Datasheet OP176 Datasheet (Analog Devices)

Page 1
NULL
–IN +IN V–
NC V+ OUT NULL
1 2 3 45
6
7
8
OP176
1 2 3 45
6
7
8
NULL
–IN +IN
V–
OP-482
NC
V+ OUT NULL
OP176
a
FEATURES Low Noise: 6 nV/Hz High Slew Rate: 25 V/ Wide Bandwidth: 10 MHz Low Supply Current: 2.5 mA Low Offset Voltage: 1 mV Unity Gain Stable SO-8 Package
APPLICATIONS Line Driver Active Filters Fast Amplifiers Integrators
GENERAL DESCRIPTION
The OP176 is a low noise, high output drive op amp that features the Butler Amplifier front-end. This new front-end design combines both bipolar and JFET transistors to attain amplifiers with the accuracy and low noise performance of bipolar transistors, and the speed and sound quality of JFETs. Total Harmonic Distortion plus Noise equals previous audio amplifiers, but at much lower supply currents.
Improved dc performance is also provided with bias and offset currents greatly reduced over purely bipolar designs. Input offset voltage is guaranteed at 1 mV and is typically less than
*Protected by U.S. Patent No. 5101126.
µ
s
Audio Operational Amplifier
OP176*
PIN CONNECTIONS
8-Lead Narrow-Body SO 8-Lead Epoxy DIP
(S Suffix) (P Suffix)
200 µV. This allows the OP176 to be used in many dc coupled or summing applications without the need for special selections or the added noise of additional offset adjustment circuitry.
The output is capable of driving 600 loads to 10 V rms while maintaining low distortion. THD + Noise at 3 V rms is a low
0.0006%. The OP176 is specified over the extended industrial (–40°C to
+85°C) temperature range. OP176s are available in both plastic DIP and SO-8 packages. SO-8 packages are available in 2500 piece reels. Many audio amplifiers are not offered in SO-8 surface mount packages for a variety of reasons, however, the OP176 was designed so that it would offer full performance in surface mount packaging.
7
RB4
RB5
QB6
Z2
Q5
QB8
R2S
5
JB1
Z1
CB1
QB4
QB1
RB2
QB2
RB1
QB3
CC1
R1L
R1A
RB3
QB5
J1
Q1
Q3
R1P1
R1P2
R1S
1
Simplified Schematic
REV. 0
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.
Q8
Q9
RB6
QB7
R4
Q7
CC2
Q10
4
Q11
RS1
QS1
QS2
R5
RS2
RB7
Q2
J2
32
CCB
Q6
Q4 QS3
CF
R3
R2L
R2P1
R2P2
R2A
One Technology Way, P.O. Box 9106, Norwood. MA 02062-9106, U.S.A. Tel: 617/329-4700 Fax: 617/326-8703
QB9
6
Page 2
OP176–SPECIFICATIONS
ELECTRICAL CHARACTERISTICS
(@ VS = ±15.0 V, TA = +25°C unless otherwise noted)
Parameter Symbol Conditions Min Typ Max Units
INPUT CHARACTERISTICS
Offset Voltage V Offset Voltage V Input Bias Current I
Input Offset Current I
Input Voltage Range V
OS
OS
B
OS
CM
Common-Mode Rejection CMRR V
Large Signal Voltage Gain A
Offset Voltage Drift V
VO
/T5µV/°C
OS
–40°C TA +85°C 1.25 mV VCM = 0 V 350 nA
= 0 V, –40°C TA +85°C 400 nA
V
CM
V
= 0 V ±50 nA
CM
= 0 V, –40°C TA +85°C ±100 nA
V
CM
–10.5 +10.5 V
= ±10.5 V,
CM
–40°C T R
= 2 k 250 V/mV
L
= 2 k, –40°C TA +85°C 175 V/mV
R
L
= 600 200 V/mV
R
L
+85°C 80 106 dB
A
1mV
OUTPUT CHARACTERISTICS
R
Output Voltage Swing V
Output Short Circuit Current I
O
SC
= 2 k, –40°C TA +85°C –13.5 +13.5 V
L
= 600 , VS = ±18 V –14.8 +14.8 V
R
L
±25 ±50 mA
POWER SUPPLY
Power Supply Rejection Ratio PSRR V
Supply Current I
Supply Current I
Supply Voltage Range V
SY
SY
S
= ±4.5 V to ±18 V 86 108 dB
S
–40°C T V
= ±4.5 V to ±18 V, V
S
= , –40°C TA +85°C 2.5 mA
R
L
V
= ±22 V, V
S
–40°C T
+85°C80 dB
A
= 0 V, R
O
+85°C 2.75 mA
A
= 0 V,
O
= ,
L
±4.5 ±22 V
DYNAMIC PERFORMANCE
Slew Rate SR R
= 2 k 15 25 V/µs
L
Gain Bandwidth Product GBP 10 MHz
AUDIO PERFORMANCE
THD + Noise V
Voltage Noise Density e Current Noise Density i
Specifications subject to change without notice.
n
n
= 3 V rms,
IN
= 2 k, f = 1 kHz 0.001 %
R
L
f = 1 kHz 6 nV/Hz f = 1 kHz 0.5 pA/Hz
–2–
REV. 0
Page 3
OP176
WAFER TEST LIMITS
(@ VS = ±15.0 V, TA = +25°C unless otherwise noted)
Parameter Symbol Conditions Limit Units
Offset Voltage V Input Bias Current I Input Offset Current I Input Voltage Range
1
OS
B
OS
V
CM
Common-Mode Rejection CMRR V
VCM = 0 V 350 nA max V
= 0 V ±50 nA max
CM
= ±10.5 V 80 dB min
CM
1 mV max
±10.5 V min
Power Supply Rejection Ratio PSRR V = ± 4.5 V to ±18 V 86 dB min
R
Large Signal Voltage Gain A Output Voltage Range V
Supply Current I
NOTES Electrical tests and wafer probe to the limits shown. Due to variations in assembly methods and normal yield loss, yield after packaging is not guaranteed for standard product dice. Consult factory to negotiate specifications based on dice lot qualifications through sample lot assembly and testing.
1
Guaranteed by CMR test.
