AD546JN [ADI]

1 pA Monolithic Electrometer Operational Amplifier; 1 pA的单片静电计运算放大器
AD546JN
型号: AD546JN
厂家: ADI    ADI
描述:

1 pA Monolithic Electrometer Operational Amplifier
1 pA的单片静电计运算放大器

运算放大器
文件: 总12页 (文件大小:457K)
中文:  中文翻译
下载:  下载PDF数据表文档文件
1 pA Monolithic Electrometer  
Operational Amplifier  
a
AD546*  
CONNECTION DIAGRAM  
8-Pin Plastic  
FEATURES  
DC PERFORMANCE  
Mini-DIP Package  
1 mV max Input Offset Voltage  
Low Offset Drift: 20 V/؇C  
1 pA max Input Bias Current  
Input Bias Current Guaranteed Over Full  
Common-Mode Voltage Range  
AC PERFORMANCE  
3 V/s Slew Rate  
1 MHz Unity Gain Bandwidth  
Low Input Voltage Noise: 4 V p-p, 0.1 Hz to 10 Hz  
Available in a Low Cost, 8-Pin Plastic Mini-DIP  
Standard Op Amp Pinout  
APPLICATIONS  
Electrometer Amplifiers  
Photodiode Preamps  
pH Electrode Buffers  
Log Ratio Amplifiers  
PRODUCT HIGHLIGHTS  
PRODUCT DESCRIPTION  
1. The input bias current of the AD546 is specified, 100%  
tested and guaranteed with the device in the fully warmed-up  
condition.  
The AD546 is a monolithic electrometer combining the virtues  
of low (1 pA) input bias current with the cost effectiveness of a  
plastic mini-DIP package. Both input offset voltage and input  
offset voltage drift are laser trimmed, providing very high perfor-  
mance for such a low cost amplifier.  
2. The input offset voltage of the AD546 is laser trimmed to  
less than 1 mV (AD546K).  
Input bias currents are reduced significantly by using “topgate”  
JFET technology. The 1015 common-mode impedance,  
resulting from a bootstrapped input stage, insures that input  
bias current is essentially independent of common-mode voltage  
variations.  
3. The AD546 is packaged in a standard, low cost, 8-pin  
mini-DIP.  
4. A low quiescent supply current of 700 µA minimizes any  
thermal effects which might degrade input bias current and  
input offset voltage specifications.  
The AD546 is suitable for applications requiring both minimal  
levels of input bias current and low input offset voltage. Appli-  
cations for the AD546 include use as a buffer amplifier for cur-  
rent output transducers such as photodiodes and pH probes. It  
may also be used as a precision integrator or as a low droop rate  
sample and hold amplifier. The AD546 is pin compatible with  
standard op amps; its plastic mini-DIP package is ideal for use  
with automatic insertion equipment.  
The AD546 is available in two performance grades, all rated  
over the 0°C to +70°C commercial temperature range, and  
packaged in an 8-pin plastic mini-DIP.  
*Covered by Patent No. 4,639,683.  
REV. A  
Information furnished by Analog Devices is believed to be accurate and  
reliable. However, no responsibility is assumed by Analog Devices for its  
use, nor for any infringements of patents or other rights of third parties  
which may result from its use. No license is granted by implication or  
otherwise under any patent or patent rights of Analog Devices.  
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.  
Tel: 617/329-4700  
Fax: 617/326-8703  
(@ +25  
؇
C and  
؎
15 V dc, unless otherwise noted)  
AD546–SPECIFICATIONS  
AD546J  
AD546K  
Typ  
Model  
Conditions  
Min  
Typ  
Max  
Min  
Max  
Units  
INPUT BIAS CURRENT1  
Either Input  
Either Input  
VCM = 0 V  
CM = ±10 V  
0.2  
0.1  
1
1
0.2  
0.2  
0.5  
0.5  
pA  
pA  
V
Either Input  
@ TMAX  
Either Input  
Offset Current  
Offset Current  
@ TMAX  
VCM = 0 V  
40  
40  
0.17  
20  
20  
0.09  
pA  
pA  
pA  
V
V
CM = ±10 V  
CM = 0 V  
VCM = 0 V  
13  
7
pA  
INPUT OFFSET  
Initial Offset  
2
1
pA  
Offset @ TMAX  
vs. Temperature  
vs. Supply  
vs. Supply  
Long-Term Stability  
3
2
mV  
20  
20  
20  
20  
µV/°C  
µV/V  
µV/V  
µV/Month  
100  
100  
100  
100  
T
MIN–TMAX  
INPUT VOLTAGE NOISE  
f = 0.1 Hz to 10 Hz  
f = 10 Hz  
f = 100 Hz  
f = 1 kHz  
f = 10 kHz  
4
4
µV p-p  
90  
60  
35  
35  
90  
60  
35  
35  
nV/Hz  
nV/Hz  
nV/Hz  
nV/Hz  
INPUT CURRENT NOISE  
f = 0.1 Hz to 10 Hz  
f = 1 kHz  
1.3  
0.4  
1.3  
0.4  
fA rms  
fA/Hz  
INPUT IMPEDANCE  
Differential  
Common Mode  
V
DIFF = ±1 V  
1013ʈ1  
1013ʈ1  
ʈpF  
ʈpF  
VCM = ±10 V  
1015ʈ0.8  
1015ʈ0.8  
OPEN LOOP GAIN  
VO = ±10 V  
R
LOAD = 10 kΩ  
VO = ±10 V  
LOAD = 10 kΩ  
VO = ±10 V  
LOAD = 2 kΩ  
300  
300  
100  
80  
1000  
800  
250  
200  
300  
300  
100  
