There are actually several different variations on the balanced-unbalanced topology.
In a balanced connection, there are two separate signals, a (+) signal and a (-) signal.
A true balanced output delivers these as two separate and active signals.
And a true balanced input subtracts the two signals to derive its output - which is how it achieves its immunity to common mode noise.
(And, if either signal is either missing or ignored at either end, then it isn't a true balanced connection.)
(And, if you simply lose one of the signals, the sum of the two will be lower by half.)
And, in a balanced component, you have two separate channels, for example two separate amplifier channels....
One channel delivers a (+) signal, the other delivers a (-) signal, and the "true output" is the difference of those two signals.
However, internally, the two channels in a balanced component may or may not be actively cross connected.
(If you have a balanced amplifier, and the halves are not cross connected, if you lose one side of your input signal, half of your amplifier becomes an active ground.)
(The side receiving a signal is active; while the half not receiving a signal will cheerfully amplify 0V and act as a ground.)
On Emotiva gear, we tend to actively convert and cross connect when it makes sense.
So, on an XSP-1, if you use an unbalanced input, it is fully converted to a balanced signal, and both the (+) and (-) balanced outputs will have the proper signal on them.
And, if you have a balanced input, both sides of the input signal contribute to both sides of the output signal.
Likewise, on one of our fully differential amps, if you were to connect an input to only one of the two balanced input lines, BOTH output modules will still be operating.
(Rather than having the (+) input drive one amplifier module, and the (-) input drive the other, both output modules have fully differential inputs, and both output modules are driven by both input signals.)
(However, loss of one of those input signals would still reduce the output of each by half.)
In certain instances, when going from balanced to unbalanced, we do not do it this way.
For example, when you use the balanced-to-unbalanced adapters on the RMC-1, the (+) balanced output is used as the unbalanced output, and the (-) balanced output is unused.
(In this case, doing it the other way, and summing the balanced signals to derive the unbalanced output, would not improve the distortion performance, and the extra active stage would actually result in a higher noise level.)
If you don't cross connect the signal along the way, then an unbalanced input signal, or an unbalanced connection anywhere along the way, will result in only half of the balanced circuitry being used.
To be fair I don't think too many folks do it that way these days... to do so would literally result in an amplifier whose rated power output with an unbalanced input was half what it was with a balanced input.
With most modern equipment, you probably should consider whether the component is balanced, and whether connections between components are balanced, to be separate questions.
(Remember that you can in fact have "a fully balanced amplifier" that only has an unbalanced input... and many amplifiers with balanced inputs are not "fully balanced amplifiers".)
Thanks, Keith -
I'm going to try the easy part first. I'll substitute "known good" interconnects for the ones in the chain & see if that gets me anything.
One other question, if I might
KeithL ? If one starts at the source (in this case, a DAC's analog output) with unbalanced connections, does a subsequent component (a preamp, for example) convert the source to balanced output if you're using the XLR outputs of the preamp? Or is a balanced chain a full "source to power amp" thing, where EVERY component must be balanced?
Thanks - Boom