Power foldback problem on graphing

During automated testing in V1.215 using THD vs power, as power increases, it suddenly folds back to a low power. It is not the DUT because I can push it hard into clipping if I increase the amplitude manually.



Image shows amp clipping around 30 volts pk into an 8 ohm load.

Thanks for any help.

Hi @Johnr6697, could you go into Edit->Settings and set the pause time between acquisition to 1 second? And then run your failing test again? Maybe a relay is switching and for whatever reason the switch wasn’t completed when the measurement was started???

I’m also having this issue and the 1 second delay doesn’t solve it. It seems to be that around -12dBV through -2dBV output much lower than they should.

Here’s the tests I did:
-13dBV output had around 1.6v input
Bumping up to -12dBV will decrease the input to 0.6v

-3dBV output brings us back up to around 1.6v input
Bumping up to -2dBV jumps to nearly 6v

I have screenshots for all those^ but as a new member on the forum its only letting me upload one image.

Lastly I shorted the output to the input and we can see the same thing happening:

Let me know if there’s anything else you’d like me to try to debug this further :slight_smile:

Hi @abcdmku, I accelerated the normal “new user” rating and I think you should be able to paste away.

I reverted to 1.215 software, and am trying to replicate the fold back. You noted you can do it in loopback. Can you please provide more instructions? I don’t quite understand in your plot how you see it replicating?

Thanks!

Apologies for the long delay. Anyhow, setting the delay to 1 second did not help. I was able to complete measurements by manually setting the attenuator to 42dB and it would finish correctly, but that is not ideal for testing low distortion amplifiers. Did not have this issue in 1.210 that I believe I was running previously.

Hi @Johnr6697, not a problem at all. Good to know the 1 second delay didn’t help. Is there a possibility you could try your settings (with Autoset enabled) with pre-release version 1.216? This is a pre-release, but it has moved to common code paths for all the tests and centralized logic. And then, you can go back to the previous release. (1.215) after the quick eval. Thanks!

Hi @matt, thanks for lifting the new member rating.

Its hard to see in the image I previously sent, I zoomed in to the area where you can see the output power cuts back and hops forward.

Here’s the photos from the original test:

-13dBV output had around 1.6v input

Bumping up to -12dBV will decrease the input to 0.6v

-3dBV output brings us back up to around 1.6v input

Bumping up to -2dBV jumps to nearly 6v

Ill give the new version a try and write back

1.216 has the same issue

Hi @abcdmku, the levels at which you are finding issues are DAC relay boundaries. So, if you just go into single-ended loopback and add the gain measurement tile, and set the generator to -13 dBV you should see

And then again at -12 you should see:

From the plots you provided, it looks like a DAC output relay might be stuck or not working?

Same thing at the -3 dBV to -2 boundary.

At both of these transitions (-13 to to -12 output level, and -3 to -2 output level) you should hear a relay click.

Can you try this:

  1. Put QA403 into single ended loopback (L+ OUT to L+ IN, short L- IN, and same for right channel)
  2. File->New Settings
  3. Run Amp Gain and Distortion versus Amplitude with the following settings:

This will exercise every relay in the DAC and ADC path, and should yield a plot as follows:

It’s hard to see in the plot, but the green right trace is right behind the left trace.

I’ve tried quite a few configurations using the plug-ins, in both loopback and with a 20 dB amp and can’t replicate your problems. But the issue you posted above with the sudden jumps while not running the tests seems to indicate a HW problem to me.

1 Like

That configuration is saved as a useful quick test I can run to test my input/output gain/attenuators and correct relay operation. :slight_smile:

Non-inverting:

Inverting: