem.
Attempting to drive a high capacitance digital data bus. The
more capacitance the output drivers have to charge for each
conversion output, the more current is required from the DV+ I/O
and DGND I/O supply pins. The large charging current spikes
can couple into the analog section and subsequently may
degrade dynamic performance of the system. Adequate
bypassing and maintaining separate analog and digital ground
planes will reduce this problem on the application system board.
Buffering the digital data outputs may be necessary if the data
bus being driven by the LM98503 is heavily loaded. Dynamic
performance may also be improved by adding series resistors of
47Ω at each digital output.
Driving the reference pins with devices that cannot source
or sink the current required by the reference resistor ladder.
As mentioned previously, any devices driving the reference
resistor ladder must source sufficient current into the top of the
ladder. Additionally, the device connected to the bottom ofladder
voltages are not stable the converter output will not generate
ladder must be able to sink the necessary amount of current to
www.national.com
24
9.00 ±0.25 TYP
36
25
12o TOP & BOTTOM
0 o MIN
R 0.08 MIN
24
37
1.6 MAX
R0.08 - 0.20
0.25
GAGE PLANE
0.08
0 o - 7o
0.05 - 0.10 MIN
0.20 MIN
0.60 ± 0.15
SEATING PLANE
(1.0)
DETAIL A
48
13
PIN #1 IDENT
1
12
0.2 - 0.05 TYP
NOTE 3
TYPICAL
OPTIONAL:
SHARP CORNERS
EXCEPT PIN 1 IDENT
CORNER
SEE DETAIL A
0.8 TYP
1.40 ±0.05
DIMENSIONS ARE IN MILLIMETERS
0.125 TYP
NOTES: UNLESS OTHERWISE SPECIFIED
7.0 ±0.1
NOTE 2
1. STANDARD LEAD FINISH:
7.62 MICROMETERS MINIMUM SOLDER PLATING (85/15)
THICKNESS ON ALLOY 42 / COPPER.
2. DIMENSION DOES NOT INCLUDE MOLD PROTRUSION.
MAXIMUM ALLOWABLE MOLD PROTRUSION 0.25mm PER SIDE.
3. DIMENSION DOES NOT INCLUDE DAMBAR PROTRUSION.
ALLOWABLE PROTRUSION SHALL BE 0.15mm.
4. REFERENCE JEDEC REGISTRATION MO-136, VARIATION BE.
DATED 4/93.
VBH48A
48 Lead LQFP Package
Order Number LM98503CCVV
NS Package Number VBH48A
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:
1. Life support devices or systems are devices or systems
which, (a) are intended for surgical implant into the body, or (b)
support or sustain life, and whose failure 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.
N
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) 180-530 85 86
Email:[email protected]
Deutsch Tel:+49 (0) 69 9508 6208
English Tel:+44 (0) 870 24 0 2171
Francais Tel: +33 (0) 1 41 91 8790
2. A critical component as any component in a life support
device or system whose failure to perform can be reasonably
expected to cause the failure of the life support device or
system, or to affect its safety or effectiveness.
National Semiconductor
Asia Pacific
Customer Response Group
Tel:
65-2544466
Fax:
65-2504466
Email: [email protected]
National Semiconductor
Japan Ltd.
Tel: 81-3-5639-7560
Fax: 81-3-5639-7507
National does not assume any responsibility for 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
LM98503 10-Bit, 18 MSPS Camera Signal Processor
Physical Dimensions (millimeters)