ABSOLUTE MAXIMUM RATINGS
Supply Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ±22 V
Input Voltage Differential Input Voltage
2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ±18 V
2
. . . . . . . . . . . . . . . . . . . . . . . ±7.5 V
VO
O
SY
1
= 2 k 250 V/mV min
L
R
= 2 k 13.5 V min
L
= ±18.0 V, RL = 600 14.8 V min
V
S
V
= ±22.0 V, V
S
= ±4.5 V to ±18 V, 2.5 mA max
V
S
VO = 0 V, R
=
L
= 0 V, R
O
= 2.75 mA max
L
DICE CHARACTERISTICS
+
V
OUT
NULL
Output Short-Circuit Duration to GND . . . . . . . . . . Indefinite
Storage Temperature Range
P, S Package . . . . . . . . . . . . . . . . . . . . . . . .–65°C to +150°C
Operating Temperature Range
OP176G . . . . . . . . . . . . . . . . . . . . . . . . . . . .–40°C to +85°C
Junction Temperature Range
P, S Package . . . . . . . . . . . . . . . . . . . . . . . .–65°C to +150°C
Lead Temperature Range (Soldering, 60 sec) . . . . . . . +300°C
Package Type θ
3
JA
θ
JC
Units
NULL
–IN
+IN
V
8-Pin Plastic DIP (P) 103 43 °C/W 8-Pin SOIC (S) 158 43 °C/W
NOTES
1
Absolute maximum ratings apply to both DICE and packaged parts, unless
otherwise noted.
2
For input voltages greater than ±7.5 V limit input current to less than 5 mA.
3
θJA is specified for the worst case conditions, i.e., θ
for P-DIP packages; θ package.
is specified for device soldered in circuit board for SOIC
JA
is specified for device in socket
JA
OP176 Die Size 0.069 × 0.067 Inch, 4,623 Sq. Mils. Substrate (Die Backside) Is Connected to V–. Transistor Count, 26.
ORDERING GUIDE
Model Temperature Range Package Description Package Option
OP176GP –40°C to +85°C 8-Pin Plastic DIP N-8 OP176GS –40°C to +85°C 8-Pin SOIC SO-8 OP176GSR –40°C to +85°C SO-8 Reel, 2500 Pieces OP176GBC +25°C DICE
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 OP176 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.
REV. 0
–3–
WARNING!
ESD SENSITIVE DEVICE
Page 4
OP176–Typical Characteristics
120
100
80
60
40
20
±VS = ±15V
–40°C ≤ T
BASED ON 300 OP AMPS
0
1
0
µ
t
VOS – µV/°C
C
+85°C
A
754362
8
30
25
20
15
10
MAXIMUM OUTPUT SWING – Volts
5
0
10k 10M1M100k1k
FREQUENCY – Hz
±VS = ±15V
TA = +25°C
RL = 2kΩ
Figure 1. Input Offset Voltage Drift Distribution @ ±15 V Figure 4. Maximum Output Swing vs. Frequency
16
VS = ±18V, +VOM, RL = 600Ω
VS = ±18V, –VOM, RL = 600Ω
15
14
VS = ±15V, –VOM, RL = 2k
13
VS = ±15V, –VOM, RL = 600
ABSOLUTE OUTPUT VOLTAGE – V
VS = ±15V, +VOM, RL = 2kΩ
VS = ±15V, +VOM, RL = 600Ω
16
VS = ±15V T
= +25°C
A
14
12
10
8
6
OUTPUT SWING – Volts
4
2
POSITIVE SWING
NEGATIVE SWING
12
–25–50
TEMPERATURE – °C
100
7550250
0
10 100 10k1k
LOAD RESISTANCE – Ω
Figure 2. Output Swing vs. Temperature Figure 5. Maximum Output Swing vs. Load Resistance
300
250
200
150
100
INPUT BIAS CURRENT – nA
50
0
–50
–25
TEMPERATURE – °C
±VS = ±15V
V
= 0V
CM
7550250
100
2.50
2.25
TA = +85°C
2.00
SUPPLY CURRENT – mA
1.75
1.50 0
TA = +25°C
TA = –40°C
±5±
SUPPLY VOLTAGE – V
±15± ±20±±10±
±25±
Figure 3. Input Bias Current vs. Temperature Figure 6. Supply Current per Amplifier vs. Supply Voltage
–4–
REV. 0
Page 5
OP176
40
20
0
1k 1M100k10k100
10
30
FREQUENCY – Hz
IMPEDANCE – Ω
AV = +100
AV = +10
AV = +1
TA = +25°C V
S
= ±15V
80
70
60
50
40
30
20
ABSOLUTE OUTPUT CURRENT – mA
10
0
SINK
SOURCE
–25–50
TEMPERATURE – °C
±VS = ±15V
7550250
Figure 7. Short Circuit Current vs. Temperature @
120
100
80
60
40
20
GAIN – dB
–20
–40
–60
PHASE
0
1k 10k 100M10M1M100k
GAIN
PHASE MARGIN = 60°
FREQUENCY – Hz
TA = +25°C VS = ±15V
R
= >600
L
100
±
15 V
90
135
180
225
POWER SUPPLY REJECTION – dB
Figure 10. Power Supply Rejection vs. Frequency
2000
1750
1500
1250
1000
PHASE – Degrees
OPEN-LOOP GAIN – V/mV
120
100
750
500
250
80
60
40
20
0
0
±VS = ±15V ±VO = ±10V
–25–50
–PSRR
1k 1M100k10k100
FREQUENCY – Hz
TEMPERATURE – °C
+PSRR
–GAIN, RL = 2kΩ
+GAIN, RL = 2k
–GAIN, RL = 600
+GAIN, RL = 600
TA = +25°C
±VS = ±15V
7550250
100
Figure 8. Open-Loop Gain & Phase vs. Frequency
50
40
30
20
10
GAIN – dB
0
–10
–20
–30
1k 10k 100M10M1M100k
FREQUENCY – Hz
TA = +25°C
V
= ±15V
S
Figure 11. Open-Loop Gain vs. Temperature
Figure 12. Closed-Loop Output Impedance vs. FrequencyFigure 9. Closed-Loop Gain vs. Frequency
REV. 0
–5–
Page 6
OP176
65
45
–75 125
60
50
–50
55
75 10050250–25
±VS = ±15V
14
6
12
8
10
TEMPERATURE – °C
PHASE MARGIN – Degrees
GAIN BANDWIDTH PRODUCT – MHz
PHASE
GAIN
40
0
100
10
5
–25–50
20
15
25
30
35
7550250
±VS = ±15V ΩRL = 2kΩ
SLEW RATE – V/µs
TEMPERATURE – °C
SR–
SR+
140
120
100
80
60
40
COMMON-MODE REJECTION – dB
20
TA = +25°C V
= ±15V
S
0
1k 1M100k10k100
FRERQUENCY – Hz
Figure 13. Common-Mode Rejection vs. Frequency
100
±VS = ±15V
90
RL = 2kΩ
VIN = 100mVp-p
80
70
60
50
40
OVERSHOOT – %
30
20
10
0
= 1
AV
CL
100
0
LOAD CAPACITANCE – pF
NEGATIVE SWING
POSITIVE SWING
1000
900800700600500400300200
Figure 14. Small Signal Overshoot vs. Load Capacitance
35
Figure 16. Gain Bandwidth Product & Phase Margin vs. Temperature
50
40
NEGATIVE SLEW RATE
30
POSITIVE SLEW RATE
20
SLEW RATE – V/µs
10
0
200
0
LOAD CAPACITANCE – pF
VS = ±15V R
= 2k
L
SWING = ±10V SLEW WINDOW = ±5V T
= +25°C
A
18001600140012001000800600400
2000
Figure 17. Slew Rate vs. Load Capacitance
30
VS = ±15V
R
= 2k
L
TA = +25°C
25
20
15
SLEW RATE – V/µs
10
5
0
0
Figure 15. Slew Rate vs. Differential Input Voltage
0.4 DIFFERENTIAL INPUT VOLTAGE – V
SR+ AND SR–
1.61.20.8
2.0
–6–
Figure 18. Slew Rate vs. Temperature
REV. 0
Page 7
25
10k
2.5
2.0
0
10 100 1k
1.5
1.0
0.5
FREQUENCY – Hz
CURRENT NOISE – pA/ Hz
VS = ±15V T
A
= +25°C
±VS = ±15V ΩTA = +25°C
20
15
10
VOLTAGE NOISE – nV/ Hz
5
0
10 100 10k1k
FREQUENCY – Hz
OP176
Figure 19. Voltage Noise Density vs. Frequency
100
90
V
OUT
(50mV/DIV)
10
0%
50mV
TIME –100ns/DIV
100nS
Figure 20. Small Signal Transient Response
Figure 21. Current Noise Density vs. Frequency
100
90
V
OUT
(5V/DIV)
10
0%
5V
TIME – 500ns/DIV
500nS
Figure 22. Large Signal Transient Response
REV. 0
–7–
Page 8
OP176
0.1
0.010
0.001
.0001
20 100 1k 10k 20k
±VS = ±18V
RL = 600Ω
10Vrms
5Vrms
3Vrms
1Vrms
APPLICATIONS Short Circuit Protection
The OP176 has been designed with output short circuit protection. The typical output drive current is ± 50 mA. This high output current and wide output swing combine to yield an excellent audio amplifier, even when driving large signals, at low power and in a small package.
