ANALOG DEVICES Low Noise, Precision Operational Amplifier OP-27 FEATURES @ Low Noise .................... 80nV,., (0.1Hz to 10Hz) ee eee ee eee ee teen een teen eens 3nV/\V/ Hz @ Low Drift 2.2.0... ccc eee eee eee 0.2nV/C @ High Speed ........................ 2.8V/ys Slew Rate eee ee en eb renee eee enes 8MHz Gain Bandwidth @ LOW Vog occ cece cece cette tne e eens 10uV @ Excellent CMRR ............... 126dB at Voy of +11V High Open-Loop Gain ..................... 1.8 Million e Fits 725, OP-07, OP-05, AD510, AD517, 5534A sockets Available in Die Form ORDERING INFORMATION ' PACKAGE T,=+25C OPERATING Vos MAX CERDIP PLASTIC LCC TEMPERATURE (nV) TO-99 8-PIN 8-PIN 20-CONTACT RANGE 25 OP27Au" OP27AZ* ~ - MIL 25 OP27EJ OP27EZ OP27EP - iND/COM 60 OP27BJ* OP27BZ" - OP27BR/883 MIL 60 OP27FJ OP27FZ OP27FP - IND/COM 100 OP27CJ OP27CZ - - MIL 100 OP27GJ OP27GZ OP27GP - XIND 100 - - op27astt - XIND * For devices processed in total compliance to MIL-STD-883, add /883 after part number. Consult factory for 883 data sheet. t Burn-in is available on commercial and industrial temperature range parts in CerDIP, plastic DIP, and TO-can packages. tt For availability and burn-in information on SO and PLCC packages, contact your local sales office. GENERAL DESCRIPTION The OP-27 precision operational amplifier combines the low offset and drift of the OP-07 with both high speed and tow noise. Offsets down to 25nV and drift of 0.6unV/ C maximum make the OP-27 ideal for precision instrumentation applica- signals. A gain-bandwidth product of 8MHz and a 2.8V/ysec slew rate provides excellent dynamic accuracy in high-speed data-acquisition systems. A low input bias current of +10nA is achieved by use of a bias-current-cancellation circuit. Over the military temper- ature range, this circuit typically holds I, and Iog to +20nA and 15nA respectively. The output stage has good load driving capability. A guaran- teed swing of + 10V into 60020 and low output distortion make the OP-27 an excellent choice for professional audio applica- tions. PSRR and CMRR exceed 120dB. These characteristics, coupled with long-term drift of 0.2.V/month, allow the circuit designer to achieve performance levels previously attained only by discrete designs. PIN CONNECTIONS BAL 8 BAL1 7 Vt -IN2 6 OUT HIN3 SNC, 8-PIN HERMETIC DIP 4.V- (CASE) (Z-Suffix) TO-99 EPOXY MINI-DIP (J-Suffix) (P-Suffix) 8-PINSO (S-Suffix) OP-27BRC/883 LCC PACKAGE tions. Exceptionally low noise, en = 3.5nV/\/ Hz , at 10Hz, a (RC-Suffix) low 1/f noise corner frequency of 2.7Hz, and high gain (1.8 million), allow accurate high-gain amplification of low-level SIMPLIFIED SCHEMATIC Ve Sr3 Ra ; 2 7 @ ns 4 Vos Adi. L {cme a, 46 > RY eR2 R23 R24 v 21 Ro Liebe aby] 20} arg q P-0 OUTPUT NON- R12 ere ce [ors . RS Ril C4 INVERTING AW TH {Koss INPUT {-) kaze Qi ai2 N 027 a28 * R1&R2 ARE PERMANENTLY ADJUSTED AT WAFER TEST FOR MINIMUM OFFSET VOLTAGE. 6 OYOP-27 Low cost, high-volume production of OP-27 is achieved by using an on-chip zener-zap trimming network. This reliable and stable offset trimming scheme has proved its effective- ness over many years of production history. The OP-27 provides excellent performance in low-noise high-accuracy amplification of low-level signals. Applica- tions include stable integrators, precision summing ampli- fiers, precision voltage-threshoid detectors, comparators, and professional audio circuits such as tape-head and microphone preamplifiers. The OP-27 is a direct replacement for 725, OP-06, OP-07 and OP-05 amplifiers; 741 types may be directly replaced by removing the 741s nulling potentiometer. ABSOLUTE MAXIMUM RATINGS (Note 4) Supply Voltage 00... sees ssseseceeeseacsssssenseesecseaene eens +22V Input Voltage (Note 1) oo... ccsssssessesseresseesensesesscsenenese +22V Output Short-Circuit Duration .......... ce ceeeseeeeercenerens Indefinite Differential Input Voltage (Note 2)... ce seseeeeeeenernees +0.7V Differential Input Current (Note 2) oo... eee eseeeeeeneees +25mA Storage Temperature Range ...............06 .. -65C to +150C Operating Temperature Range OP-27A, OP-27B, OP-27C (J, Z, RC)........ -55C to +125C OP-27E, OP-27F (Uy Z) eessssssssssssssssssssseseee -25C to +85C OP-27E, OP-27F (P) esssssssssssssssssssssstssssssssseees 0C to +70C OP-27G (P, S.J, Z) cescesessssssssssssssssssssnssssusness 40C to +85C Lead Temperature Range (Soldering, 60 sec) .............. 300C Junction Temperature 0.0... eeeeeeseeeeeeeees -65C to +150C PACKAGE TYPE @,_ (Note 3) ic UNITS TO-99 (J) 150 18 CAV 8-Pin Hermetic DIP (Z) 148 16 CW 8-Pin Plastic DIP (P) 103 43 C/W 20-Contact LCC (RC) 98 38 C/W 8-Pin SO (S) 158 43 C NOTES: 1. For supply voltages less than +22V, the absolute maximum input voltage is equal to the supply voltage. The OP-27's inputs are protected by back-to-back diodes. Current limiting resistors are not used in order to achieve low noise. If differential input voltage exceeds +0.7V, the input current should be limited to 25mA. @., is specified for worst case mounting conditions, i.e., @., is specified for device in socket for TO, CerDIP, P-DIP, and LCC packages; 8 A is specified for device soldered to printed circuit board for SO package. Absolute maximum ratings apply toe both DICE and packaged parts, unless otherwise noted. 2. ELECTRICAL CHARACTERISTICS at Vo = +15V, Ta = 25C, unless otherwise noted. OP-27A/E OP-27B/F OP-27C/G PARAMETER SYMBOL CONDITIONS MIN TYP MAX MIN TYP MAX MIN TYP MAX UNITS Input Offset Voltage Vog (Note 1) _ 10 25 _ 20 60 _ 30 100 BV Long-Term Vog . Vos/T! Notes 2, 3 _ 0.2 1.0 _ 0.3 16 _ 0.4 2.0 V/M Stability os/Time (Notes 2, 3) eva Input Offset Current log _ 7 35 _ 9 50 _ 12 75 nA Input Bias Current Ip _ +10 +40 _ +12 55 _ +15 +80 nA 0.1Hz to 10Hz Input Noise Voltage = @np_ - 0.08 0.18 - 0.08 0.18 - 0.09 0.25 Vp- p 92 &np-p (Notes 3, 5) BNPSP input Noise fo = 10Hz (Note 3) _ 3.5 5.5 _ 3.5 5.5 _ 3.8 8.0 e. fo = 30Hz (Note 3 _ 3.4 45 _ 34 4.5 - 3.3 5.6 V/V H Voitage Density a o = SOHz (Note 3) " 2 fg = 1000Hz (Note 3) _ 3.0 3.8 _ 3.0 3.8 _ 3.2 4.5 . fo = 10Hz (Notes 3,6) - 1.7 4.0 _ 1.7 40 _ 1.7 =- Input Noise fo = 30Hz (Notes 3, 6) 1.0 23 1.0 23 1.0 Af He Current Density n oT ( , ~ . . . , FP fo = 1000Hz (Notes 3, 6) - 0.4 0.6 - 0.4 0.6 _ 0.4 0.6 Input Resistance R 7 1. 6 - 0.94 5 - 0.7 4 _ MQ Differential-Mode | /N (Note 7) 8 Input Resistance R _ 3 _ _ 2.5 - _ 2 _ Go Common-Mode INCM Input Voltage Range iVR 110 +123 _ +110 =12.3 _ +110 =+12.3 V Cc -Mod ommon-Nloge CMRR Voy=ET1V 114 126 406 428 _ 100 120 dB Rejection Ratio Power Supp! ver SUPP y PSRR s- Vg = 44V to + 18V _ 1 40 _ 1 10 _ 2 20 uv Rejection Ratio Large-Signal A R,_ 2 2k9, Vo = +10V 1000 1800 _ 1000 1800 - 700 1500 _ vim Voltage Gain vo Ry 2 6000, Vo = +10V 800 =: 1500 800 = 1500 - 600 1500 _ OQutput Voltage Vv R_ 2ekon +12.0 +13.8 ~ +12.0 +13.8 _ 4114.5 413.5 _ V Swing o R, = 6000 +10.0 11.56 _ 10.0 211.5 _ 10.60 +115 - Slew Rate SR R, 2 2k (Note 4) 1.7 2.8 _ 7 2.8 _ 17 2.8 _ V/sOP-27 ELECTRICAL CHARACTERISTICS at Vs = +15V, Ta = 25C, unless otherwise noted. (Continued) OP-27A/E OP-27B/F OP-27C/G PARAMETER SYMBOL CONDITIONS MIN TYP MAX MIN TYP MAX MIN TYP MAX UNITS Gain Bandwidth Prod. GBW : (Note 4} 5.0 8.0 _ 5.0 8.0 _ 5.0 8.0 _ MHz Open-Loop Output = = _ 70 - _ 70 - Resistance Yo=0lo=0 70 a Power Consumption Py Vo , _ 90 140 - 90 140 _ 100 170 mw Offset Adjustment : Rp = 10kn _ +4.0 _ _ +40 - = +4.0 _ mV Range . NOTES: days are typically 2.54V refer to typical performance curve. 1. Input offset voltage measurements are performed ~ 0.5 seconds after 3. Sample tested. application of power. A/E grades guaranteed fully warmed-up. 4. Guaranteed by design. ; 2. Long-term input offset voltage stability refers to the average trend line of 5. See test circuit and frequency response curve for 0.1Hz to 10Hz tester. Vog vs. Time over extended periods after the first 30 days of operation. 