80  
1000  
800  
250  
200  
V/mV  
V/mV  
V/mV  
V/mV  
T
T
MIN–TMAX  
R
R
MIN–TMAX  
VO = ±10 V  
RLOAD = 2 kΩ  
INPUT VOLTAGE RANGE  
Differential3  
±20  
±20  
V
Common-Mode Voltage  
Common-Mode Rejection Ratio  
–10  
80  
76  
+10  
–10  
84  
76  
+10  
V
dB  
dB  
V
CM = ±10 V  
90  
80  
100  
80  
TMIN to TMAX  
OUTPUT CHARACTERISTICS  
Voltage  
R
R
LOAD = 10 kΩ  
LOAD = 2 kΩ  
–12  
–10  
15  
+12  
+10  
35  
–12  
–10  
15  
+12  
+10  
35  
V
V
mA  
pF  
Current  
Load Capacitance Stability  
Short Circuit  
Gain = +1  
20  
4000  
20  
4000  
–2–  
REV. A  
AD546  
AD546J  
Typ  
AD546K  
Typ  
Model  
Conditions  
Min  
Max  
Min  
Max  
Units  
FREQUENCY RESPONSE  
Gain BW, Small Signal  
Full Power Response  
Slew Rate, Unity Gain  
Settling Time  
G = –1  
VO = 20 V p-p  
G = –1  
to 0.1%  
to 0.01%  
0.7  
2
1.0  
50  
3
4.5  
5
0.7  
2
1.0  
50  
3
4.5  
5
MHz  
kHz  
V/µs  
µs  
µs  
Overload Recovery  
50% Overdrive  
Gain = –1  
2
2
µs  
POWER SUPPLY  
Rated Performance  
Operating Range  
Quiescent Current  
Transistor Count  
±15  
±15  
V
V
mA  
؎5  
؎18  
0.7  
؎5  
؎18  
0.7  
0.60  
50  
0.60  
50  
# of Transistors  
PACKAGE OPTIONS  
Plastic Mini-DIP (N-8)  
AD546JN  
AD546KN  
NOTES  
1Bias current specifications are guaranteed maximum, at either input, after 5 minutes of operation at TA = +25°C. Bias current increases by a factor of 2.3 for  
every 10°C rise in temperature.  
2Input offset voltage specifications are guaranteed after 5 minutes of operation at TA = +25°C.  
3Defined as max continuous voltage between inputs, such that neither exceeds ±10 V from ground.  
Specifications subject to change without notice.  
Specifications in boldface are tested on all production units at final electrical test. Results from those tests are used to calculate outgoing quality levels. All min and  
max specifications are guaranteed, although only those shown in boldface are tested on all production units.  
ABSOLUTE MAXIMUM RATINGS1  
Supply Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ±18 V  
Internal Power Dissipation . . . . . . . . . . . . . . . . . . . . . .500 mW  
Input Voltage2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ±18 V  
Output Short Circuit Duration . . . . . . . . . . . . . . . . . Indefinite  
Differential Input Voltage . . . . . . . . . . . . . . . . . . +VS and –VS  
Storage Temperature Range . . . . . . . . . . . . .65°C to +125°C  
Operating Temperature Range . . . . . . . . . . . . . . 0°C to +70°C  
Lead Temperature Range (Soldering 60 sec) . . . . . . . . +300°C  
NOTES  
1Stresses above those listed under “Absolute Maximum Ratings” may cause  
permanent damage to the device. This is a stress rating only and functional  
operation of the device at these or any other conditions above those indicated in the  
operational section of this specification is not implied. Exposure to absolute  
maximum rating conditions for extended periods may affect device reliability.  
2For supply voltages less than ±18 V, the absolute maximum input voltage is equal  
to the supply voltage.  
CAUTION  
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily  
accumulate on the human body and test equipment and can discharge without detection.  
Although the AD546 features proprietary ESD protection circuitry, permanent damage may  
occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD  
precautions are recommended to avoid performance degradation or loss of functionality.  
WARNING!  
ESD SENSITIVE DEVICE  
–3–  
REV. A  
(V = ؎15 V, unless otherwise noted)  
AD546–Typical Characteristics  
S
30  
25  
20  
20  
15  
10  
5
o
+25 C  
R
= 10kΩ  
L
15  
10  
5
+V  
–V  
OUT  
20  
15  
10  
+V  
IN  
V
= ± 15 VOLTS  
S
OUT  
–V  
IN  
5
0
0
0
10  
100  
1k  
10k  
100k  
0
5
10  
15  
20  
+15  
7
0
5
10  
15  
20  
SUPPLY VOLTAGE ± V  
LOAD RESISTANCE – Ω  
SUPPLY VOLTAGE ± V  
Figure 2. Output Voltage Range  
vs. Supply Voltage  
Figure 3. Output Voltage Swing  
vs. Resistive Load  
Figure 1. Input Voltage Range  
vs. Supply Voltage  
120  
110  
100  
90  
3000  
1000  
800  
700  
600  
500  
400  
300  
R
= 10kΩ  
L
80  
70  
100  
–15  
0
5
10  
15  
20  
–10  
0
+10  
0
5
10  
15  
20  
SUPPLY VOLTAGE ± V  
INPUT COMMON MODE VOLTAGE – V  
SUPPLY VOLTAGE ± V  
Figure 5. CMRR vs. Input  
Common-Mode Voltage  
Figure 6. Open Loop Gain vs.  
