Page 1 of 1

Discrete Op Amp clipping behavior

Posted: Fri Jul 15, 2022 9:35 am
by lsburden
I'm working on a simple discrete Class-A op amp. This isn't the first time I have worked on a discrete amplifier but there is a clipping behavior I would like to dig into this time.

The amplifier design is simple - run of the mill input differential pair, single transistor VAS (active or bootstrap load doesn't seem to matter), and single transistor emitter follower output with active load.

When pushing signal beyond initial voltage rail clipping the output signal starts to flip over to the opposite rail. I'm sorry if this is a poor description.
I'm sure there is a term for this behavior but I don't know it. If I knew the term I would search for it.

I don't recall this type of behavior with any IC op amp that I have used. Is this something that is just inherent with discrete amplifiers? Is there anything that can be done about this behavior?

Again, sorry if this is a poor description. I can provide more information if needed.

Re: Discrete Op Amp design

Posted: Fri Jul 15, 2022 11:13 am
by flyboy71
What you describe is phase reversal which occurs when the common mode voltage is exceeded. There are many newer enhanced versions of older opamps that the datasheet indicates as not having phase reversal. The TL072 datasheet is one example. You will definitely see this on a general purpose part such as an LM358.

Re: Discrete Op Amp clipping behavior

Posted: Fri Jul 15, 2022 11:18 am
by lsburden
I had somebody point me to the OPA4197 datasheet because there was a description of this behavior that sounded similar.
In the datasheet it is called Phase Reversal -

"Many op amps exhibit a phase reversal when the input is driven beyond its linear common-mode range. This condition is most often encountered in noninverting circuits when the input is driven beyond the specified common-mode voltage range, causing the output to reverse into the opposite rail. The OPAx197 is a rail-to-rail input op amp; therefore, the common-mode range can extend up to the rails. Input signals beyond the rails do not cause phase reversal; instead, the output limits into the appropriate rail."

This sounds right.
Now to try to see if it's possible to reduce or eliminate this behavior.
Anybody have any suggestions?

Re: Discrete Op Amp design

Posted: Fri Jul 15, 2022 11:19 am
by lsburden
flyboy71 wrote: Fri Jul 15, 2022 11:13 am What you describe is phase reversal which occurs when the common mode voltage is exceeded. There are many newer enhanced versions of older opamps that the datasheet indicates as not having phase reversal. The TL072 datasheet is one example. You will definitely see this on a general purpose part such as an LM358.
Thanks!
So this is a common issue then.
Then maybe not worth spending too much time on.

Re: Discrete Op Amp clipping behavior

Posted: Fri Jul 15, 2022 11:21 am
by lsburden

Re: Discrete Op Amp clipping behavior

Posted: Fri Jul 15, 2022 7:19 pm
by Audio1Man
Phase reversal is very common in most FET IC when over driven. Try a non FET IC
Duke

Re: Discrete Op Amp clipping behavior

Posted: Sat Jul 16, 2022 5:44 am
by mediatechnology
I do wish the manufacturers would stop calling this effect "phase" reversal when in fact it is polarity reversal.

TI's new TL072 "H" version has apparently addressed this problem.