Conrad Johnson ACT2 against Audio Research Ref 3

Bingo Doc. But it could be even worse than that. . . the button likely shuts off one half of the Ref 3's circuit--that is the signal return path. The 6DB signal drop off is but a sideeffect.
I should like to point out that when the event happened, raising the volume on the remote did not correct the degradation of the signal. Furthermore, I was not alone at the time, several other people, including Aris Audio's Scott haver observed the degradation.
On the other hand, it is quite possible that a sound difference might be detectable only on some totally balanced systems.

Oneobgyn, Guido -
As you are both owners, see if you can find out (from ARC?) or elsewhere, what the purpose of having that button/circuit change is for. Given that there is an audible degredation of sound, it appears that the REF3 works best in balanced mode rather than in SE mode, or am I mistaken.

I disagree. IMO you had a REF 3 using balanced cables and not SE cables. Switching to SE with the remote resulted in faulty signal. I am surprised that you even heard any sound at all

"I am surprised that you even heard any sound at all."

I'm not. It comes down to how ARC has implmented the SE input signal. As with the LS5 that started it all with ARC, the design here is a truly balanced implementation. Only ARC has since added SE input and output connections. The question is how do they support SE signals:

1) Is the SE input converted to a balanced signal through a phase-splitter circuit which then runs the signal "balanced" from then on; or 2) Does the SE signal simply run through the "+" phase of the balanced design from input to output? My guess is the former as ARC is a big proponent of the balanced topology. In the former case, the SE/Bal switch would do nothing more than "activate" that phase-splitter stage. In the latter case, the SE/Bal switch would simply ground the input of the "-" phase just like an adaptor would do if using an SE source on a balanced-only connection like the LS5. In both cases, there is a signal on the "+" phase as the ARC products are non-inverting.

And then the output: 1) is the SE output simply the "+" phase; or 2) is there an additional stage that brings the two phases together for a SE output? The benefit here is using both phases throughout but at a cost of adding that final stage to produce an SE output.

In any of these cases, with balanced cables, the sound will come through just fine with the "+" phase and the "-" phase possibly having no signal. But this is no different than when I used an adaptor on the LS5 to run a SE power amp; the "-" phase was simply not being used.

Knowing exactly the implementation here would give the user a better idea as to the benefits, or more likely, the losses, of using the SE inputs or outputs vs. the balanced connections for a more "pure" sound.

John

Excellent points John. I will call Leonard at ARC when I have a minute. I will post his answers here.