Supply Voltage  
Figure 4. Quiescent Current vs.  
Supply Voltage  
3000  
1000  
300  
300  
250  
200  
150  
30  
25  
20  
15  
10  
5
R
= 10kΩ  
L
o
+25 C  
100  
–55  
100  
0
–25  
5
35  
65  
o
95  
125  
–10  
–5  
0
5
10  
0
1
2
3
4
5
6
TEMPERATURE –  
C
COMMON-MODE VOLTAGE – Volts  
WARM-UP TIME – Minutes  
Figure 7. Open Loop Gain vs.  
Temperature  
Figure 9. Input Bias Current vs.  
Common-Mode Voltage  
Figure 8. Change in Offset  
Voltage vs. Warm-Up Time  
–4–  
REV. A  
AD546  
160  
140  
120  
100  
80  
300  
250  
200  
150  
100  
100k  
10k  
1k  
WHENEVER JOHNSON NOISE IS GREATER THAN  
AMPLIFIER NOISE, AMPLIFIER NOISE CAN BE  
CONSIDERED NEGLIGIBLE FOR THE APPLICATION.  
1 kHz BANDWIDTH  
RESISTOR JOHNSON NOISE  
100  
10  
10 Hz  
BANDWIDTH  
60  
o
1
40  
+25 C  
AMPLIFIER GENERATED NOISE  
20  
0.1  
100k 1M  
100  
1k  
10k  
0
5
10  
15  
20  
10  
10M 100M 1G  
10G 100G  
SUPPLY VOLTAGE ± VOLTS  
FREQUENCY – Hz  
SOURCE RESISTANCE – Ohms  
Figure 12. Noise vs. Source  
Resistance  
Figure 10. Input Bias Current  
vs. Supply Voltage  
Figure 11. Input Voltage Noise  
Spectral Density vs. Frequency  
40  
100  
80  
60  
40  
20  
0
100  
80  
60  
40  
20  
0
35  
30  
25  
20  
15  
10  
5
–20  
–40  
–20  
–40  
0
10  
100  
1k  
10k  
100k  
1M  
10  
100  
1k  
10k 100k  
1M  
10M  
FREQUENCY – Hz  
FREQUENCY – Hz  
Figure 13. Open Loop Frequency  
Response  
Figure 14. Large Signal Frequency  
Response  
Figure 15. CMRR vs. Frequency  
Figure 17. Output Settling Time vs.  
Output Swing and Error Voltage  
Figure 16. PSRR vs. Frequency  
REV. A  
–5–  
AD546  
Figure 18. Unity Gain Follower  
Figure 20. Unity Gain Follower  
Small Signal Pulse Response  
Figure 19. Unity Gain Follower  
Large Signal Pulse Response  
Figure 21. Unity Gain Inverter  
Figure 22. Unity Gain Inverter  
Large Signal Pulse Response  
Figure 23. Unity Gain Inverter  
Small Signal Pulse Response  
MINIMIZING INPUT CURRENT  
On-chip power dissipation will raise chip operating temperature  
causing an increase in input bias current. Due to the AD546’s  
low quiescent supply current, chip temperature when the  
(unloaded) amplifier is operated with 15 V supplies, is less than  
3°C higher than ambient. The difference in input current is  
negligible.  
The AD546 is guaranteed to have less than 1 pA max input bias  
current at room temperature. Careful attention to how the am-  
plifier is used will reduce input currents in actual applications.  
The amplifier operating temperature should be kept as low as  
possible to minimize input current. Like other JFET input am-  
plifiers, the AD546’s input current is sensitive to chip tempera-  
ture, rising by a factor of 2.3 for every 10°C rise. This is  
illustrated in Figure 24, a plot of AD546 input current versus  
ambient temperature.  
However, heavy output loads can cause a significant increase in  
chip temperature and a corresponding increase in input current.  
Maintaining a minimum load resistance of 10 kis recom-  
mended. Input current versus additional power dissipation due  
to output drive current is plotted in Figure 25.  
Figure 24. AD546 Input Bias Current vs  
Ambient Temperature  
Figure 25. AD546 Input Bias Current vs.  
Additional Power Dissipation  
–6–  
REV. A  
AD546  
Circuit Board Notes  
The AD546 is designed for through hole mount into PC boards.  
Maintaining picoampere level resolution in that environment re-  
quires a lot of care. Since both the printed circuit board and the  
amplifier’s package have a finite resistance, the voltage differ-  
ence between the amplifier’s input pin and other pins (or traces  
on the PC board) will cause parasitic currents to flow into (or  
out of) the signal path (see Figure 26). These currents can easily  
exceed the 1 pA input current level of the AD546 unless special  
precautions are taken. Two successful methods for minimizing  
leakage are guarding the AD546’s input lines and maintaining  
adequate insulation resistance.  
The AD546’s positive input (Pin 3) is located next to the nega-  
tive supply voltage pin (Pin 4). The negative input (Pin 2) is  
next to the balance adjust pin (Pin 1) which is biased at a poten-  
tial close to the negative supply voltage. The layouts shown in  
Figures 27a and 27b for the inverter and follower connections  