Total Harmonic Distortion
Total Harmonic Distortion + Noise (THD + N) of the OP176 is well below 0.001% with any load down to 600 . However, this is dependent upon the peak output swing. In Figure 23 it is seen that the THD + Noise with 3 V rms output is below
0.001%. In the following Figure 24, THD + Noise is below
0.001% for the 10 k and 2 k loads but increases to above
0.01% for the 600 load condition. This is a result of the output swing capability of the OP176. Notice the results in Figure 25, showing THD vs. V
(V rms).
IN
0.1
±VS = ±15V
VO = 3Vrms
0.010
0.001
.0001
20 100 1k 10k 20k
600Ω
FIGURE 23. THD + Noise vs. Frequency
0.1
±VS = ±18V
VO = 10Vrms
Figure 25. THD + Noise vs. Output Amplitude (V rms)
The output of the OP176 is designed to maintain low harmonic distortion while driving 600 loads. However, driving 600 loads with very high output swings results in higher distortion if clipping occurs.
To attain low harmonic distortion with large output swings, supply voltages may be increased. Figure 26 shows the perfor­mance of the OP176 driving 600 loads with supply voltages varying from ±18 volts to ±20 volts. Notice that with ± 18 volt supplies the distortion is fairly high, while with ± 20 volt supplies it is a very low 0.0007%.
0.1
RL = 600Ω
0.010
VO = ±18V
0.010
0.001
0.0001 20 100 1k 10k 20k
Figure 24. THD + Noise vs. R
600Ω
2k
10k
LOAD
0.001
VO = ±19V
0.0001 20 100 1k 10k 20k
VO = ±20V
Figure 26. THD + Noise vs. Supply Voltage
–8–
VO = ±22V
REV. 0
Page 9
OP176
OP176
1.4kΩ
1.4kΩ
+
2
3
6
Noise
The voltage noise density of the OP176 is below 6 nV/Hz from 30 Hz. This enables low noise designs to have good perfor­mance throughout the full audio range. Figure 27 shows a typical OP176 with a 1/f corner at 6 Hz.
10.0 µV /DIVCH A: 80.0 µV FS
Hz
50Hz /
BW:
300 mHz
\ 0 Hz MKR:
MKR: 15.9 µV/
5.4 Hz
Figure 27. 1/f Noise Corner
Noise Testing
For audio applications the noise density is usually the most important noise parameter. For characterization the OP176 is tested using an Audio Precision, System One. The input signal to the Audio Precision must be amplified enough to measure accurately. For the OP176 the noise is gained by approximately 1020 using the circuit shown in Figure 28. Any readings on the Audio Precision must then be divided by the gain. In imple­menting this test fixture, good supply bypassing is essential.
100
OP176
909
100
OP37
909
490
OP37
4.42k
OUTPUT
Figure 28. Noise Test
Upgrading “5534‘’ Sockets
The OP176 is a superior amplifier for upgrading existing designs using the industry standard 5534. In most application circuits, the OP176 can directly replace the 5534 without any modifications to the surrounding circuitry. Like the 5534, the OP176 follows the industry standard, single op amp pinout. The difference between these two devices is the location of the null pins and the 5534’s compensation capacitor.
The 5534 normally requires a 22 pF capacitor between Pins 5 and 8 for stable operation. Since the OP176 is internally compensated for unity gain operation, it does not require external compensation. Nevertheless, if the 5534 socket already includes a capacitor, the OP176 can be inserted without removing it. Since the OP176’s Pin 8 is a “NO CONNECT’’ pin, there is no internal connection to that pin. Thus, the 22 pF capacitor would be electrically connected through Pin 5 to the internal nulling circuitry. With the other end left open, the capacitor should have no effect on the circuit. However, to avoid altogether any possibility for noise injection, it is recom­mended that the 22 pF capacitor be cut out of the circuit entirely.
If the original 5534 socket includes offset nulling circuitry, one would find a 10 k to 100 k potentiometer connected between Pins 1 and 8 with said potentiometer’s wiper arm connected to V+. In order to upgrade the socket to the OP176, this circuit should be removed before inserting the OP176 for its offset nulling scheme uses Pins 1 and 5. Whereas the wiper arm of the 5534 trimming potentiometer is connected to the positive supply, the OP176’s wiper arm is connected to the negative supply. Directly substituting the OP176 into the original socket would inject a large current imbalance into its input stage. In this case, the potentiometer should be removed altogether, or, if nulling is still required, the trimming potentiometer should be rewired to match the nulling circuit as illustrated in Figure 29.
