Curious problem measuring distortion with QA403 and Alpine PDX F4 amp

Let me start this by stating I am a EE but I graduated in 1989 and have not done much with analog electronics since the mid 1990’s. I am having problems with high noise floor and distortion on a couple of PDX F4 amps but only on channels 2 and 4. The noise signal is only present when the QA403 output is connected to the amplifer input for these channels and it is being amplified along with the 1 kHz test tone. All results so far have been testing single ended from QA403 output to PDX input and single ended from the PDX output to the QA403 input. It has occured to me if I test with differential inputs and outputs this problem may go away. I don’t have any cables configured for this but can make some.

If I remove the QA403 from the test and only use a signal generator and a battery powered scope, all 4 channels reproduce the test signal with very small amount of switching frequency ripple riding on the sine wave. I am still trying to figure out why the noise is only manifesting on 2 channels. I have attached multiple picture to show what I am seeing. The first shows how the bridged outputs are configured. I feel that it can’t be a coincidence that the channels with the problem are the ones where the positive output is at the same potential as the 12 VDC input ground but I still don’t understand the mechanism.

The noise signal is there even if the QA403 generator is off. I only need to connect the QA403 output to the amp input for the noise to be there.



Here is a good channel measurement and I have attached a screen shot of the noise signal when QA403 generator is off. If I disconnect the QA403, this noise signal goes away or is so small in amplitude, I don’t see it on a 1 kHz test signal. It is 155 kHz and 2.5 volts rms at the speaker output.