will guard against the effects of low surface resistance of the  
board. Note that the guard traces should be placed on both sides  
of the board. In addition the input trace should be guarded on  
both of its edges along its entire length.  
Figure 26. Sources of Parasitic Leakage Currents  
Figure 27a. Guarding Scheme—Inverter  
Figure 27b. Guarding Scheme—Follower  
–7–  
REV. A  
AD546  
Figure 28. Input Pin to Insulating Standoff  
Table I. Insulating Materials and Characteristics  
Leakage through the bulk of the circuit board will still occur  
with the guarding schemes shown in Figures 27a and 27b. Stan-  
dard “G10” type printed circuit board material may not have  
high enough volume resistivity to hold leakages at the sub-  
picoampere level particularly under high humidity conditions.  
One option that eliminates all effects of board resistance  
is shown in Figure 28. The AD546’s sensitive input pin (either  
Pin 2 when connected as an inverter, or Pin 3 when connected  
as a follower) is bent up and soldered directly to a Teflon* insu-  
lated standoff. Both the signal input and feedback component  
leads must also be insulated from the circuit board by Teflon  
standoffs or low-leakage shielded cable.  
Volume  
Minimal  
Minimal  
Resistance  
Resistivity Triboelectric Piezoelectric to Water  
Material1  
(
–CM)  
Effects  
Effects  
Absorption  
Teflon*  
Kel-F**  
Sapphire  
1017–1018  
1017–1018  
1016–1018  
W
W
M
M
W
W
W
G
W
M
G
G
G
G
Polyethylene 1014–1018  
G
M
M
W
W
G
Polystyrene  
Ceramic  
1012–1018  
1012–1014  
M
M
M
M
G
Glass Epoxy 1010–1017  
PVC  
1010–1015  
105–1012  
Contaminants such as solder flux on the board’s surface and on  
the amplifier’s package can greatly reduce the insulation resis-  
tance between the input pin and those traces with supply or sig-  
nal voltages. Both the package and the board must be kept clean  
and dry. An effective cleaning procedure is to first swab the sur-  
face with high grade isopropyl alcohol, then rinse it with deion-  
ized water and, finally, bake it at 80°C for 1 hour. Note that if  
either polystyrene or polypropylene capacitors are used on the  
printed circuit board, a baking temperature of 70°C is safer,  
since both of these plastic compounds begin to melt at approxi-  
mately +85°C.  
Phenolic  
W
W
G–Good with Regard to Property.  
M–Moderate with Regard to Property.  
W–Weak with Regard to Property.  
1Electronic Measurements, pp.15-17, Keithley Instruments, Inc., Cleveland,  
Ohio, 1977.  
*Teflon is a registered trademark of E.I. du Pont Co.  
**Kel-F is a registered trademark of 3M Company.  
OFFSET NULLING  
The AD546’s input offset voltage can be nulled by usingbalance  
Pins 1 and 5, as shown in Figure 29. Nulling the input offset  
voltage in this fashion will introduce an added input offset volt-  
age drift component of 2.4 µV/°C per millivolt of nulled offset.  
Other guidelines include making the circuit layout as compact  
as possible and reducing the length of input lines. Keeping cir-  
cuit board components rigid and minimizing vibration will re-  
duce triboelectric and piezoelectric effects. All precision high  
impedance circuitry requires shielding from electrical noise and  
interference. For example, a ground plane should be used under  
all high value (i.e., greater than 1 M) feedback resistors. In  
some cases, a shield placed over the resistors, or even the entire  
amplifier, may be needed to minimize electrical interference  
originating from other circuits. Referring to the equation in Fig-  
ure 26, this coupling can take place in either, or both, of two  
different forms—coupling via time varying fields:  
dV  
CP  
dT  
or by injection of parasitic currents by changes in capacitance  
due to mechanical vibration:  
dCp  
V
dT  
Figure 29. Standard Offset Null Circuit  
Both proper shielding and rigid mechanical mounting of compo-  
nents help minimize error currents from both of these sources.  
Table I lists various insulators and their properties.  
The circuit in Figure 30 can be used when the amplifier is used  
as an inverter. This method introduces a small voltage in series  
with the amplifier’s positive input terminal. The amplifier’s  
–8–  
REV. A  
AD546  
input offset voltage drift with temperature is not affected. How-  
ever, variation of the power supply voltages will cause offset  