6. See test circuit-for current noise measurement. Excluding the initial hour of operation, changes in Vog during the first 30 7. Guaranteed by input bias current. ELECTRICAL CHARACTERISTICS for Vs = + 15V, -55C < Ta S + 125C, unless otherwise noted. OP-27A OP-27B _ OP-27C PARAMETER SYMBOL CONDITIONS MIN TYP MAX MIN TYP MAX MIN TYP MAX UNITS Input Offset Voitage Vog (Note 1) _ 30 60 ~ 50 200 _ 70 300 BV Average Input TCVos (Note 2) - .2 . . . ~ 4 1.8 Offset Drift TCVogn _(Note 3) os (08 18 0 WiC Input Offset Current log - 15 50 - 22 85 _ 30 135 nA Input Bias Current Ip _ +20 +60 - +28 95 - +35 +150 nA Input Voltage Range IVR 410.3 #115 _ 10.3 115 _ +10.2 +11.5 = v c -Mod ommonoee CMAR Vom = +10V 108 122 100 119 - 4. 116 dB Rejection Ratio Pi i ower Supply PSAR Vg = 24.5V to +18V _ 2 16 _ 2 20 _ 4 51 LVN Rejection Ratio Large-Signal - A Rx, = 2kQ, Vo= +10V 600 1200 _ 500 1000 ~ 300 800 ~ V/mV Voltage Gain vo L 9 Output Voltage eee swing 9 Vo Ri 2 2kn 4115 185 - +110 132 - +105 +13.0 - v ELECTRICAL CHARACTERISTICS at V, =+15V,-25C 2k +17 +136 - +14 135 ~ t11.0 +133 _ v Swing NOTES: 2. The TCVog performance is within the specifications unnulled or when 1. Input offset voltage measurements are performed by automated test nulled with Rp = 8k to 20k. TCVog is 100% tested for A/E grades, equipment approximately 0.5 seconds after application of power. A/E sample tested for B/C/F/G grades. grades guaranteed fully warmed-up. 3. Guaranteed by design.OP-27 DICE CHARACTERISTICS NULL . (-) INPUT (+) INPUT y- OUTPUT V+ NULL PND Pwr s DIE SIZE 0.109 x 0.055 inch, 5995 sq. mils (2.77 X 1.40mm, 3.88 sq. mm) WAFER TEST LIMITS at Vs = +15V, Ta = 25C for OP-27N, OP-27G, and OP-27GR devices; Ta = 125C for OP-27NT and OP-27GT devices, unless otherwise noted. OP-27N OP-27GT OP-27G OP-27GR OP-27NT PARAMETER SYMBOL CONDITIONS LIMIT LIMIT LIMIT LIMIT LIMIT UNITS Input Offset Voltage Vos (Note 1) 60 35 200 60 100 uV MAX Input Offset Current los 50 35 85 50 76 nA MAX Input Bias Current Ig +60 +40 +95 +55 +80 nA MAX Input Voltage Range IVR +10.3 n +10.3 +n +11 V MIN Common-Mode , = , 1 4 Rejection Ratio CMRR Vom= IVR 08 1 100 106 100 dB MiN Power Supply . =t +18V _ 10 _ 1 Rejection Ratio PSRR Vg 4V to +18 0 20 uV/V MAX Large-Signal RL 2k0,Vo=+10V 600 ~S=WN000 500 1000 700 . V/A Voltage Gain Avo R, = 6000, Vo=+10V - 800 = 800 600 mv MIN ae - Ri 2 2kn : +115 +12.0 11.0 120 , 11.5 a Output Voltage Swing Vo R= 6000 _ +10.0 _ +100 +100 V MIN Power Consumption 140 _ 140 170 mW MAX NOTE: Py Vo= 0 _ Electrical tests are performed at 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 qualification through sample lot assembly and testing. TYPICAL ELECTRICAL CHARACTERISTICS at Vs = + 15V, Ta = +25C, unless otherwise noted. OP-27N OP-27G OP-27GR PARAMETER SYMBOL CONDITIONS TYPICAL TYPICAL TYPICAL UNITS Average Input Offset TCVog or Nulled or Unnulled 0.2 Re R e Voltage Drift TCVosn Rp = 8k0 to 20k 04 wie Average Input Offset Current Drift TClog 80 130 180 PAG Average Input Bias 5 Current Drift TClp 100 160 200 pA/ec Input Noise fo = 10HzZ 3.5 3.5 3.8 Voltage Densit en {9 = 30Hz 3.1 341 3.3 nV// Hz $ y fg = 1000Hz 3.0 3.0 3.2 Input Noise fo = 10Hz 17 WwW 17 current eneit in fo= 30Hz 1.0 1.0 1.0 pA// Hz y fo = 1000Hz 0.4 0.4 04 Input Noise Voltage @np-p 0.1Hz to 10Hz 0.08 0.08 0.09 uVp-p Slew Rate SR R, = 2k 2.8 28 2.8 V/ps Gain Bandwidth Product GBW 8 8 8 MHz NOTE: 1. Input offset voltage measurements are performed by automated test equipment approximately 0.5 seconds after application of power.OP-27 TYPICAL PERFORMANCE CHARACTERISTICS 0.1Hz TO 10Hzy., NOISE TESTER FREQUENCY RESPONSE 100 90 GAIN (dB) S$ 3s 8 50 TEST TIME OF 10 SEC FURTHER LIMITS LOW FREQUENCY (<0.1Hz) 40 GAIN 30 0.01 0.1 1.0 10 100 FREQUENCY (Hz) INPUT WIDEBAND VOLTAGE NOISE vs BANDWIDTH (0.1Hz TO FREQUENCY INDICATED) 10 Ta = 25C Vs = +15V S 7 lo 1 o 2 Ww QQ = EF a Qo > 4 9.1 = a 0.01 100 1k 10k 100k BANDWIDTH (Hz) VOLTAGE NOISE DENSITY vs SUPPLY VOLTAGE . | Ta = 25C iS 4 > AT 10Hz : pe a - AT 1kHz o 3 2 Ww 9 < FE J Q2 1 oO 40 20 30 40 TOTAL SUPPLY VOLTAGE (V+ V) (VOLTS) VOLTAGE