How to measure OP-Amp Oscillation?

[Pippi Lindenstroem]

How to measure OP-Amp Oscillation?

I would like to be able to check if a new installed op-amp oscillates in a non-hearable frequency spectrum.

I already have some basic tools like an oscilloscope and signal generator

How can I go on?
Measure device?
Where to measure ?
How to measure (Do I need a signal? -Adjustment of the oscilloscope etc.)?


Your advice / a step by step Dummie guide would be very much appreciated

Many thanks
Pipp

[Enzo]

Look at its output pin with a scope. There is either a strong high frequency signal sitting there inaudibly or there is not. Now try it with a signal, looking in particular at peaks. See if there are any parasitic glitches on those peaks.

On your scope, audio usually looks like a wiggly line across the screen. if a stage oscillates, it usually does so at a high level as well as a high frequency. SO if your scope is set up to look at a volt of audio, a 50kHz signal 10 volts tall might be off the screen top and bottom, so make sure to turn the scope vertical down enough to include rail to rail voltage of the op amps. For example with +/-15v rails, then 5V/div on the scope will put the entire +15 to -15 range on the screen in 6 divisions vertical. And when set up for audio frequencies, that 50kHz might just look like a wide band of light on the screen. At least until you speed up the sweep to resolve it into a waveform.

But the waveform itself is not usually that important when we are simply looking for oscillation or no oscillation.

[Pippi Lindenstroem]

Originally Posted by Enzo

Look at its output pin with a scope. There is either a strong high frequency signal sitting there inaudibly or there is not. Now try it with a signal, looking in particular at peaks. See if there are any parasitic glitches on those peaks.

On your scope, audio usually looks like a wiggly line across the screen. if a stage oscillates, it usually does so at a high level as well as a high frequency. SO if your scope is set up to look at a volt of audio, a 50kHz signal 10 volts tall might be off the screen top and bottom, so make sure to turn the scope vertical down enough to include rail to rail voltage of the op amps. For example with +/-15v rails, then 5V/div on the scope will put the entire +15 to -15 range on the screen in 6 divisions vertical. And when set up for audio frequencies, that 50kHz might just look like a wide band of light on the screen. At least until you speed up the sweep to resolve it into a waveform.

But the waveform itself is not usually that important when we are simply looking for oscillation or no oscillation.

Thank you very much Sir, for your very helpful informations and the time you have spent

[ABO]

Measuring with a scope is difficult and might bring on oscillation caused by the proximity of the proble. Specially with fast opamps.

You might want to look for changes in supply current or check that the value is in accordance with the datasheet value. Usually oscillating opamps have elevated supply currents.

Oh yeah, HF oscillations can also change DC values. So if your 2 mV device suddenly exhibits 5 mV offset, you know what's going on.

[Dave]

You can also look at how it responds to a squarewave. Ringing indicates the circuit may have stability issues even though it isn't continuously oscillating.

[cliffforrest]

Originally Posted by ABO

Measuring with a scope is difficult and might bring on oscillation caused by the proximity of the proble. Specially with fast opamps.

You might want to look for changes in supply current or check that the value is in accordance with the datasheet value. Usually oscillating opamps have elevated supply currents.

Oh yeah, HF oscillations can also change DC values. So if your 2 mV device suddenly exhibits 5 mV offset, you know what's going on.

That is just SO WRONG and MISLEADING !!

Yes, I know, there are domains where 10pF can provoke instability, but not here! Learn to use your 'scope as your (electronic) eye! That is what the bloody things are for! So that you don't have feel how warm it is and guess!

Sorry, rant mode off!

[ABO]

Originally Posted by cliff

That is just SO WRONG and MISLEADING !!

Yes, I know, there are domains where 10pF can provoke instability, but not here! Learn to use your 'scope as your (electronic) eye! That is what the bloody things are for! So that you don't have feel how warm it is and guess!

Sorry, rant mode off!

Why not here? Did I miss some information?

Given the question, I am assuming the poster has little experience and it is an obvious mistake to make. I have seen plently of examples. And if you measure across a fast opamp (think 100 Mhz GBW) even close proximity (i.e. not touching) of the scope could invoke oscillation.
With your apparent experience you should know that.