The Obsolescence Brief Sign in Subscribe
← All briefs Parts Teardown

OP27 replacement: the OP37, OPA227, AD797 and AD8671 options compared

The OP27 precision op-amp is NRND at ADI. The deciding difference is compensation, worked out part by part.

OP27: Not Recommended (NRND) OP37: decompensated (G≥5) OPA227: pin-for-pin, unity-gain AD8671: ADI-named alternative

The OP27 low-noise precision op-amp is Not Recommended for New Designs at Analog Devices. The cross-reference identified the OP37, the TI OPA227/228, the TI OPA27, the AD797 and the AD8671. None is a drop-in; the deciding difference is compensation, whether the part holds a unity-gain loop, with noise grade and DC offset close behind. The sections below work through those differences, part by part.

01

Status // The part in question

Lifecycle status

ADI lists the OP27 as Not Recommended for New Designs and names the AD8671, AD8675 and ADA4004-1 (all in production) as suggested replacements. The line still carries 12 part models, and each orderable carries its own lifecycle flag: the plastic OP27GPZ (PDIP) is still active at distribution. Confirm the exact orderable you use against its own status.

NRND is not obsolete, so this is a planned migration, not an emergency. Identify the specific OP27 grade you carry first; it sets the DC error budget and the requalification target for whichever cross you pick.

Last-time-buy timing and stock are dated data for the email, not this page.

02

Candidates // Why these

The candidate set

The parts engineers cross-shop for the OP27, each with the fit-catch to validate in your own circuit:

OP37ADI · the decompensated siblingNRND
Noise 3.0-3.2 nV/√Hz
Min gain 5 (decompensated)
Speed 63 MHz · 17 V/µs

Catch: stability. Same specs, pinout and noise as the OP27, and faster, but decompensated: stable only at a closed-loop gain of 5 or more, so in a unity-gain socket it can oscillate. Use it only where your gain is already 5 or higher.

OPA227/228Texas Instruments · pin-for-pinACTIVE
Noise 3 nV/√Hz · 75 µV
OPA227 unity-gain (G≥1)
OPA228 decompensated (G≥5)

Catch: DC spec and the sibling trap. The OPA227 is the closest functional match: unity-gain stable, 3 nV/√Hz, marketed pin-for-pin. Check offset against your grade. Do not grab the OPA228 by mistake, that is the decompensated (G≥5) OP37 analog.

OPA27Texas Instruments · direct 2nd sourceACTIVE
Pinout same 8-pin single
Stability unity-gain stable
Grade offset & noise graded

Catch: grade matching. TI's datasheet calls the OPA27 an improved replacement for the OP-27, same pinout, unity-gain stable, in production. Match the OPA27 grade to the OP27 grade you carry.

AD797ADI · ultralow-noise step upACTIVE
Noise 0.9 nV/√Hz (1.2 max)
Speed 110 MHz · 20 V/µs
Comp external decomp cap

Catch: a noise grade above the OP27. At 0.9 nV/√Hz it is a step up, not a match, and it needs an external decompensation cap rather than being flatly unity-gain stable. Reach for it when you need lower noise, and design in its compensation network.

AD8671ADI · named modern alternativeACTIVE
Noise 2.8 nV/√Hz
Offset 75 µV · 0.5 µV/°C
Stability unity-gain stable

Catch: different DC numbers. A modern unity-gain-stable part ADI names as an OP27 alternative, but its noise, offset and supply range differ. Validate against your stage's error budget rather than assuming the cross-reference closes it.

Also cross-shopped: OPA37 (TI, the OP37 analog), HA-5127 (Renesas), OP270 (ADI dual).

03

Pinout // Pin 8 is the tell

Same 8-pin outline, pin 8 is not the same

All six share the 8-pin single-op-amp outline, so a cross-reference on package alone reads them as interchangeable. The non-obvious result: the OP37, OPA227 and OPA27 are pin-identical to the OP27, yet the deciding difference,compensation, is not a pin at all. Where the pinout does move is the trim/null pins: the AD797 turns pin 8 into a decompensation node, and the AD8671 drops the trim pins entirely.

8-pin single op-amp pinout · pins 2, 3, 4, 6, 7 shared · the trim pins and pin 8 differ

OP27ADI · subject
NRND
1Trim
2−IN
3+IN
4V−
Trim8
V+7
OUT6
NC5
OP37Analog Devices
NRND
1Trim
2−IN
3+IN
4V−
Trim8
V+7
OUT6
NC5
OPA227Texas Instruments
ACTIVE
1Trim
2−IN
3+IN
4V−
Trim8
V+7
OUT6
NC5
OPA27Texas Instruments
ACTIVE
1Trim
2−IN
3+IN
4V−
Trim8
V+7
OUT6
NC5
AD797Analog Devices
ACTIVE
1Null
2−IN
3+IN
4V−
DECOMP8
V+7
OUT6
Null5
AD8671Analog Devices
ACTIVE
1NC
2−IN
3+IN
4V−
NC8
V+7
OUT6
NC5

Pin color,green same function · amber trim pin rearranged or removed · red function changed

COMPENSATION IS THE TELLThe OP37, OPA227 and OPA27 are pin-identical to the OP27; the deciding difference is compensation, which is not a pin.
AD797 CHANGES PIN 8Pin 8 becomes a decompensation node, not an offset trim, and at 0.9 nV/√Hz it is a noise step up, not a match.
AD8671 DROPS THE TRIMPins 1 and 8 are no-connects, no external offset trim, so any trace to the old trim pins must be checked.

