I have a multimeter in series with the amplifier I am testing setup to measure current which seems to be the culprit for a large ripple.

Input: 2Vpp
Differential gain: 2
Output: 8Vpp
Load: 8Ω
Pout: 1W

When I use mini grabbers from the regulated PSU to the DUT with long cables the ripple is ~ 100mVp. When I solder short, heavy gauge wires in parallel for for each direct PSU connection (power, gound) the ripple is ~10mVp. With the multimeter in series to measure current even with the heavy gauge wire (mini grabber from multimeter to heavy gauge wire) the ripple is 300mVp.

A large ripple will wreck havok with THD+N performance. Does anyone know what the innards of a multimeter are like where it would cause a ripple on the PSU to the DUT? The multimeter is an Agilent 34401A.

I would expect long wires to have inductance to be a factor, but the multimeter makes it worse.

 

Probably the cause there is some resistance of the multimeter (as ammeter). A resistance in series with the load will cause the ripple to be more severe, since the capacitor will discharge slower and provide less backup to sustain the voltage. Getting a more sensitive galvanometer and using it as ammeter would not be a bad idea. Or you can solve this issue with a second filter capacitor after the ammeter, making a resistive Pi filter (the ammeter is the resistor of the Pi filter).

It your amplifier an audio amplifier (seems to be the case)? If it is, you shouldn't worry too much about ripple voltage. They normally work well under a ripple of up to 10%.

 

Most multi meters measure current by measuring the voltage across a small well-calibrated internal resistance. Even though this resistance is less than a tenth of an Ohm, it still would become significant in series with a small load such as the one described here.

 

This is a case where an oscilloscope would give a much better picture of the actual condition.