More Grounding Questions

Will ground loops degrade sound quality even though it does not hum?

In addition to producing hum, ground loops can also result in high frequency noise and buzz, by coupling harmonic distortion components and/or noise that is typically present on the ac lines into signal return paths. However, if you don't hear that when no music is playing, or if any low level buzz or noise that you do hear is the same whether or not the equipment is temporarily floated with a cheater plug, I doubt that it would have any effect on sound quality. I suppose it is slightly conceivable that inaudible ultrasonic noise that is introduced by a ground loop could somehow intermodulate with the music and have audible effects, but it's easy enough to rule out that possibility -- just compare the sonics with and without the ac safety ground temporarily lifted via a cheater plug.

When you "float" your power amp, do you float the signal/circuit ground from the chassis or you float AC ground from the chassis?

I would never isolate circuit ground from chassis, assuming the equipment was designed with them common. And I'm not sure how it would even be practical to do that, unless the design had them connected together at only one point, and the rca jacks were the type that has isolated ground sleeves.

Assuming that circuit ground and chassis are connected at more than one point, isolating those connections would most likely result in significant impedances, especially at high frequencies, between circuit points that should be common. In other words, grounds at different points in the circuit would no longer be truly common. Also, stray capacitances between various circuit points and ground would be different than those which were present when the initial design and development was performed, in effect changing the design in unpredictable ways. Also, instead of the chassis functioning as a shield, it would very conceivably function as a path for emi/rfi to couple into internal circuit points.

Of course, as you probably realize floating the ac safety ground is in principle not a good idea either, due to the safety risk it creates. Although a lot of people do it anyway. As far as I am aware the only way to break a problematic ground loop that is both effective and completely safe is with something like a Jensen Isomax.

This Jensen paper will probably be of interest, if you haven't seen it already.

Best regards,
-- Al

But I am not sure it (ground loop) does not hurt sound quality when the music is playing.

As I suggested, I think that the only way to be sure is to try it both ways (with and without cheater plugs), and compare the sound. I don't think comparing experiences with others will be helpful on that question, because it figures to be highly dependent on the components, the cables, and the AC power distribution.

The Jensen paper may be technically correct. But it may not describe all real life scenarios. Take Figure 1 in that paper as an example. It tried to explain the hum voltage potential at the power amp and suggested that it would cause hum. This may well be correct. But both my power amps had the same grounding scheme before the mod. However, only one of them hummed. So his theory cannot be true at all time.

The Jensen paper explains that the root cause of ground loop hum and buzz is stray capacitance (and in some cases also designed-in capacitance) between the ac line and chassis. Particularly stray capacitance within the power transformer. If your two amps did not have identical transformers, and identical electrical and mechanical design of the the surrounding circuitry, wiring, and mechanical structure, there is no reason to expect their susceptibility to ground loop problems to be the same, even if their basic grounding schemes were similar.

So for chassis to act like a shield, do you have to wire the chassis to the signal ground or AC ground?

If the chassis is connected to AC safety ground but not to circuit ground, it seems to me that any and all noise voltages that may be introduced onto the chassis from the AC safety ground wiring, or as a result of emi/rfi, will be free to couple via stray capacitances from the chassis to arbitrary and unpredictable points within the circuits. If the chassis and circuit ground are common, there will be controlled low impedance pathways for that noise to be shunted to ground.

I thought the Aiken paper was excellent; thanks! I note that he addressed the issue of isolated rca jacks that I had raised.

You may find the following paper and its references to also be of interest:

http://www.analog.com/static/imported-files/tutorials/MT-095.pdf

Best regards,
-- Al