NOISE DENSITY vs FREQUENCY Ta = 25C Vs = +15V OB VoOod \/f CORNER = 2.7Hz 2 VOLTAGE NOISE (nVA/Hz; 1 10 100 1000 FREQUENCY (Hz) TOTAL NOISE vs SOURCE RESISTANCE Ta = 25C a1 Vg = #15V ne Hz) AT 19Hz TOTAL NOISE (nVA AT 1kHz RESISTOR NOISE ONLY 100 1k 10k SOURCE RESISTANCE (2) CURRENT NOISE DENSITY vs FREQUENCY CURRENT NOISE (pAA/Hz} \/f CORNER = 140Hz 10 100 1k 10k FREQUENCY (Hz) A COMPARISON OF OP AMP VOLTAGE NOISE SPECTRA 741 Hz) Vf CORNER LOW NOISE AUD 10 If CORNER OP AMP 2.7Hz if CORNER OP-27 VOLTAGE NOISE (nV INSTRUMENTATION AUDIO RANGE To oc TO 10 100 1000 FREQUENCY (Hz) VOLTAGE NOISE DENSITY vs TEMPERATURE 5 | T Vs = t15V 4 AT 10Hz z Ly Ln 4 | 33 AT kHz Fo oe 5 3 >2 1 -50 -26 0 25 50 75 100 125 TEMPERATURE (C} SUPPLY CURRENT vs SUPPLY VOLTAGE 5.0 < 0 Ta = 125C Ee |_ 2 WwW f= 30} Ta = 25C = > a & a2 a 20 ia | Ta = -55C 10 | 5 15 25 35 45 TOTAL SUPPLY VOLTAGE (VOLTS)OP-27 TYPICAL PERFORMANCE CHARACTERISTICS OFFSET VOLTAGE DRIFT OF LONG-TERM OFFSET EIGHT REPRESENTATIVE UNITS VOLTAGE DRIFT OF SIX vs TEMPERATURE REPRESENTATIVE UNITS 60 Nw L 5 o oN FO OFFSET VOLTAGE (uV} & o ~40 F TRIMMING WITH 10k POT DOES CHANGE IN OFFSET VOLTAGE (xV) CHANGE IN INPUT OFFSET VOLTAGE (xuV) NOT CHANGE 60 TCVos 4 OP-27C 6 -75-50 -25 0 25 50 75 100 125 150 175 0 1 2 3 4 5 6 7 TEMPERATURE (C) TIME (MONTHS} OFFSET VOLTAGE CHANGE _ INPUT BIAS CURRENT _ DUE TO THERMAL SHOCK vs TEMPERATURE = & 30 50 w 2 Vg =+15V 25 8 40 Vg=etsv LJ Z aq & ry = IN E 2 20 5 \ a 5 FTN op.2ze f be oc N \ . 2 > 15 3 5 2 TN yi 2 2 20 \ N ~ 44 if = a w 10 bE NI NON OP-27B 3 Q 5 NL NSS E Zz a m.| Ms] ap > z 2 nd Poa] T a Bo 5 10 = Ww 5 DEVICE IMMERSED OP-27A 2 IN 70C OIL BATH | | | H 0 a @ 20 0 20 40 60 30 = 100 -50 -25 0 2 50 75 100 125 150 TIME (SEC) TEMPERATURE (C) SLEW RATE, GAIN-BANDWIDTH OPEN-LOOP GAIN vs PRODUCT, PHASE MARGIN vs FREQUENCY TEMPERATURE = 10 6 WwW a z Vs = 15V z c = a << 9 F a = S = iu 2 2 4 3 a zt x c a] ao a a = w 8x 2 g 3 a 6 - 8 S 8 = Q > 2 2 tu a & Tz a 4 z o Wd oad wo ~10 6 10 1 10 100 1k 10k 100k 1M 10M 100M 75 50-25. 0 25 50 75 100 125 1M FREQUENCY (Hz) TEMPERATURE (@C) WARM-UP OFFSET VOLTAGE DRIFT Ta = 425C Vg = 15V = o OP-27 C/G OP-27 B/F a OP-27 A/E Qo 1 2 3 4 TIME AFTER POWER ON (MINUTES) INPUT OFFSET CURRENT vs TEMPERATURE b a w S Ny = o 0 -75 -50 ~-25 0 25 50 75 100 TEMPERATURE (C) GAIN, PHASE SHIFT vs FREQUENCY Ta = 25C Vg = 15V Vg = +15V 80 100 120 140 160 180 200 220 10M 100M FREQUENCY (Hz) 12 BHAGE GHIFT [NFROP-27 TYPICAL PERFORMANCE CHARACTERISTICS OPEN-LOOP VOLTAGE GAIN vs SUPPLY VOLTAGE 25 T RL = 2kQ 5 fTA= 25C 7 AL = sion 2, LZ g 4 8 L 3 1.0 HW 5 0.5 A 0.0 0 10 20 30 40 60 TOTAL SUPPLY VOLTAGE (VOLTS) SMALL-SIGNAL OVERSHOOT vs CAPACITIVE LOAD 100 80 b 8 = 60 Lo @ Ww > a kK 5 40 Ww Oo faa uu a Vg = +15V 20 VIN = 100mV / Ave 4 0 | | 0 500 1000 1500 2000 2500 CAPACITIVE LOAD (pF) SHORT-CIRCUIT CURRENT vs TIME 60 q 50 = Ty = 425C _ Zz NN Vg = #15 = 40 leo (-) e sc oO E 2 2 30 9 Igo) c oc & 20 10 0 1 2 3 4 5 TIME FROM OUTPUT SHORTED TO GROUND (MINUTES) PEAK-TO-PEAK AMPLITUDE (VOLTS) f= 5omv isn ov une -50mV CMRR (dB) MAXIMUM OUTPUT SWING vs FREQUENCY 10k 100k FREQUENCY (Hz) Ta = 25C Vg = +15V 1M 10M SMALL-SIGNAL TRANSIENT RESPONSE am = +1, C, = 15pF Vg = 15V in= 25C 100 CMRR vs FREQUENCY 10k FREQUENCY (Hz) 100k COMMON-MODE RANGE (VOLTS) MAXIMUM OUTPUT (VOLTS) MAXIMUM OUTPUT VOLTAGE vs LOAD RESISTANCE POSITIVE SWING NEGATIVE SWING Ta = 25C Vg = 15V 100 1k 10k LOAD RESISTANCE (82) LARGE-SIGNAL TRANSIENT RESPONSE COMMON-MODE INPUT RANGE vs SUPPLY VOLTAGE Ta = 55C Ta = 125C Ta = 128C 6 45 +10 +15 +20 SUPPLY VOLTAGE (VOLTS)OP-27 TYPICAL PERFORMANCE CHARACTERISTICS VOLTAGE NOISE TEST CIRCUIT (0.1Hz-TO-10Hz) | O.1p.F 100kQ 402 NOTE: ALL CAPACITOR VALUES ARE FOR NON POLARIZED CAPACITORS ONLY. OPEN-LOOP VOLTAGE GAIN vs LOAD RESISTANCE 2.4 22 Ta = 26C Vs = 15V 2.0 s 1.8 a OPEN-LOOP VOLTAGE GAIN (V/xV)} p 1k LOAD RESISTANCE (2} 10k 100k APPLICATIONS INFORMATION OP-27 Series units may be inserted directly into 725, OP-06, OP-07 