shifts.  
Figure 32. Inverter Pulse Response with 1 MSource and  
Feedback Resistance  
Figure 30. Alternate Offset Null Circuit for Inverter  
AC RESPONSE WITH HIGH VALUE SOURCE AND  
FEEDBACK RESISTANCE  
Source and feedback resistances greater than 100 kwill  
magnify the effect of input capacitances (stray and inherent to  
the AD546) on the ac behavior of the circuit. The effects of  
common-mode and differential-input capacitances should be  
taken into account since the circuit’s bandwidth and stability  
can be adversely affected.  
Figure 33. Inverter Pulse Response with 1 MSource and  
Feedback Resistance, 1 pF Feedback Capacitance  
COMMON-MODE INPUT VOLTAGE OVERLOAD  
The rated common-mode input voltage range of the AD546 is  
from 3 V less than the positive supply voltage to 5 V greater  
than the negative supply voltage. Exceeding this range will de-  
grade the amplifier’s CMRR. Driving the common-mode volt-  
age above the positive supply will cause the amplifier’s output to  
saturate at the upper limit of output voltage. Recovery time is  
typically 2 µs after the input has been returned to within the  
normal operating range. Driving the input common mode volt-  
age within 1 V of the negative supply causes phase reversal of  
the output signal. In this case, normal operation is typically  
resumed within 0.5 ms of the input voltage returning within  
range.  
In a follower, the source resistance, RS, and input common-  
mode capacitance, CS (including capacitance due to board and  
capacitance inherent to the AD546), form a pole that limits cir-  
cuit bandwidth to 1/2 π RSCS. Figure 31 shows the follower  
pulse response from a 1 Msource resistance with the  
amplifier’s input pin isolated from the board, only the effect of  
the AD546’s input common-mode capacitance is seen.  
DIFFERENTIAL INPUT VOLTAGE OVERLOAD  
A plot of the AD546’s input current versus differential input  
voltage (defined as VIN+ –VIN–) appears in Figure 34. The  
Figure 31. Follower Pulse Response from 1 MSource  
Resistance  
In an inverting configuration, the differential input capacitance  
forms a pole in the circuit’s loop transmission. This can create  
peaking in the ac response and possible instability. A feedback  
capacitance can be used to stabilize the circuit. The inverter  
pulse response with RF and RS equal to 1 M, and the input pin  
isolated from the board appears in Figure 32. Figure 33 shows  
the response of the same circuit with a 1 pF feedback capaci-  
tance. Typical differential input capacitance for the AD546  
is 1 pF.  
Figure 34. Input Current vs. Differential Input Voltage  
REV. A  
–9–  
AD546  
input current at either terminal stays below a few hundred  
femtoamps until one input terminal is forced higher than 1 V to  
1.5 V above the other terminal. Under these conditions, the  
input current limits at 30 µA.  
than 1 pA), such as the FD333’s should be used, and should be  
shielded from light to keep photocurrents from being generated.  
Even with these precautions, the diodes will measurably increase  
the input current and capacitance.  
In order to achieve the low input bias currents of the AD546, it  
is not possible to use the same on-chip protection as used in  
other Analog Devices op amps. This makes the AD546 sensitive  
to handling and precautions should be taken to minimize ESD  
exposure whenever possible.  
INPUT PROTECTION  
The AD546 safely handles any input voltage within the supply  
voltage range. Subjecting the input terminals to voltages beyond  
the power supply can destroy the device or cause shifts in input  
current or offset voltage if the amplifier is not protected.  
A protection scheme for the amplifier as an inverter is shown in  
Figure 35. The protection resistor, RP, is chosen to limit the  
current through the inverting input to 1 mA for expected tran-  
sient (less than 1 second) overvoltage conditions, or to 100 µA  
for a continuous overload. Since RP is inside the feedback loop,  
and is much lower in value than the amplifier’s input resistance,  
it does not affect the inverter’s dc gain. However, the Johnson  
noise of the resistor will add root sum of squares to the  
amplifier’s input noise.  
Figure 35. Inverter with Input Current Limit  
In the corresponding version of this scheme for a follower,  