+V
S
7
2
V
OUT
3
OP176
1
4
–V
P1
S
6
5
P1 = 10kΩ
V
TRIM RANGE = ±2mV
OS
Figure 29. Offset Voltage Nulling Scheme
Input Overcurrent Protection
The maximum input differential voltage that can be applied to the OP176 is determined by a pair of internal Zener diodes connected across its inputs. They limit the maximum differen­tial input voltage to ±7.5 V. This is to prevent emitter-base junction breakdown from occurring in the input stage of the OP176 when very large differential voltages are applied. However, in order to preserve the OP176’s low input noise voltage, internal resistances in series with the inputs were not used to limit the current in the clamp diodes. In small signal applications, this is not an issue; however, in applications where large differential voltages can be inadvertently applied to the device, large transient currents can flow through these diodes. Although these diodes have been designed to carry a current of ±5 mA, external resistors as shown in Figure 30 should be used in the event that the OP176’s differential voltage were to exceed ±7.5 V.
Figure 30. Input Overcurrent Protection
REV. 0
–9–
Page 10
OP176
+15V
+
10µF
0.1µF
2
3
7
6
4
V
IN
V
OUT
R
L
2k
–15V
10µF
0.1µF
OP176
Output Voltage Phase Reversal
Since the OP176’s input stage combines bipolar transistors for low noise and p-channel JFETs for high speed performance, the output voltage of the OP176 may exhibit phase reversal if either of its inputs exceeds the specified negative common-mode input voltage. This might occur in some applications where a trans­ducer, or a system, fault might apply very large voltages upon the inputs of the OP176. Even though the input voltage range of the OP176 is ±10.5 V, an input voltage of approximately –13.5 V will cause output voltage phase reversal. In inverting amplifier configurations, the OP176’s internal 7.5 V clamping diodes will prevent phase reversal; however, they will not prevent this effect from occurring in noninverting applications. For these applications, the fix is a 3.92 k resistor in series with the noninverting input of the device and is illustrated in Figure 31.
RFB*
V
R
L
2k
OUT
V
IN
*R
FB
2
3
R
S
3.92k
IS OPTIONAL
OP176
6
Figure 31. Output Voltage Phase Reversal Fix
Overdrive Recovery
The overdrive recovery time of an operational amplifier is the time required for the output voltage to recover to a rated output level from a saturated condition. This recovery time is impor­tant in applications where the amplifier must recover quickly after a large abnormal transient event. The circuit shown in Figure 32 was used to evaluate the OP176’s overload recovery time. The OP176 takes approximately 1 µs to recover to V +10 V and approximately 900 ns to recover to V
2
3
R2
10kΩ
OP176
6
V
IN
4Vp-p
@100Hz
R1
1kΩ
R
S
909
= –10 V.
OUT
V
OUT
R
L
2.43kΩ
Figure 32. Overload Recovery Time Test Circuit
High Speed Operation
As with most high speed amplifiers, care should be taken with supply decoupling, lead dress, and component placement. Recommended circuit configurations for inverting and noninverting applications are shown in Figure 33 and Figure 34.
OUT
Figure 33. Unity Gain Follower
+15V
10µF
+
0.1µF
10pF
4.99k
V
IN
2.49k
=
2
OP176
3
–15V
4.99k
7
4
0.1µF
10µF
V
6
+
OUT
2k
Figure 34. Unity Gain Inverter
In inverting and noninverting applications, the feedback resistance forms a pole with the source resistance and capaci­tance (R shown in Figure 35. With R
and CS) and the OP176’s input capacitance (CIN), as
S
and R
S
in the k range, this pole
F
can create excess phase shift and even oscillation. A small capacitor, C setting R
, in parallel with RFB eliminates this problem. By
FB
+ CIN) = R
S (CS
, the effect of the feedback pole is
FB CFB
completely removed.
C
FB
R
FB
R
S
C
S
C
IN
V
OUT
–10–
Figure 35. Compensating the Feedback Pole
REV. 0
Page 11
OP176
6
2 3
A2
6
3
2
A1
3
2
6
A3
V
IN
V
O1
V
O2
R3
2k
R9
50
R11
1k
P1
10k
R12 1k
R10 50
R8
2k
R2
2k
R5
2k
R4
2k
R1
2k
R7
2k
R6
2k
V
O2
– VO1 = V
IN
A1, A2, A3 = OP176 GAIN = SET R2, R4, R5 = R1 AND R6, R7, R8 = R3
R3 R1
Attention to Source Impedances Minimizes Distortion
Since the OP176 is a very low distortion amplifier, careful attention should be given to source impedances seen by both inputs. As with many FET-type amplifiers, the p-channel JFETs in the OP176’s input stage exhibit a gate-to-source capacitance that varies with the applied input voltage. In an inverting configuration, the inverting input is held at a virtual ground and, as such, does not vary with input voltage. Thus, since the gate-to-source voltage is constant, there is no distor­tion due to input capacitance modulation. In noninverting applications, however, the gate-to-source voltage is not constant. The resulting capacitance modulation can cause distortion above 1 kHz if the input impedance is > 2 k and unbalanced.
Figure 36 shows some guidelines for maximizing the distortion performance of the OP176 in noninverting applications. The best way to prevent unwanted distortion is to ensure that the parallel combination of the feedback and gain setting resistors (R
and R
F
) is less than 2 k. Keeping the values of these
G
resistors small has the added benefits of reducing the thermal noise of the circuit and dc offset errors. If the parallel combina­tion of R
F
resistor, R
V
and R
IN
is larger than 2 k, then an additional
G
, should be used in series with the noninverting
S
R
G
RS*
R
F
OP176
* RS = RG//RF IF RG//RF > 2kΩ FOR MINIMUM DISTORTION
V
OUT
Figure 36. Balanced Input Impedance to Mininize Distortion in Noninverting Amplifier Circuits
R
V
IN
R
= R
X
R
=I+
C
F
R
G
[ ]
F
C
F
R
R
X
+
R
F
G
CLR
)
R
F
O
OP176
WHERE RO = OPEN-LOOP OUTPUT RESISTANCE
O RG
F
I
(
)(
|
A
|
CL
V
C
OUT
L
Figure 37. In-the-Loop Compensation Technique for Driving Capacitive Loads
APPLICATIONS USING THE OP176 A High Speed, Low Noise Differential Line Driver
The circuit of Figure 38 is a unique line driver widely used in many applications. With ±18 V supplies, this line driver can deliver a differential signal of 30 V p-p into a 2.5 k load. The high slew rate and wide bandwidth of the OP176 combine to yield a full power bandwidth of 130 kHz while the low noise front end produces a referred-to-input noise voltage spectral density of 15 nV/
Hz. The circuit is capable of driving lower impedance loads as well. For example, with a reduced output level of 5 V rms (14 V p-p), the circuit exhibits a full-power bandwidth of 190 kHz while driving a differential load of 249 ! The design is a transformerless, balanced transmission system where output common-mode rejection of noise is of paramount importance. Like the transformer-based design, either output can be shorted to ground for unbalanced line driver applications without changing the circuit gain of 1. Other circuit gains can be set according to the equation in the diagram. This allows the design to be easily set for noninverting, inverting, or differential operation.
input. The value of R tion of R
and RG to maintain the low distortion performance of
F
is determined by the parallel combina-
S
the OP176. For a more generalized treatment on circuit impedances and their effects on circuit distortion, please review the section on Active Filters at the end of the Applications section.