and OP-05 sockets with or without removal of external compensation or nulling components. Additionally, the OP- 27 may be fitted to unnulled 741-type sockets; however, if conventional 741 nulling circuitry is in use, it should be modi- fied or removed to ensure correct OP-27 operation. OP-27 offset voltage may be nulled to zero (or other desired setting) using a potentiometer (see Offset Nulling Circuit). The OP-27 provides stable operation with load capacitances of up to 2000pF and + 10V swings; larger capacitances should be decoupled with a 500 resistor inside the feedback loop. The OP-27 is unity-gain stable. Thermoelectric voltages generated by dissimilar metals at the input terminal contacts can degrade the drift perfor- mance. Best operation will be obtained when both input contacts are maintained at the same temperature. OFFSET VOLTAGE ADJUSTMENT The input offset voltage of the OP-27 is trimmed at wafer level. However, if further adjustment of Vos is necessary, a 10kQ trim potentiometer may be used. TCVogis not degraded LOW-FREQUENCY NOISE VOLTAGE NOISE (nV) Observation time limited to 10 seconds. PSRR vs FREQUENCY NEGATIVE 80 SUPPLY POSITIVE SUPPLY Q & 8 POWER SUPPLY REJECTION RATIO (dB) i 3 o 1 70 700 1k 10k 100k 1M FREQUENCY (Hz) 10M 100M (see Offset Nulling Circuit). Other potentiometer values from 1kQ to 1MQ can be used with a slight degradation (0.1 to 0.2pV/C) of TCVos. Trimming to a value other than zero creates a drift of approximately (Vogs/300) uV/ C. For exam- ple, the change in TCVgg will be 0.33uV/ C if Vog is adjusted to 100uV. The offset-voltage adjustment range with a 10k0 potentiometer is +4mV. If smaller adjustment range is re- quired, the nulling sensitivity can be reduced by using a smaller pot in conjuction with fixed resistors. For example, the network below will have a +280zV adjustment range. 1kQ. POT v Vt ATK? ATk2 B NOISE MEASUREMENTS To measure the 80nV peak-to-peak noise specification of the OP-27 in the 0.1Hz to 10Hz range, the following precautions must be observed: 7 (1) The device has to be warmed-up for at least five minutes. As shown in the warm-up drift curve, the offset voltageOP-27 typically changes 4,V due to increasing chip temperature after power-up. In the 10-second measurement interval, these temperature-induced effects can exceed tens-of- nanovolts. (2) For similar reasons, the device has to be well-shielded from air currents. Shielding minimizes thermocouple effects. (3) Sudden motionin the vicinity of the device can also feed- through to increase the observed noise. (4) The test time to measure 0.1Hz-to-10Hz noise should not exceed 10 seconds. As shown in the noise-tester frequency- response curve, the 0.1Hz corner is defined by only one zero. The test time of 10 seconds acts asan additional zero to eliminate noise contributions from the frequency band below 0.1Hz. (5) A noise-voltage-density test is recommended when measuring noise on a large number of units. A 10Hz noise-voltage-density measurement will correlate well with a 0.1Hz-to-10Hz peak-to-peak noise reading, since both results are determined by the white noise and the location of the 1/f corner frequency. UNITY-GAIN BUFFER APPLICATIONS When RS 1002 and the inputis driven with a fast, large signal pulse (>1V), the output waveform will look as shown in the pulsed operation diagram below. During the fast feedthrough-like portion of the output, the input protection diodes effectively short the output to the input and a current, limited only by the output short-circuit protection, will be drawn by the signal generator. With Ry 2 5000, the output is capable of handling the current requirements (I, = 20mA at 10V); the amplifier will stay in its active mode and a smooth transition will occur. When R; > 2kQ, a pole will be created with Ry and the amplifiers input capacitance (8pF) that creates additional phase shift and reduces phase margin. A small capacitor (20 to 50pF) in parallel with R will