shown in Figure 36, RP and the capacitance at the positive input  
terminal will produce a pole in the signal frequency response at  
a f = 1/2 π RC. Again, the Johnson noise of RP will add to the  
amplifier’s input voltage noise.  
Figure 37 is a schematic of the AD546 as an inverter with an in-  
put voltage clamp. Bootstrapping the clamp diodes at the invert-  
ing input minimizes the voltage across the clamps and keeps the  
leakage due to the diodes low. Low leakage diodes (less  
Figure 38. Sample and Difference Circuit for Measuring  
Electrometer Leakage Currents  
MEASURING ELECTROMETER LEAKAGE CURRENTS  
There are a number of methods used to test electrometer leak-  
age currents, including current integration and direct current to  
voltage conversion. Regardless of the method used, board and  
interconnect cleanliness, proper choice of insulating materials  
(such as Teflon or Kel-F), correct guarding and shielding tech-  
niques and care in physical layout are essential for making accu-  
rate leakage measurements.  
Figure 36. Follower with Input Current Limit  
Figure 38 is a schematic of the sample and difference circuit  
which is useful for measuring the leakage currents of the AD546  
and other electrometer amplifiers. The circuit uses two AD549  
electrometer amplifiers (A and B) as current to voltage convert-  
ers with high value (1010 ) sense resistors (RSa and RSb). R1  
and R2 provide for an overall circuit sensitivity of 10 fA/mV  
(10 pA full scale). CC and CF provide noise suppression and  
loop compensation. CC should be a low leakage polystyrene ca-  
pacitor. An ultralow-leakage Kel-F test socket is used for con-  
Figure 37. Input Voltage Clamp with Diodes  
–10–  
REV. A  
AD546  
tacting the device under test. Rigid Teflon coaxial cable is used  
to make connections to all high impedance nodes. The use of  
rigid coax affords immunity to error induced by mechanical vi-  
bration and provides an outer conductor for shielding. The en-  
tire circuit is enclosed in a grounded metal box.  
Input current, IB, will contribute an output voltage error, VE1,  
proportional to the feedback resistance:  
VE1 = IB × RF  
The op amp’s input voltage offset will cause an error current  
through the photodiode’s shunt resistance, RS:  
The test apparatus is calibrated without a device under test  
present. A five minute stabilization period after the power is  
turned on is required. First, VERR1 and VERR2 are measured.  
These voltages are the errors caused by offset voltages and leak-  
age currents of the current to voltage converters.  
I = VOS/RS  
The error current will result in an error voltage (VE2) at the  
amplifier’s output equal to:  
VE2 = (1 +RF/RS) VOS  
VERR1 = 10 (VOSA IBA × RSa)  
VERR2 = 10 (VOSB IBB × RSb)  
Given typical values of photodiode shunt resistance (on the or-  
der of 109 ), RF/RS can be greater than one, especially if a large  
feedback resistance is used. Also, RF/RS will increase with tem-  
perature, as photodiode shunt resistance typically drops by a  
factor of two for every 10°C rise in temperature. An op amp  
with low offset voltage and low drift helps maintain accuracy.  
Once measured, these errors are subtracted from the readings  
taken with a device under test present. Amplifier B closes the  
feedback loop to the device under test, in addition to providing  
current to voltage conversion. The offset error of the device un-  
der test appears as a common-mode signal and does not affect  
the test measurement. As a result, only the leakage current of  
the device under test is measured.  
VA VERR1 = 10[RSa × IB(+)]  
VX VERR2 = 10[RSb × IB(–)]  
Although a series of devices can be tested after only one calibra-  
tion measurement, calibration should be updated periodically to  
compensate for any thermal drift of the current-to-voltage con-  
verters or changes in the ambient environment. Laboratory re-  
sults have shown that repeatable measurements within 10 fA can  
be realized when this apparatus is properly implemented. These  
results are achieved in part by the design of the circuit, which  
eliminates relays and other parasitic leakage paths in the high  
impedance signal lines, and in part by the inherent cancellation  
of errors through the calibration and measurement procedure.  