Driving Capacitive Loads
As with any high speed amplifier, care must be taken when driving capacitive loads. The graph in Figure 14 shows the OP176’s overshoot versus capacitive load. The test circuit is a standard noninverting voltage follower; it is this configuration that places the most demand on an amplifier’s stability. For capacitive loads greater than 400 pF, overshoot exceeds 40% and is roughly equivalent to a 45° phase margin. If the applica­tion requires the OP176 to drive loads larger than 400 pF, then external compensation should be used.
Figure 37 shows a simple circuit which uses an in-the-loop compensation technique that allows the OP176 to drive any capacitive load. The equations in the figure allow optimization of the output resistor, R optimal circuit stability. One important note is that the circuit bandwidth is reduced by the feedback capacitor, C given by:
REV. 0
, and the feedback capacitor, CF, for
X
1
BW =
2 π R
F CF
, and is
F
Figure 38. A High Speed, Low Noise Differential Line Driver
–11–
Page 12
OP176
1.0
0.1
0.010
0.001
20 100 1k 10k 20k
G = 2000
G = 200
G = 20
G = 4
±VS = ±18V
80kHz LPF
A Low Noise Microphone Preamplifier with a Phantom Power Option
Figure 39 is an example of a circuit that combines the strengths of the SSM2017 and the OP176 into a variable gain micro­phone preamplifier with an optional phantom power feature. The SSM2017’s strengths lie in its low noise and distortion, and gain flexibility/simplicity. However, rated only for 2 k or higher loads, this makes driving 600 loads somewhat limited with the SSM2017 alone. A pair of OP176s are used in the circuit as a high current output buffer (U2) and a DC servo stage (U3). The OP176’s high output current drive capability provides a high level drive into 600 loads when operating from ±18 V supplies. For a complete treatment of the circuit design details, the interested reader should consult application note AN-242, available from Analog Devices.
This amplifier’s performance is quite good over programmed gain ranges of 2 to 2000. For a typical audio load of 600 , THD + N at various gains and an output level of 10 V rms is illustrated in Figure 40. For all but the very highest gain, the THD + N is consistent and well below 0.01%, while the gain of 2000 becomes more limited by noise. The noise performance of the circuit is exceptional with a referred-to-input noise voltage spectral density of 1 nV/
Hz at a circuit gain of 1000.
Figure 40. Low Noise Microphone Preamplifier THD + N Performance at Various Gains (V
= 10 V rms and
OUT
RL = 600 Ω)
+
C8 47µF/ 63V
R9
6.81k
–IN
TO MICROPHONE
COMMON
+IN
NOTES:
1) Z1–Z4 1N752 (SEE TEXT).
2) C
3) GAIN = G = 2 x ((10k/RG ) +1) (SEE TEXT).
4) ALL RESISTORS 1% METAL FILM.
5) DOTTED PHANTOM POWER RELATED COMPONENTS OPTIONAL (SEE TEXT).
+48V
R10
100
R8
6.81k
+
R
C
1
P
IN
49.9
47µF/
63V
+
RP2
C
2
IN
49.9
47µF/
63V
, CGX LOW LEAKAGE ELECTROLYTIC TYPES (SEE TEXT).
INX
PHANTOM POWER SUPPLY CONNECTIONS, INTERLOCKED WITH +/–V
Z1
Z2
1)
RB1 10k
4.7nF/ FILM
R
2
B
10k
Z3
C
N
2200µF/ 10V
+
C
1
G
3)
R
G
C
2
G
+
2200µF/ 10V
Z4
1
(SEE NOTE 5).
S
+V
S
CRF2 100pF
7
3
8
1
5
2
4
C
1
RF
100pF
–V
S
U1
SSM2017P
6
R1
10k
R7
1k
10k
+V
S
–V
R2
4
7
C1
7
4
–V
OP176
+V
S
S
U2
R3
6
S
2
3
49.9
R4 221k
D1
1N458
D2
1N458 R5
221k
C2
1µF FILM
C5
33pF
20k
–V
S
2 3
+V
S
R6
1µF FILM
U3
OP176
6
C6
0.1µF
C7
0.1µF
+
C3 100µF/25V
+
C4 100µF/25V
OUT COMMON
+18V
–18V
OUTPUT
Figure 39. A Low Noise Microphone Preamplifier
–12–
REV. 0
Page 13
OP176
C1
100pF
R1
5.76k
R2
5.62k
R5
51
C2 100pF
R3
5.49k
R4
5.62k
R6
51
C4
0.01µF
U1
U2
C3
0.0047µF
C5
0.01µF
BALANCED
INPUTS
= AG, PIN 10 OR 18
V
IN
V
IN
+
TO AD1878/ AD1879 SIGMA­DELTA ADC L & R INPUTS
U1, U2 = OP176
–12V ANALOG
(+)
(–)
NOTES C1–C5 = NPO CERAMIC, NON-INDUCTIVE, C3-C5 CLOSE TO ADC R1–R6 = 1% METAL FILM
0.1µF
0.1µF
100µ/25V
+12V ANALOG
COM
+V
S
–V
S
TO
U1, U2
100µ/25V
(+)
USE FOR SINGLE-ENDED INPUTS
5k
5k
A Low Noise, +5 V/+10 V Reference
In many high resolution applications, voltage reference noise can be a major contributor to overall system error. Monolithic voltage references often exhibit too much wide band noise to be used alone in these systems. Only through careful filtering and buffering of these monolithic references can one realize wide­band microvolt noise levels. The circuit illustrated in Figure 41 is an example of a low noise precision reference optimized for both ac and dc performance around the OP176. With a +10 V reference (the AD587), the circuit exhibits a 1 kHz spot output noise spectral density < 10 nV/
Hz. The reference output voltage is selectable between 5 V and 10 V, depending only on the selection of the monolithic reference. The output table illustrated in the figure provides a selection of monolithic references compatible with this circuit.
