You might just ground the input to the sound card to see if it's supply or sound card.

 

Good idea. It's funny but when I use the circuit in post #20 instead of the one in post #15, NO SPIKES. What could that mean?

 

Doesn't make any sense, because it looks like your sound card is already AC coupled. Unless the DC voltage is to high for the sound card.

 

You might just ground the input to the sound card to see if it's supply or sound card.

Recorded through grounded sound card for 12+ hours. No spikes.

 

Doesn't make any sense, because it looks like your sound card is already AC coupled. Unless the DC voltage is to high for the sound card.

If I understand the difference between circuit post#15 (which spikes) and circuit post#20 (which does not spike), the former references the capacitor to the negative rail and the latter does not. Wouldn't this make the negative rail of the supply suspect?

 

Wouldn't this make the negative rail of the supply suspect?

If the sound card ground and negative rail are commoned, any spikes on the negative rail shouldn't get recorded.

 

If the sound card ground and negative rail are commoned, any spikes on the negative rail shouldn't get recorded.

In that case, I am at a loss.

 

Not altogether relevant to this discussion, but extremely interesting and something I did not know: http://powerelectronics.com/mag/709PET21.pdf