Figure 40. Photodiode Preamp DC Error Sources  
Photodiode Preamp Noise  
Noise limits the signal resolution obtainable with the preamp.  
The output voltage noise divided by the feedback resistance is  
the minimum current signal that can be detected. This mini-  
mum detectable current divided by the responsivity of the pho-  
todiode represents the lowest light power that can be detected  
by the preamp.  
PHOTODIODE INTERFACE  
The AD546’s 1 pA current and low input offset voltage make it  
a good choice for very sensitive photodiode preamps (Figure  
39). The photodiode develops a signal current, IS, equal to:  
Noise sources associated with the photodiode, amplifier, and  
feedback resistance are shown in Figure 41; Figure 42 is the  
voltage spectral density versus frequency plot of each of the  
noise source’s contribution to the output voltage noise (circuit  
parameters in Figure 40 are assumed). Each noise source’s rms  
contribution to the total output voltage noise is obtained by in-  
tegrating the square of its spectral density function over fre-  
quency. The rms value of the output voltage noise is the square  
root of the sum of all contributions. Minimizing the total area  
under these curves will optimize the preamplifier’s resolution for  
a given bandwidth.  
IS = R × P  
where P is light power incident on the diode’s surface in watts  
and R is the photodiode responsivity in amps/watt. RF converts  
the signal current to an output voltage:  
VOUT = RF × IS  
Figure 39. Photodiode Preamp  
DC error sources and an equivalent circuit for a small area  
(0.2 mm square) photodiode are indicated in Figure 40.  
Figure 41. Photodiode Preamp Noise Sources  
REV. A  
–11–  
AD546  
Figure 42. Photodiode Preamp Noise Sources’ Spectral  
Density vs. Frequency  
The photodiode preamp in Figure 39 can detect a signal current  
of 26 fA rms at a bandwidth of 16 Hz, which assuming a photo-  
diode responsivity of 0.5 A/W, translates to a 52 fW rms mini-  
mum detectable power. The photodiode used has a high source  
resistance and low junction capacitance. CF sets the signal band-  
width with RF and also limits the “peak” in the noise gain that  
multiplies the op amp’s input voltage noise contribution. A  
single pole filter at the amplifier’s output limits the op amp’s  
output voltage noise bandwidth to 26 Hz, a frequency compa-  
rable to the signal bandwidth. This greatly improves the  
preamplifier’s signal to noise ratio (in this case, by a factor of  
three).  
Figure 43. Photodiode Array Processor  
Photodiode Array Processor  
The AD546 is a cost effective preamp for multichannel applica-  
tions, such as amplifying signals from photo diode arrays, as il-  
lustrated in Figure 43. An AD546 preamp converts each of the  
diodes’ output currents to a voltage. An 8 to 1 multiplexer  
switches a particular preamp output to the input of an AD1380  
16-bit sampling ADC. The output of the ADC can be displayed  
or put onto a databus. Additional preamps and muxes can be  
added to handle larger arrays. Layout of multichannel circuits is  
critical. Refer to “PC board notes” for guidance.  
Figure 44. pH Probe Amplifier  
OUTLINE DIMENSIONS  
Dimensions shown in inches and (mm).  
pH PROBE AMPLIFIER  
Mini-DIP (N) Package  
A pH probe can be modeled as a mV-level voltage source with a  
series source resistance dependent upon the electrode’s compo-  
sition and configuration. The glass bulb resistance of a typical  
pH electrode pair falls between 106 and 109 . It is, therefore,  
important to select an amplifier with low enough input currents  
such that the voltage drop produced by the amplifier’s input  
bias current and the electrode resistance does not become an  
appreciable percentage of a pH unit.  
The circuit in Figure 44 illustrates the use of the AD546 as a  
pH probe amplifier. As with other electrometer applications, the  
use of guarding, shielding, Teflon standoffs, etc., is a must in  
order to capitalize on the AD546’s low input current. If an  
AD546J (1 pA max input current) is used, the error contributed  
by input current will be held below 10 mV for pH electrode  
source impedances up to 109 . Input offset voltage (which can  
be trimmed) will be below 2 mV. Refer to AD549 data sheet for  
temperature compensated pH probe amplifier circuit.  
–12–  
REV. A  