OUTPUT TABLE
4
TOLERANCE (+/–mV)
5 TO 10 30 TO 100 30 TO 50 5 TO 30 2 TO 10
2.5 TO 20 15 TO 50 15 TO 25
6 5
R1
1k
R2
10k
R
TRIM
10k
(OPTIONAL)
U2
OP176
R3
100
C1 100µF/25V
C2 100µF/25V
R4
100
4
2
C3
100µF/25V
R5
1.1k
10µF/25V
6
7
3
V
OUT
R6
3.3
C5
REF COMMON
+15V
V 10V
10V 10V 10V 5V 5V 5V 5V
OUT
8
C4
0.1µF
U1 AD587
REF01 REF10 AD581 REF195 AD586 REF02 REF05
2
U1
A Differential ADC Driver
High performance audio sigma-delta ADCs, such as the stereo 16-bit AD1878 and the 18-bit AD1879, present challenging design problems with regards to input interfacing. Because of an internal switched capacitor input circuit, the ADC input structure presents a difficult dynamic load to the drive amplifier with fast transient input currents due to their 3 MHz ADC sampling rate. Also, these ADCs inputs are differential with a rated full-scale range of ±6.3 V, or about 4.4 V rms. Hence, the ADC interface circuit of Figure 42 is designed to accept a balanced input signal to drive the low dynamic impedances seen at the inputs of these ADCs. The circuit uses two OP176
Figure 41. A Low Noise, +5 V/+10 V Reference
In operation, the basic reference voltage is set by U1, either a 5 V or 10 V 3-terminal reference chosen from the table. In this case, the reference used is a 10 V buried Zener reference, but all U1 IC types shown can plug into the pinout and can be optionally trimmed. The stable 10 V from the reference is then applied to the R1-C1-C2
noise filter, which uses electrolytic capacitors for a low corner frequency. When electrolytic capacitors are used for filtering, one must be cognizant of their dc leakage current errors. Here, however, a dc bootstrap of C1 is used, so this capacitor sees only the small R2 dc drop as bias, effectively lowering its leakage current to negligible levels. The resulting low noise, dc-accurate output of the filter is then buffered by a low noise, unity gain op amp using an OP176. With the OP176’s low V the dc performance of this circuit is quite good and will not compromise voltage reference accuracy and/or drift. Also, the OP176 has a typical current limit of 50 mA, so it can provide higher output currents when compared to a typical IC reference alone.
REV. 0
and control of the source resistances,
OS
Figure 42. A Balanced Driver Circuit for Sigma-Delta ADCs
amplifiers as inverting low-pass filters for their speed and high output current drive. The outputs of the OP176s then drive the differential ADC inputs through an RC network. This RC network buffers the amplifiers against step changes at the ADC sampling inputs using one differential (C3) and two common­mode connected capacitors (C4 and C5). The 51 series resistors isolate the OP176s from the heavily capacitive loads, while the capacitors absorb the transient currents. Operating on ±12 V supplies, this circuit exhibits a very low THD + N of
0.001% at 5 V rms outputs. For single-ended drive sources, a third op amp unity gain inverter can be added between R2’s (+) input terminal and R4. For best results, short-lead, noninduc­tive capacitors are suggested for C3, C4, and C5 (which are placed close to the ADC), and 1% metal-film types for R1 through R6. For surface mount PCBs, these components can be NPO ceramic chip capacitors and thin-film chip resistors.
–13–
Page 14
OP176
An RIAA Phono Preamp
Figure 43 illustrates a simple phono preamplifier using RIAA equalization. The OP176 is used here to provide gain and is chosen for its low input voltage noise and high speed perfor­mance. The feedback equalization network (R1, R2, C1, and C2) forms a three time constant network, providing reasonably accurate equalization with standard component values. The input components terminate a moving magnet phono cartridge as recommended by the manufacturer, the element values shown being typical. When this ac coupled circuit is built with a low noise bipolar input device such as the OP176, amplifier bias current makes direct cartridge coupling difficult. This circuit uses input and output capacitor coupling to minimize biasing interactions.
Input ac coupling to the amplifier is provided via C5, and the low frequency termination resistance, R
, is the parallel equiva-
T
lent of R6 and R7. R3 of the feedback network is ac grounded via C4, a large value electrolytic. Additionally, this resistor is set to a low value to minimize circuit noise from nonamplifier sources. These design measures reduce the dc offset at the output of the OP176 to a few millivolts. The output coupling network of C3 and R4 is shown as suitable for wide band response, but it can be set to a 7950 µs time constant for use as a 20 Hz rumble filter.
The 1 kHz gain (“G”) of this circuit, controlled by R3, is calculated as:
G (@ 1 kHz) = 0.101 × 1 +
R1
R3
For an R3 of 200 Ω, the circuit gain is just under 50 × (≈ 34 dB), and higher gains are possible by decreasing R3. For any value of R3, the R5-C6 time constant should be equal to R3 and the series equivalent of C1 and C2.
Using readily available standard values for network elements (R1, R2, C1, and C2) makes the design easily reproducible and inexpensive. These components are ideally high quality precision types, for low equalization errors and minimum
parasitics. One percent metal-film resistors and two percent film capacitors of polystyrene or polypropylene are recom­mended. Using the suggested values, the frequency response relative to the ideal RIAA characteristic is within ± 0.2 dB over 20 Hz–20 kHz. Even tighter response can be achieved by using the alternate values, shown in brackets “[ ],” with the trade-off of a non off-the-shelf part.
As previously mentioned, the OP176 was chosen for three reasons: (1) For optimal circuit noise performance, the amplifier used should exhibit voltage and current noise densities of 5 nV/
Hz and 1 pA/Hz, respectively. (2) For high gain
accuracy, especially at high stage gains, the amplifier should exhibit a gain bandwidth product in excess of 5 MHz. (3) Equally important because of the 100% feedback through the network at high frequencies, the amplifier must be unity gain stable. With the OP176, the circuit exhibits low distortion over the entire range, generally well below 0.01% at outputs levels of
rms using ±18 V supplies. To achieve maximum perfor-
5 V mance from this high gain, low level circuit, power supplies should be well regulated and noise free, and care should be taken with shielding and conductor layout.
Active Filter Circuits Using the OP176
A general active filter topology that lends itself to both high-pass (HP) and low-pass (LP) filters is the well known Sallen-Key (SK) VCVS (Voltage-Controlled, Voltage Source) architecture. This filter type uses the op amp as a fixed gain voltage follower at either unity or a higher gain. Discussed here are simplified 2­pole, unity gain forms of these filters, which are attractive for several reasons: One, at audio frequencies, using an amplifier with a 10 MHz bandwidth such as the OP176, these filters exhibit reasonably low sensitivities for unity gain and high damping (low Q). Second, as voltage followers, they are also inherently gain accurate within their pass band; hence, no gain resistor scaling errors are generated. Third, they can also be made “dc accurate,” with output dc errors of only a few millivolts. The specific filter response in terms of HP, LP and damping is determined by the RC network around the op amp, as shown in Figure 44a.
