I was looking through Doug Self's book and Sam Groner's treatise on opamps and noticed that in nearly every THD vs. freq. plot I've seen (at relatively high amplitudes),  common-mode distortion rises with frequency as expected, then appears to peak and begins to roll off at extremely high frequencies.  Why does that happen?


JohnRoberts

Re: Quick question about common-mode distortion in op-amps.
« Reply #1 on: May 11, 2020, 03:22:21 PM »
probably a combination of capacitive coupling (causing the increase) and falling open loop gain at HF tapering off.

Using op amps inverting can mitigate CM issues.

One complaint I had about uber-low distortion opamps is not only that they measure them inverting, they measure them with the NF padded to scale the distortion.

JR
Don't only half-ass tune your drums. Visit https://circularscience.com to hear what properly "cleared" drums sound like.

Re: Quick question about common-mode distortion in op-amps.
« Reply #2 on: May 11, 2020, 03:36:30 PM »
Cool, John, thanks!  I suspected it had something to do with gain/bandwidth.  For fun I was running a couple of devices through some THD tests, swapping out op-amps (BJTs) and testing again.

Regarding manipulation of test results, I also wonder how much smoothing manufacturers apply to their results to soften ugly peaks/dips.  I recently saw on a certain mic's datasheet that the frequency response plot was much "friendlier" looking than it was the last time I saw it, oh, say 10 years ago.

JohnRoberts

Re: Quick question about common-mode distortion in op-amps.
« Reply #3 on: May 12, 2020, 10:45:45 AM »
Cool, John, thanks!  I suspected it had something to do with gain/bandwidth.  For fun I was running a couple of devices through some THD tests, swapping out op-amps (BJTs) and testing again.

Regarding manipulation of test results, I also wonder how much smoothing manufacturers apply to their results to soften ugly peaks/dips.  I recently saw on a certain mic's datasheet that the frequency response plot was much "friendlier" looking than it was the last time I saw it, oh, say 10 years ago.
Microphones and loudspeakers have far more response irregularities than active electronics.  FFT signal processing has windowing functions that can smooth response. Marketers can be tempted to make transducer curves look better than they are.

JR   

Don't only half-ass tune your drums. Visit https://circularscience.com to hear what properly "cleared" drums sound like.

Re: Quick question about common-mode distortion in op-amps.
« Reply #4 on: May 13, 2020, 06:14:30 PM »
To add to what John said, distortion measurements are often made in a 30kHz or 80kHz bandwidth (or even less when using a sound card), so the analyzer itself will roll off the graph towards the top end.

JohnRoberts

Re: Quick question about common-mode distortion in op-amps.
« Reply #5 on: May 14, 2020, 09:57:37 AM »
To add to what John said, distortion measurements are often made in a 30kHz or 80kHz bandwidth (or even less when using a sound card), so the analyzer itself will roll off the graph towards the top end.
Risking TMI but that is how I roll... Back in the 70s trying to improve RIAA phono preamp designs I observed the difficulty using simple harmonic distortion as a performance metric. Harmonic distortion occurs and higher frequency incremental multiples of a fundamental. For 20kHz distortion products fall at 40kHz, 80kHz etc.  NF amplifiers exhibit higher distortion with HF due to falling open loop gain. Then throw in the RIAA playback EQ that further attenuates HF and we get measurements that look better than they should. 

My solution was to roll my own two tone IMD test using 19kHz and 20 kHz 1:1. The distortion product from any nonlinearity expresses at the 1kHz difference well down in the audio band where humans can hear it. Further with RIAA equalization 1kHz is reproduced with 20dB more gain than 19kHz or 20 kHz.  I found this test very revealing for improving phono preamps, too bad that is obsolete technology now.  ::)

JR
Don't only half-ass tune your drums. Visit https://circularscience.com to hear what properly "cleared" drums sound like.


JohnRoberts

Re: Quick question about common-mode distortion in op-amps.
« Reply #7 on: May 20, 2020, 11:24:19 AM »
BTW I wrote about common-mode distortion recently:
https://www.electronicdesign.com/technologies/analog/article/21131760/input-protection-for-lowdistortion-opamp-circuits
Nice I used to read that magazine back in the day...

IIRC Bob Pease (RIP) may have done some work in that area of low leakage protection, back in the day.

JR
Don't only half-ass tune your drums. Visit https://circularscience.com to hear what properly "cleared" drums sound like.


cuelist

Re: Quick question about common-mode distortion in op-amps.
« Reply #9 on: May 24, 2020, 08:26:41 AM »
Mr K


 

Related Topics

  Subject / Started by Replies Last post
7 Replies
5154 Views
Last post October 23, 2005, 12:48:22 PM
by bcarso
4 Replies
1700 Views
Last post January 06, 2013, 02:32:15 PM
by dfuruta
26 Replies
6681 Views
Last post February 10, 2013, 06:23:45 PM
by tv
0 Replies
890 Views
Last post April 13, 2013, 03:32:41 PM
by JessJackson