相关型号:

AD546K

1 pA Monolithic Electrometer Operational Amplifier
ADI

AD546KN

1 pA Monolithic Electrometer Operational Amplifier
ADI

AD547

High Performance, BiFET Operational Amplifiers
ADI

AD547JCHIPS

IC IC,OP-AMP,SINGLE,BIPOLAR/JFET,DIE, Operational Amplifier
ADI

AD547JH

High Performance, BiFET Operational Amplifiers
ADI

AD547JH

OP-AMP, 1000uV OFFSET-MAX, 1MHz BAND WIDTH, MBCY8, HERMETICALLY SEALED, METAL CAN, TO-99, 8 PIN
ROCHESTER

AD547JH/+

IC IC,OP-AMP,SINGLE,BIPOLAR/JFET,CAN,8PIN,METAL, Operational Amplifier
ADI

AD547KH

High Performance, BiFET Operational Amplifiers
ADI

AD547KH

OP-AMP, 500uV OFFSET-MAX, 1MHz BAND WIDTH, MBCY8, HERMETICALLY SEALED, METAL CAN, TO-99, 8 PIN
ROCHESTER

AD547KH/+

Voltage-Feedback Operational Amplifier
ETC

AD547LH

High Performance, BiFET Operational Amplifiers
ADI

AD547LH

OP-AMP, 250uV OFFSET-MAX, 1MHz BAND WIDTH, MBCY8, HERMETICALLY SEALED, METAL CAN, TO-99, 8 PIN
ROCHESTER