Balanced cables

Atmasphere 2-28-2018
Its probably not the amp that is causing you to hear the cables so much as the preamp. But there is a recent period of ARC amps that had me scratching my head. I’m not sure which models they are. They were out when Kalvin Dahl (with whom I went to school) was still at ARC (about 2-3 years ago). Apparently the amp has a very low CMRR (Common Mode Rejection Ratio) so it only has balanced inputs. Apparently also if you try to run it single-ended the power goes down and the distortion goes up.

I can’t think of a good reason for a low CMRR in a differential amplifier (which is what these amps use). You wind up leaving performance on the table (I’ve been designing differential circuits since the mid 1980s).

Ralph, I’ve never been able to find a schematic for any of ARC’s amps which only provide balanced inputs, such as the recent Reference series amps, but I’ve had the suspicion that instead of using differential stages they basically have a separate signal path through the amp (up to the output transformer primary) for each of the two signals in the balanced signal pair they receive for each channel. That would be consistent with a very dramatic reduction of power capability as well as an increase in distortion if the amp were to be provided with unbalanced inputs via RCA-to-XLR adapters or adapter cables, as was found to be the case with the Ref 150 used by the OP in this thread we had participated in some time ago. I believe it would also be consistent with low CMRR, due to the gain and other characteristics of the two paths not matching precisely.

I have no idea why they might have chosen that kind of approach, and I can’t say for sure that they did, but if I am correct in suspecting that they did so it would seem to explain some or all of the things you mentioned about their amps that you referred to above.

Best regards,
-- Al

It depends on who you believe

but I’ve had the suspicion that instead of using differential stages they basically have a separate signal path through the amp (up to the output transformer primary) for each of the two signals in the balanced signal pair they receive for each channel. That would be consistent with a very dramatic reduction of power capability as well as an increase in distortion if the amp were to be provided with unbalanced inputs via RCA-to-XLR adapters or adapter cables, as was found to be the case with the Ref 150 used by the OP in this thread we had participated in some time ago. I believe it would also be consistent with low CMRR, due to the gain and other characteristics of the two paths not matching precisely.

Al, that makes a lot of sense to me, but both Warren Gehl (of ARC) and Kalvin told me that the amp employed a differential amplifier at its input. But they may have been reading off of the same cue sheet, which may not have been accurate. Although I've known both of them for 40 years, on this point I'm more willing to believe your theory as it is consistent with the amp's behavior. I never had the heart to to tell them that if what they said was true that a lot of performance was being left on the table. But who knows- maybe that will be part of the next iteration.

At any rate, that aspect of the amp's performance is well-known and acknowledged by ARC. So in the case of a balanced interconnect, if noise were able to impinge the cable, the amp would not be very good at rejecting it (in the old days this was often handled by an input transformer, which is usually very good at CMRR). So this would seem to make the characteristics of the cable more audible. For such an amplifier, I would recommend a cable that is double shielded.

Al, that makes a lot of sense to me, but both Warren Gehl (of ARC) and Kalvin told me that the amp employed a differential amplifier at its input. But they may have been reading off of the same cue sheet, which may not have been accurate. Although I've known both of them for 40 years, on this point I'm more willing to believe your theory as it is consistent with the amp's behavior. I never had the heart to to tell them that if what they said was true that a lot of performance was being left on the table. But who knows- maybe that will be part of the next iteration.

At any rate, that aspect of the amp's performance is well-known and acknowledged by ARC. So in the case of a balanced interconnect, if noise were able to impinge the cable, the amp would not be very good at rejecting it (in the old days this was often handled by an input transformer, which is usually very good at CMRR). So this would seem to make the characteristics of the cable more audible. For such an amplifier, I would recommend a cable that is double shielded.

At the risk of coming across as overly defensive (I own and love ARC gear), this sounds just a bit too conjectural/speculative. Warren Gehl to my knowledge does some very limited design work for ARC and is key in QC by reportedly listening to every piece of gear hooked up to his reference system of a Ref 6 and Ref 150se and some old, large Magnepans before the piece leaves the factory. It was, however, Ward Fiebiger who took over the reigns from Bill Johnson in the actual engineering/circuit layouts of ARC's top-level gear including the Ref series of preamps and amps. Now does that make a difference? Probably not. Surely Warren knows enough to answer the question being posed here. I actually tried to ask ARC the very same question, more or less. Guess what they said in response to my email? That I should contact my dealer! Very nice, huh? 

FWIW, I'm looking at my manual to my Ref 150 SE and it does not include a schematic and instead only has some specs, including that the output polarity is non-inverting with "Balanced input pin 2+ (IEC-268)"
Also interesting to me is this verbiage;
"IMPORTANT
Use the best available speaker wires and interconnects. Audio Research cannot emphasize this enough. As better components and systems are developed, it becomes increasingly important to avoid the limitations of inferior system interconnections"