Pinout is necessary, not sufficient. Suggested candidates to validate in your own circuit, not drop-in equivalents, the load-bearing difference here is compensation, off the pin map.

04

Parametrics // Side by side

Comparison table

OP27 (NRND)OP37OPA227AD797AD8671
LifecycleNRNDNRNDActiveActiveActive
Min stable gainG≥1G≥5G≥1G=1 ext. capG≥1
Noise @1kHz3.0-3.2 nV3.0-3.2 nV3 nV0.9 nV2.8 nV
Offset (max)25-100 µV25-100 µV75 µV80 µV75 µV
Offset drift0.6 µV/°C0.6 µV/°C0.6-2 µV/°C1.0 µV/°C0.5 µV/°C
GBW8 MHz63 MHz8 MHz110 MHz10 MHz
Slew rate2.8 V/µs17 V/µs2.3 V/µs20 V/µs4 V/µs
Supply±4 to ±18V±4 to ±18V±2.5 to ±18V±5 to ±15V±5 to ±15V
Package8-pin8-pin8-pin8-pin8-pin

Compensation is the row that decides a swap: the OP27, OPA227 and AD8671 are unity-gain stable, while the OP37 and OPA228 need a closed-loop gain of at least 5, and the AD797 reaches unity gain only through its own external decompensation network. All share the 8-pin single-op-amp pinout. Specs are vendor-reported, not independently audited.

05

FAQ // Frequently asked

Frequently asked questions

What is the replacement for the OP27?

There is no single drop-in. For a unity-gain or low-gain stage, the unity-gain-stable OPA227 (TI, marketed pin-for-pin) or the AD8671 (an ADI-named alternative) are the candidates to validate; for a stage already at a closed-loop gain of 5 or more, the OP37 or OPA228 add bandwidth. TI's OPA27 is a same-pinout second source. Each needs validation against your offset, noise and stability, not a straight swap.

Is the OP37 a drop-in for the OP27?

No. The two share the same specifications, pinout and 3 nV/√Hz noise, but the OP37 is decompensated and stable only at a closed-loop gain of 5 or more, so in a unity-gain OP27 socket it can oscillate. It is a valid swap only where your gain is already 5 or higher.

What is the difference between the OP27 and OP37?

The difference is compensation. Both share the same specifications, pinout, 3 nV/√Hz noise and offset, but the OP27 is unity-gain stable while the OP37 is decompensated with a minimum stable gain of 5. That decompensation gives the OP37 higher speed: 63 MHz GBW and 17 V/µs slew, versus 8 MHz and 2.8 V/µs.

What is the difference between the OP27 and OPA27?

The OPA27 is Texas Instruments' second source for the OP27, on the same 8-pin pinout and unity-gain stable; TI's datasheet calls it an improved replacement for the OP-27, and it is in production. The practical difference is grade, the OPA27's guaranteed offset and noise are specified by grade, so match the OPA27 grade to the OP27 grade you are replacing. Unlike the OP37, it does not change compensation.

Can the OPA227 replace the OP27?

It is a candidate to validate, not a guaranteed drop-in. TI markets the OPA227 as a pin-for-pin replacement for the OP27; it is unity-gain stable, in production, and specified at 3 nV/√Hz. Confirm its input offset and bias current against the specific OP27 grade you are replacing before committing.

Is the OP27 obsolete or still in production?

Not obsolete, but Not Recommended for New Designs (NRND) at Analog Devices. Orderable models such as the plastic OP27GPZ still show as active at distribution, and ADI names the AD8671, AD8675 and ADA4004-1 as suggested replacements. Treat the OP27 as a planned migration and confirm your specific orderable's status.

06

Sources // Public, primary

Sources

Editorially independent and from public sources only; lifecycle, pin and rating data is vendor-reported, not independently audited. Alternates are candidates to validate in your own design. Confirm every detail with the manufacturer before committing.

Hermann Njike

Hermann Njike is a systems obsolescence manager who leads the cross-functional response when critical components go end-of-life, and is the single point of contact for thousands of production parts.

Related briefs
The Brief, weekly

One obsolete part, worked out in full.