eliminate this problem. PULSED OPERATION Re WA > OP.27 __14 2.8V/yusec COMMENTS ON NOISE The OP-27 is a very low-noise monolithic op amp. The out- standing input voltage noise characteristics of the OP-27 are achieved mainly by operating the input stage ata high quies- cent current. The input bias and offset currents, which wouid normally increase, are held to reasonable values by the input- bias-current cancellation circuit. The OP-27A/E has Ig and log of only +40nA and 35nA respectively at 25C. This is particularly important when the input has a high source- resistance. In addition, many audio amplifier designers prefer to use direct coupling. The high Ip, Vos, TCVos of previous designs have made direct coupling difficult, if not impossible, to use. Voltage noise is inversely proportional to the square-root of bias current, but current noise is proportional to the square- root of bias current. The OP-27s noise advantage disappears when high source-resistors are used. Figures 1, 2, and 3 compare OP-27 observed total noise with the noise perfor- mance of other devices in different circuit applications. Total noise = [(Voltage noise)? + (current noise X Rs)? + (resistor noise)?]'1 Figure 1 shows noise-versus-source-resistance at 1000Hz. The same plot applies to wideband noise. To use this plot, just multiply the vertical scale by the square-root of the bandwidth. NOISE vs SOURCE RESISTANCE (INCLUDING RESISTOR NOISE) AT 1000Hz. 100 50 1 Rg UNMATCHED 2.9. RS=RS1=10k,RS2=0 2 Rg MATCHED e,g.Rg= 10k, Rg1=Rg275k Rs1 Rs2 Y 500 1k Rg SOURCE RESISTANCE (&) TOTAL NOISE (nVA/Rz) 1 50 100 Sk (10k 60k Figure 1 At Rg < 1k, the OP-27s low voltage noise is maintained. With Rg> 1k, total noise increases, but is dominated by the resistor noise rather than current or voltage noise. It is only beyond Rg of 20k that current noise starts to dominate. The argument can be made that current noise is not important for applications with low-to-moderate source resistances. The crossover between the OP-27 and OP-07 and OP-08 noise occurs in the 15-to-40k0 region. Figure 2 shows the 0.1Hz-to-10Hz peak-to-peak noise. Here the picture is less favorable; resistor noise is negligible, cur- rent noise becomes important because it is inversely propor- tional to the square-root of frequency. The crossover with the OP-07 occurs in the 3-to-5k0 range depending on whether balanced or unbalanced source resistors are used (at 3k the Ip, los error also can be three times the Vos spec.).OP-27 PEAK-TO-PEAK NOISE (0.1 to 10Hz) vs SOURCE RESISTANCE (INCLUDES RESISTOR NOISE). 1k 500 10Hz NOISE vs SOURCE RESISTANCE (INCLUDES RESISTOR NOISE). 100 50 Hz} 2 -08/108 8 z g y x 100 5 10 s 2 a a 9 ft e 50 1 Rg UNMATCHED Fo5 1 Rg UNMATCHED v e.g.8S=RS1=10k,Rs2-0 2 eg.Rg=Rg1=10k,Rg2-0 ft 2 Rs MATCHED 2 Rg MATCHED w eg. Rg=10k,RS1=Rg2=5k 27/37 eg.RS=10k,RS1=Rg2=5k RESISTOR ese ONLY 10 1 50 100 500 1k Sk 10k 60k 50 100 500 1k Sk 10k 50k Figure 2 Rg SOURCE RESISTANCE (9) Figure 3 Rg + SOURCE RESISTANCE (9) Therefore, for low-frequency applications, the OP-07 is bet- AUDIO APPLICATIONS ter than the OP-27/37 when Rs> 3k. The only exception is when gain error is important. Figure 3 illustrates the 10Hz noise. As expected, the results are between the previous two figures. For reference, typical source resistances of some signal sources are listed in Table 1. Table 1 SOURCE DEVICE IMPEDANCE COMMENTS Strain gauge <5000 Typically used in low-frequency applications. Magnetic <15000 Low lg very important to reduce tapehead self-magnetization problems when direct coupling is used. OP-27 Ip can be neglected. Magnetic <15000 Similar need for low Ig in direct phonograph coupled applications. OP-27 will not cartridges introduce any self-magnetization problem. Linear variable <15000 Used in rugged servo-feedback differential applications. Bandwidth of interest is transformer 400Hz to 5kHz. OPEN-LOOP GAIN FREQUENCY OP-07 OP-27 OP-37 AT: 3Hz 100dB 124dB 125dB 10Hz 100dB 120dB 