Balanced interconnects and non-differential amps

FWIW there is a simple method of getting the amp to process both phases of the incoming signal without changing the character of the amp. If the XLR connection is already in place this is very easy to do.

All that need happen is a coupling cap from the unused pin of the XLR (likely pin 3) that goes to the cathode of the input tube.

Its not balanced but it is differential and the CMRR number should actually be pretty good.

Ya gotta love Mr. Al, always a cogent, well-reasoned reply. Contact Odyssey for the definitive answer...

-RW-

Thanks again, Al.

Hi Howard,

I couldn't find enough information to determine if the signals sent into the XLR input of the Kismet are received by a balanced receiver stage or not. As I indicated, more often than not that will be the case even if the rest of the internal signal path is unbalanced, at least in quality designs. But I'm not at all certain that is the case here.

The only specs I found in this case that may have some relevance to that question are contained in the statement in the review that "the overall gain is about 30dB with an input impedance of 22kΩ and input sensitivity of 1V RMS." If a balanced receiver stage were provided for the amp's XLR input, I would expect the input impedance of the XLR input to usually (although not always) be different than the input impedance of the RCA input, with the impedance often (although not always) being a factor of 2 greater for the XLR input than for the RCA input. And likewise I would expect that the gain would often (although not always) be 6 db greater for the XLR input than for the RCA input.

So given that only one set of numbers for those parameters seems to be available, with no indication as to whether it applies to the RCA input or the XLR input or both, in this case I think it would be best to present the question directly to Odyssey, and see what they have to say.

Best regards,
-- Al

Thanks so much, Al.

Here's what prompted my question: I've been looking into the Odyssey Audio amps, and read the following in a 6moons.com review of the Kismet monoblocks: "On the input side of the amp, the Kismets come fitted with XLR receptacles although these are provided as a convenience for those with balanced line-level cabling. The amp is not a differential design. Internally Groneberg’s Quattro Reference New Generation cabling connects the line-level inputs to the first stage of the amplifier."

As I'm nobody's electrical engineer -- my knowledge of such things is quite sketchy, in fact -- I wasn't sure what exactly to make of this statement. And it got me to thinking, and wondering about balanced connections more generally. I appreciate your answer -- and the link.

-- Howard

If one of the generally accepted benefits of long runs of balanced interconnects between a preamp and an amp is noise rejection, is this still true whether or not the amp in question is a differential unit?

Yes, provided that the driver and receiver circuits that interface with the cable are balanced (which is usually but not always the case if XLR connectors are provided), and of course as long as the cable is balanced. A balanced pair of signals do not necessarily have to be provided, but the impedances in the two interface circuits between each of the two signal lines and ground have to be equal (or close to it), for the noise rejection benefit of a balanced connection to occur.

Also, regarding balanced vs. unbalanced connections when the components do not have internally balanced or differential signal paths, see my post here.

Best regards,
-- Al