MOVING MAGNET PICKUP
100kΩ
+V
S
1%
100µF
100µF
100µF/25V
C1
0.03µF 2%
R2
0.1µF
0.1µF
–V
S
7
6
4
100k
[97.6k
R1 1%
]
S
[7.87k]
8.25k
R3
C4 1000µF/16V
C5
100µF/25V
R6
C
t
150pF
Rt = R6| |R7 ~ 50k
R7
100k
+V
3
U1
OP176
2
–V
S
200Ω (34dB)100Ω (40dB)
C3
C2
0.01µF 2%
100k
R4
+18V
–18V
R5
499
V
C6 3nF
OUT
Figure 43. An RIAA Phono Preamplifier Circuit
–14–
REV. 0
Page 15
OP176
100 50k 10k 1k 20
10.000
–30.00
–70.00
–50.00
–10.00
–20.00
–40.00
–60.00
0.0
LP
HP
FREQUENCY – Hz
dBr
High Pass Sections
Figure 44a illustrates the high-pass form of a 2-pole SK filter using an OP176. For simplicity and practicality, capacitors C1 and C2
are set equal (“C”), and resistors R2 and R1 are
adjusted to a ratio, N, which provides the filter damping coefficient, α, as per the design expressions. This high pass design is begun with selection of standard capacitor values for C1 and C2 and a calculation of N. The values for R1 and R2 are then determined from the following expressions:
R1
=
π×
2
1
FREQ×C×N
and
R2=N×R1
C1
0.01µF
C2
0.01µF
R2
22k
(22.508k)
OP176
3 2
Z
COMP
IN (–)
R1
11k
(11.254k)
+V
S
7
6
4
–V
S
Z
COMP
(HIGH PASS)
OUTPUT
R2
C2
C1
R1
OUT IN
GIVEN: α, FREQ SET C1 = C2 = C
1
2
α = =
R1 =
R2 = N x R1 1 kHz BW SHOWN
Q
N
4
R2
N = =
2
R1
α
2 π FREQ x C x N
1
Low Pass Sections
In the LP SK arrangement of Figure 44b, the R and C elements are interchanged where the resistors are made equal. Here, the ratio of C2/C1 (“M”) is used to set the filter α, as noted. Otherwise, this filter is similar to the HP section, and the resulting 1 kHz LP response is shown in Figure 45. The design begins with a choice of a standard capacitor value for C1 and a calculation of M. This then forces a value of “M × C1” for C2. Then, the value for R1 and R2 (“R”) is calculated according to the following equation:
1
FREQ×C1×M
C1
0.02µF
S
6
S
R1
GIVEN: α, FREQ
α
= =
M = =
CHOOSE C1
C2 = M x C1
R =
1 kHz BW SHOWN
+V
7
4
–V
Z
COMP
(LOW PASS)
OUTPUT
C2
C1
OUT IN
1
2
Q
M
4
C2
2
C1
α
1
2 π FREQ x C1 x M
R1
11k
(11.254k)
R2
11k
(11.254k)
C2
0.01µF
R
=
2
π×
OP176
3 2
Z
COMP
IN (–)
R2
Figures 44a. Two-Pole Unity Gain HP/LP Active Filters
In this examples, circuit α (or 1/Q) is set equal to 2, providing a Butterworth (maximally flat) characteristic. The filter corner frequency is normalized to 1 kHz, with resistor values shown in both rounded and (exact) form. Various other 2-pole response shapes are possible with appropriate selection of α, and fre­quency can be easily scaled, using inversely proportional R or C values for a given α. The 22 V/µs slew rate of the OP176 will support 20 V p-p outputs above 100 kHz with low distortion. The frequency response resulting with this filter is shown as the dotted HP portion of Figure 45.
Figures 44b. Two-Pole Unity Gain HP/LP Active Filters
Figure 45. Relative Frequency Response of 2-Pole, 1 kHz Butterworth LP (Left) and HP (Right) Active Filters
REV. 0
–15–
Page 16
OP176
Passive Component Selection for Active Filters
The passive components suitable for active filters deserve more than casual attention. Resistors should be 1%, low TC, metal­film types of the RN55 or RN60 style. Capacitors should be 1% or 2% film types preferably, such as polypropylene or polysty­rene, or NPO (COG) ceramic for smaller values.
Active Filter Circuit Subtleties
In designing active filter circuits with the OP176, moderately low values (10 k or less) for R1 and R2 can be used to minimize the effects of Johnson noise when critical. The practical tradeoff is, of course, capacitor size and expense. DC errors will result for larger values of resistance, unless compen­sation for amplifier input bias current is used. To add bias compensation in the HP filter section of Figure 42a, a feedback compensation resistor equal to R2 can be used. This will minimize bias current induced offset to the product of the OP176’s I
and R2. For an R2 of 25 k, this produces a typical
OS
compensated offset voltage of 50 µV. Similar compensation is applied to Figure 42b, using a resistance equal to R1+ R2. Using dc compensation, filter output dc errors using the OP176 will be dominated by its V
, which is typically 1 mV or less. A
OS
caveat here is that the additional resistors can increase noise substantially. For example, a 10 k resistor generates ~ 12 nV/
Hz of noise and is about twice that of the OP176. These
resistors can be ac bypassed to eliminate their noise using a simple shunt capacitor chosen such that its reactance (X
C
) is
much less than R at the lowest frequency of interest. A more subtle form of ac degradation is also possible in these
filters, namely nonlinear input capacitance modulation. This issue was previously covered for general cases in the section on minimizing distortion. In active filter circuits, a fully compen­sating network (for both dc and ac performance) can be used to minimize this distortion. To be most effective, this network
) should include R1 through C2 as noted for either filter
(Z
COMP
type, of the same style and value as their counterparts in the forward path. The effects of a Z
network on the THD + N
COMP
performance of two 1 kHz HP filters is illustrated in Figure 46. One filter (A) is the example shown in Figure 44a (Curves A1 and A2), while the second (B) uses RC values scaled 10 times upward in impedance (Curves B1 and B2). Both filters operate with a 2 V rms input, ±18 V supplies, 100 k loading, and analyzer bandwidth of 80 kHz.
1
0.1
0.010
THD +N – %
0.001
0.0001 20 100 20k 1k
FREQUENCY – Hz
B1
A1 B2
A2
10 k
Figure 46. THD + N (%) vs. Frequency for Various 1 kHz HP Active Filters Illustrating the Effects of the Z
Curves A1 and B1 show performance with Z while curves A2 and B2 illustrate operation with Z
COMP
Network
COMP
shorted,
COMP
active. For the “A” example values, distortion in the pass band of 1 kHz–20 kHz is below 0.001% compensated, and slightly higher uncompensated. With the higher impedance “B” net­work, there is a much greater difference between compensated and uncompensated responses, underscoring the sensitivity to higher impedances. Although the positive effect of Z
COMP
is seen for both “A” and “B” cases, there is a buffering effect which takes place with lower impedances. As case “A” shows, when using larger capacitance values in the source, the amplifier’s nonlinear C-V input characteristics have less effect on the signal.
Thus, to minimize the necessity for the complete Z
COMP
com-
pensation, effective filter designs should use the lowest capaci­tive impedances practical, with an 0.01 µF lower value limit as a goal for lowest distortion (while lower values can certainly be used, they may suffer higher distortion without the use of full compensation). Since most designs are likely to use low relative impedances for reasons of low noise and offset, the effects of CM distortion may or may not actually be apparent to a given application.
