channel separation

I remain puzzled about how the channels can bleed together at all in an amplifier (not a phono cartridge) if the right and left circuits are truly separated from input to output.

When two circuits are close to each other, just think of how a transformer or a capacitor works.

P.S: One more thing I noticed in the photo of the underside of the preamp is that at the extreme right there appears to be a bundle of many wires running from the small circuit board near the input jacks at the rear, all the way to the input selector switch at the front. Those wires presumably carry the input signals for both channels, for all of the inputs, and all of those wires are neatly (and tightly!) bundled together. More opportunity for crosstalk, and one that IMO is likely to be a particularly significant contributor to the 50 db number.

Regards,
-- Al

How can the channels bleed into one another when the entire layout is dual mono?

The answer, IMO, is that the design is not truly dual mono. If you look at the photos of the preamp at the PL site, especially the photo of the underside, the following things can be seen:

1) The input jacks for both channels are located close to one another, on one side of the preamp.

2) Those jacks appear likely to be wired to a single small printed circuit board, serving both channels. That isn’t entirely clear in the photo, however.

3) There is a single volume control, serving both channels.

4) There is a single input selector switch, serving both channels.

5) The output jacks for the two channels are located close to one another.

6) As a consequence of the foregoing factors several wires can be seen running from one side of the preamp to the other. And in one case, near the control at the top left of the photo of the underside, it appears that wires associated with both channels are physically bundled together. That is good with respect to the neatness and appearance of the underside, but bad with respect to crosstalk.

7) A printed circuit board in the power supply section, the board located toward the rear of the amp that among other things has four large black capacitors on it, serves both channels. Its relatively small size means that circuit points on it that are related to both channels are in close proximity. Also, ground and power "planes" that are probably within that board probably serve both channels.

Undoubtedly there are more such examples. And all of these things create opportunities for crosstalk between the channels, especially at high frequencies.

In any event, though, I agree with the others that 50 db of channel separation at 20 kHz (and >77 db below 1 kHz, per Stereophile’s measurements) is not in itself a significant issue.

Regards,
-- Al

Interesting. It sounds like channel separation may not be causally related to sound quality,

I wouldn’t say this, exactly, but that 50 dB is a very high amount of channel separation.

As a scientist - but not an engineer - I remain puzzled about how the channels can bleed together at all in an amplifier (not a phono cartridge) if the right and left circuits are truly separated from input to output.

They are galvanically isolated, that is, there is no DC path from one to the other, however electromatnetic (inductive) and electrostatic (capacitive) coupling are quite well understood.

https://en.wikipedia.org/wiki/Capacitive_coupling

Best,

Erik

Interesting.  It sounds like channel separation may not be causally related to sound quality, but can this provide some evaluative metric of the underlying engineering?  For example, the Pass XP-22 preamp measured >110 db separation below 4 kHz and 96 db at 25 kHz.  If this simply means that they paid extra attention in keeping the channels electrically separated, then maybe it indicates other characteristics that are also beneficial to the sound.  As a scientist - but not an engineer - I remain puzzled about how the channels can bleed together at all in an amplifier (not a phono cartridge) if the right and left circuits are truly separated from input to output.

Do you have any idea what 50 dB channel separation is like?
For comparison, a lot of ultra high end phono cartridges have only about 30 dB. 50 is huge.

Channel separation is but one of a whole long list of more or less irrelevant measurements. Actually "more or less" is being kind. 

BTW, I did the math.

Assuming one channel has 1V, -50 dB is about 0.003 volts.

It's just really a matter of how well isolated the audio as well as the power supplies are.

At 20 kHz it probably doesn't matter. 50 dB is pretty good!

It's in my mind the midrange where separation is more important.