125dB 30Hz 90dB 110dB 124dB For further information regarding noise calculations, see Minimization of Noise in Op-Amp Applications, Application Note AN-15. -10- The following applications information has been abstracted from a PMI article in the 12/20/80 issue of Electronic Design magazine and updated. Figure 4 is an example of a phono pre-amplifier circuit using the OP-27 for Ay; Ry-Re-C4-Co form a very accurate RIAA network with standard component values. The popular method to accomplish RIAA phono equalization is to employ frequency-dependent feedback around a high-quality gain block. Properly chosen, an RC network can provide the three necessary time constants of 3180, 318, and 75ys.! For initial equalization accuracy and stability, precision metal-film resistors and film capacitors of polystyrene or polypropylene are recommended since they have low voitage coefficients, dissipation factors, and dielectric absorption.* (High-K ceramic capacitors should be avoided here, though low-K ceramicssuch as NPO types, which have exceilent dissipation factors, and somewhat lower dielectric absorption can be considered for small values.) C4 (2) 220nF yI ow RS 100K&2 MOVING MAGNET LF ROLLOFF = CARTRIDGE INPUT O Lt G = 1kHz GAIN - R1 =0.101 {1+ ( * Ra) = 98.677 (39,9 dB) AS SHOWN Figure 4OP-27 The OP-27 brings a 3.2nV/\/ Hz voltage noise and 0.45 pA/.\/ Hz current noise to this circuit. To minimize noise from other sources, Rg is set to a value of 100, which generates a voltage noise of 1.38nW./ Hz . The noise in- creases the 3.2nW./ Hz of the amplifier by only 0.7dB. With a 1kQ source, the circuit noise measures 63dB below a 1mV reference level, unweighted, in a 20kHz noise bandwidth. Gain (G) of the circuit at 1kHz can be calculated by the expression: = Ry G=0.101 (1+ Ra ) For the values shown, the gain is just under 100 (or 40dB). Lower gains can be accommodated by increasing Rg, but gains higher than 40dB will show more equalization errors because of the 8MHz gain-bandwidth of the OP-27. This circuit is capable of very low distortion over its entire range, generally below 0.01% at levels up to 7V rms. At 3V output levels, it will produce less than 0.03% total harmonic distortion at frequencies up to 20kHz. Capacitor C3 and resistor R4 form a simple -6dB-per-octave rumble filter, with a corner at 22Hz. As an option, the switch- selected shunt capacitor C4, a nonpolarized electrolytic, bypasses the low-frequency rolloff. Placing the rumble fil- ters high-pass action after the preamp has the desirable result of discriminating against the RIAA-amplified low- frequency noise components and pickup-produced low- frequency disturbances. A preamplifier for NAB tape playback is similar to an RIAA phono preamp, though more gain is typically demanded, along with equalization requiring a heavy low-frequency boost. The circuit in Fig. 4 can be readily modified for tape use, as shown by Fig. 5. 0.47F TAPE R HEAD 3 18kQ T1 = 3180us T2 = 50ps Figure 5 While the tape-equalization requirement has a flat high- frequency gain above 3kHz (To = 50us), the amplifier need not be stabilized for unity gain. The decompensated OP-37 provides a greater bandwidth and slew rate. For many appli- cations, the idealized time constants shown may require trimming of Ry and Ro to optimize frequency response for nonideal tape-head performance and other factors. 5 -11- The network values of the configuration yield a 50dB gain at 1kHz, and the dc gain is greater than 70dB. Thus, the worst- case output offset is just over 500mvV. A single 0.47yF output capacitor can block this level without affecting the dynamic range. The tape head can be coupled directly to the amplifier input, since the worst-case bias current of 80nA with a 400mH, 100 nin, head (such as the PRB2H7K) will not be troublesome. One potential tape-head problem is presented by amplifier bias-current transients which can magnetize a head. The OP-27 and OP-37 are free of bias-current transients