–16–
REV. 0
Page 17
OP176
97
E
I
1
P
–IN
+IN
98
CM2
2
C
IN
I
OS
1
Q1
D1
36
D2
CM1
98
R
5
7
EN E
5
R
3
4
R
6
8
9
Q2
OS
3
35
V
N1
D
N1
10
D
N2
V
N2
11
12
V
N3
D
N3
13
C
D
N4
V
N4
N1
14
15
V
N5
D
N5
16
D
N6
V
N6
C
1
17
6
C
2
R
4
E
M
97
V
2
19
D
3
G
1
R
7
G
C
2
3
98
21
R
9
R
G
22
8
C
4
23
3
R10C
G
5
24
4
R11C
E
6
C7
25
R
2
26
12
R
13
98
D
4
20
V
3
E
REF
51
99
G6
33
G7
D8
G8
V
4
F
1
V
5
F
2
G9
D10
R
17
L
2
29
34
R
18
D7
R
15
I
SY
28
D5
30
27
G5
C8
R
14
98
R
16
C
9
31
D6
32
D9
50
Figure 47. OP176 Spice Model Schematic
REV. 0
–17–
Page 18
OP176
OP176 SPICE Model
* * Node Assignments * Noninverting Input * | Inverting Input * | | Positive Supply * | | | Negative Supply * | | | | Output *||||| *||||| .SUBCKT OP176 1 2 99 50 34 * * INPUT STAGE & POLE AT 100 MHz * R3 5 51 2.487 R4 6 51 2.487 CIN 1 2 3.7E-12 CM1 1 98 7.5E-12 CM2 2 98 7.5E-12 C2 5 6 320E-12 I1 97 4 100E-3 IOS 1 2 1E-9 EOS 9 3 POLY(1) (26,28) 0.2E-3 1 Q1 5 2 7 QX Q2 6 9 8 QX R5 7 4 1.970 R6 8 4 1.970 D1 2 36 DZ D2 1 36 DZ EN 3 1 (10,0) 1 GN1 0 2 (13,0) 1E-3 GN2 0 1 (16,0) 1E-3 * EREF98 0 (28,0) 1 EP 97 0 (99,0) 1 EM 51 0 (50,0) 1 * * VOLTAGE NOISE SOURCE * DN1 35 10 DEN DN2 10 11 DEN VN1 35 0 DC 2 VN2 0 11 DC 2 * * CURRENT NOISE SOURCE * DN3 12 13 DIN DN4 13 14 DIN VN3 12 0 DC 2 VN4 0 14 DC 2 * * CURRENT NOISE SOURCE * DN5 15 16 DIN DN6 16 17 DIN VN5 15 0 DC 2 VN6 0 17 DC 2 * * GAIN STAGE & DOMINANT POLE AT 32 Hz * R7 18 98 1.243E6 C3 18 98 4E-9 G1 98 18 (5,6) 4.021E-1 V2 97 19 1.35 V3 20 51 1.35 D3 18 19 DX D4 20 18 DX *
* POLE/ZERO PAIR AT 1.5 MHz/2.7 MHz * R8 21 98 1E3 R9 21 22 1.25E3 C4 22 98 47.2E-12 G2 98 21 (18,28) 1E-3 * * POLE AT 100 MHz * R10 23 98 1 C5 23 98 1.59E-9 G3 98 23 (21,28) 1 * * POLE AT 100 MHz * R11 24 98 1 C6 24 98 1.59E-9 G4 98 24 (23,28) 1 * * COMMON-MODE GAIN NETWORK WITH ZERO AT 1 kHz * R12 25 26 1E6 C7 25 26 60E-12 R13 26 98 1 E2 25 98 POLY(2) (1,98) (2,98) 0 2.50 2.50 * * POLE AT 100 MHz * R14 27 98 1 C8 27 98 1.59E-9 G5 98 27 (24,28) 1 * * OUTPUT STAGE * R15 28 99 58.333E3 R16 28 50 58.333E3 C9 28 50 1E-6 ISY 99 50 1.743E-3 R17 29 99 100 R18 29 50 100 L2 29 34 1E-9 G6 32 50 (27,29) 10E-3 G7 33 50 (29,27) 10E-3 G8 29 99 (99,27) 10E-3 G9 50 29 (27,50) 10E-3 V4 30 29 1.74 V5 29 31 1.74 F1 29 0 V4 1 F2 0 29 V5 1 D5 27 30 DX D6 31 27 DX D7 99 32 DX D8 99 33 DX D9 50 32 DY D10 50 33 DY * * MODELS USED * .MODEL QX PNP(BF=5E5) .MODEL DX D(IS=1E-12) .MODEL DY D(IS=1E-15 BV=50) .MODEL DZ D(IS=1E-15 BV=7.0) .MODEL DEN D(IS=1E-12 RS=4.35K KF=1.95E-15 AF=1) .MODEL DIN D(IS=1E-12 RS=268 KF=1.08E-15 AF=1) .ENDS OP176
–18–
REV. 0
Page 19
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
8-Lead Plastic DIP (N-8)
OP176
PIN 1
0.210
(5.33)
MAX
0.160 (4.06)
0.115 (2.93)
0.022 (0.558)
0.014 (0.356)
0.0098 (0.25)
0.0040 (0.10)
8
1
0.430 (10.92)
0.348 (8.84)
0.100 (2.54)
BSC
5
4
0.070 (1.77)
0.045 (1.15)
0.280 (7.11)
0.240 (6.10)
0.060 (1.52)
0.015 (0.38)
0.130 (3.30) MIN
SEATING PLANE
8-Lead Narrow-Body SO (SO-8)
5
4
0.0192 (0.49)
0.0138 (0.35)
0.1574 (4.00)
0.1497 (3.80)
0.2440 (6.20)
0.2284 (5.80)
0.0688 (1.75)
0.0532 (1.35)
0.0098 (0.25)
0.0075 (0.19)
PIN 1
8
1
0.1968 (5.00)
0.1890 (4.80)
0.0500 (1.27)
BSC
0.325 (8.25)
0.300 (7.62)
0.015 (0.381)
0.008 (0.204)
0.0196 (0.50)
0.0099 (0.25)
8° 0°
0.195 (4.95)
0.115 (2.93)
x 45°
0.0500 (1.27)
0.0160 (0.41)
REV. 0
–19–
Page 20
C1878–10–1/94
–20–
PRINTED IN U.S.A.
Page 21
FOR CATALOG
ORDERING GUIDE
Model Temperature Range Package Description Package Option*
OP176GP –40°C to +85°C 8-Pin Plastic DIP N-8 OP176GS –40°C to +85°C 8-Pin SOIC SO-8 OP176GSR –40°C to +85°C SO-8 Reel, 2500 Pieces OP176GBC +25°C DICE
*For outline information see Package Information section.
OP176
REV. 0
–21–
Loading...