10-05-10. Magfan: As near as I understand it.....Carver nulled the amps by somehow comparing thru a single speaker. When the speaker made NO sound....or some large value 'down', they amps were producing the SAME sound. The louder the speaker, the greater the difference between amps. This is the source of the 't-mod' which was part of my now ancient Carver Cube, the M-400t.
I don't remember the exact numbers, but just for example, when the speaker being used was say......nulled to -60db, you could say they were the 'same' to 1 part in 10,000 or some such.
ONE possible flaw in this ointment is that of the speaker used. If it was a highly reactive load, the tube amp would not be at its 'best' while a speaker of the single driver/full range persuasion may put the SS at a disadvantage. I doubt there is an 'easy' or even 'best' answer here.
You make an excellent point IMO, Magfan. A perfect or near perfect null with one set of speakers, or even a variety of speakers that represent relatively easy loads, would not be meaningful with speakers that are highly reactive, or that have impedance dips that demand more current than the amp's power supply can provide.
As you probably realize, the target amplifier whose transfer function the M400t attempted to duplicate was the Mark Levinson ML2, a pure Class A solid state amp, which although low powered (due to voltage swing limitations) could provide enormous amounts of current into low impedance loads. That effort preceded by a couple of years or so the Stereophile "challenge" where the target was a C-J tube amp, and was performed for and reported by Peter Aczel in "The Audio Critic." That was well before Mr. Aczel's later unfortunate conversion to the belief that all amplifiers meeting certain basic criteria sound the same.
The technique actually involved three speakers. One speaker was connected to amplifier A in the usual manner, except through long cables that allowed the speaker to be placed at an inaudible location. Another identical speaker was connected to amplifier B in a similar manner, and placed at an inaudible location. A third speaker, as well as measuring instruments, were connected between the red terminals of the two amps. It was reported that a 74 db null was achieved, with a variety of music, with the monitor speaker remaining dead silent except for a small ground loop hum. Which is pretty remarkable.
According to a follow-up article in "The Audio Critic" a few years later (Issue 10, in 1987), while the Stereophile challenge initially resulted in a 70db null between the prototype Carver 1.0t and the C-J amp, a later test of that same C-J amp and a production 1.0t resulted in only a 28db null. Unfortunately it wasn't determined if that was because the production 1.0t didn't match the prototype, or because of tube aging or other changes in the C-J, even though Carver had returned the original prototype to Stereophile to enable that determination.
I too used to use an M400t, which I still have although I haven't used it in several years. I found it's sound, with easy to drive speakers, to be quite respectable, but to me far less preferable than the sound of several tube amps I've had. The main differences being that it projected a considerably smaller image (that surprised me, and I have no idea why it would be so), and, not surprisingly, it lacked what I would refer to as the harmonic richness of the tube amps.
Best regards,
-- Al