upon power up or power down. However, itis always advantageous to control the speed of power supply rise and fall, to elimi- nate transients. In addition, the dc resistance of the head should be carefully controlled, and preferably below 1kQ. For this configura- tion, the bias-current-induced offset voltage can be greater than the 100uV maximum offset if the head resistance is not sufficiently controlled. A simple, but effective, fixed-gain transformerless micro- phone preamp (Fig. 6) amplifies differential signals from low- impedance microphones by 50dB, and has an input impe- dance of 2k0. Because of the high working gain of the circuit, an OP-37 helps to preserve bandwidth, which will be 110kHz. As the OP-37 is a decompensated device (minimum stable gain of 5), a dummy resistor, Rp, may be necessary, if the microphone is to be unplugged. Otherwise the 100% feed- back from the open input may cause the amplifier to oscillate. Common-mode input-noise rejection will depend upon the match of the bridge-resistor ratios. Either close-tolerance (0.1%) types should be used, or Ra should be trimmed for best CMRR. All resistors shouid be metal-film types for best sta- bility and low noise. Noise performance of this circuitis limited more by the input resistors R, and Re than by the op amp, as R; and Re each generate a 4nW/ Hz noise, while the op amp generates a 3.2nWV/ Hz noise. The rms sum of these predominant noise sources will be about 6nW./ Hz , equivalent to 0.9zV in a 20kHz noise bandwidth, or nearly 61dB below a 1mV input signal. Measurements confirm this predicted performance. RI c1 R3 R6 1k2 | $n LOW IMPEDANCE MICROPHONE INPUT {Z = 50 TO 20022) P. R7 OUTPUT OP-27/37 RT a R3_ R4 Ri R2 Figure 6OP-27 For applications demanding appreciably lower noise, a high- quality microphone-transformer-coupled preamp (Fig. 7) incorporates the internally-compensated OP-27. T; is a JE-115K-E 1500/15k0 transformer which provides an opti- mum source resistance for the OP-27 device. The circuit has an overall gain of 40dB, the product of the transformers voltage setup and the op amps voltage gain. 1800pF R2 A 110082 OUTPUT _ SOURCE *71 = JENSEN JE-115K-E JENSEN TRANSFORMERS 10735 Burbank Blvd. N. Hollywood, CA 91601 Gain may be trimmed to other levels, if desired, by adjusting R2 or Ry. Because of the low offset voltage of the OP-27, the output offset of this circuit will be very low, 1.7mV or less, fora 40dB gain. The typical output blocking capacitor can be BURN-IN CIRCUIT eliminated in such cases, but is desirable for higher gains to eliminate switching transients. Capacitor C2 and resistor Ro form a 2us time constant in this circuit, as recommended for optimum transient response by the transformer manufacturer. With Coin use, A; must have unity-gain stability. For situations where the 2us time con- stant is not necessary, C, can be deleted, allowing the faster OP-37 to be employed. Some comment on noise is. appropriate to understand the capability of this circuit. A 1500 resistor and R,; and Ro gain resistors connected toa noiseless amplifier will generate 220 nV of noise in a 20kHz bandwidth, or 73dB below a imV reference level. Any practical amplifier can only approach this noise level; it can never exceed it. With the OP-27 and T, specified, the additional noise degradation will be close to 3.6dB (or 69.5 referenced to 1mV). References 1. Lipshitz, S.P.,, On RIAA Equalization Networks, JAES, Vol. 27, June 1979, p. 458-481. 2. Jung, W.G., /(C Op Amp Cookbook, 2nd Ed., H.W. Sams and Company, 1980. 3. Jung, W.G., Audio IC Op Amp Applications, 2nd Ed., H.W. Sams and Company, 1978. 4. Jung, W.G., and Marsh, R.M., Picking Capacitors, Audio, February & March, 1980. 5. Otala, M., Feedback-Generated Phase Nonlinearity in Audio Amplifiers, London AES Convention, March 1980, preprint 1976. Stout, D.F., and Kaufman, M., Handbook of Operational Amplifier Circuit Design, New York, McGraw Hill, 1976. OFFSET NULLING CIRCUIT -12-