Thanks for the input and well wishes, Keith. One thing still baffles me about an amp's ability to drive low impedence loads is the notion that the amp should double the power as the impedence halves. This is supposed to be indicative of high current delivery capability as the load changes. Yet I see amps like the 7/14B that only increase 50% from 8 to 4 ohms and actually decrease by about 30% into 2 ohms. I've even read Jim Thiel make this statement when describing a high current amp that would be suitable to drive his CS3.7's while citing Bryston as a favored amp vendor (see James' description of the Thiel room at CES this year). Confusing to say the least. How does the 7B do so well given this paradox?
Bill
Hi Bill;
The 'paradox' is only apparent, not real. The correct answer is to go back to what an amplifier 'does for a living', i.e., what is its design goal? Its actual task of course is to amplify a musical signal and present it to loudspeakers in such a way that there is nothing audibly added to or subtracted from. No noise, no distortion, just the music with all its subleties and nuances intact.
The fact that this is an extremely challenging goal is evident in several ways: First, music is an exquisitely complex and dynamic phenomenon. It's freqency range spans over a 1000:1 ratio, its dynamics range over ratios of many millions-to-one, and music changes many times per millisecond. It is by definition never static, always in flux, always changing.
Second, a loudspeaker is itself a very changeable entity, with impedance spans of many Ohms, dynamic back-forces, energy-storage, phase-shifts over the frequency range, and altering with temperature as well. With almost no exceptions, loudspeakers range from the extreme low of a couple of Ohms, up to a few tens of Ohms over their working frequency range, with combinations of reactances and resistances in that impedance curve. That's what the amplifier has to deal with.
Third, of course we have seen over many years one or another supposed 'final answer' to amplifier design come along, again and again, and ultimately lose favor, again and again. It always turns out that the simple assumptions are not complete, not necessarily valid, and that if some is good, more is not better forever. Slew rate was once taken to an extreme, with manufacturers claiming 1000V/uS was necessary for music reproduction. MOSFETS outputs were considered the 'only' way to go for a while. Tube vs. transistors was a hot topic for a long time. Now we have some manufacturers claiming that an amplifier has to 'double' its power rating every time the impedance is halved, down to 1 Ohm, or 1/2 Ohm, or whatever extreme impedance level is in fashion this year. (The amplifiers described this way don't actually double their power capacity with each halving of impedance, of course; the laws of physics always subtract substantial losses for heat and unavoidable resistances in the power train. The maunfacturers just measure the power at the lowest impedance they wish to advertise, and then cut their published rating in half for each doubling of the impedance, going up).
As mentioned above, loadspeakers operate over known impedance ranges. An amplifier has to drive them with preferably inaudible distortions over that whole range. Loudspeakers can display narrow-band dips in impedance to a couple of Ohms, so an amplifier has to have current reserves to handle them without distortion. Bryston amplifiers have output stages that can deliver pulses of current up to the limits required by a 2-Ohm full-band load on a dynamic basis, with real music. The output stages, power-supplies and heatsinking have been designed to an optimum size range to handle those musical demands. The steady-state power output is not as high as the dynamic power, however.
A 4B SST2, for instance, is rated conservatively at 300 Wpc at 8 Ohms, 500 at 4 Ohms. Close to double, but not quite. We could theroetically rate the amplifier at 250 Ohms into 8 Ohms so it would seem to 'double' at 4 ohms, to 500, but that just seems like gamesmanship, not honest ratings. And, although the amplifier can indeed deliver pulses of over 1000 Watts into 2-Ohm loads, its 2-Ohm steady-state measurements would be quite a bit lower. In order to supply 1000 or more Watts steady-state into 2 Ohms, the size, weight, cost and number of output devices would need to double, and double again for 1-Ohm ratings. A 4B SST would then weigh 200 pounds, be the size of a coffee-table, and cost $16,000. But here is the important part: It would NOT sound better on real-world loudspeakers! In fact, with the increase in number of output transistors that would be required, the HF distortion and transient response could suffer.
Bryston makes an amplifier that costs $16,000 and weighs 200 pounds for a stereo pair; it's called the 28B SST2. It's made to drive loudspeakers rated from 4 to 8 Ohms, with lots of provision for impedance dips to 2 Ohms or lower. It took a lot of time to figure out how to provide audibly perfect transparency at very low to very high output levels with its large and complex output stage. We could have designed that exact amplifier package to be rated at 1000 Watts into 1 Ohm, 500 into 2, 250 into 4 and 125 Watts into 8 Ohms. What would it have sounded like on a real-world loudspeaker? Like a 125-Watt amplifier, naturally. Instead, we think the real 28B is the most accurate, most musical, most dynamically thrilling and riveting amplifier made, period. And an awful lot of reviewers seem to agree.
My point is that, for equivalent 8-Ohm ratings, I will stack any Bryston amplifier up to any other product, and it will much more than hold its own. There will indeed be amps that are much larger, much heavier and much more expensive for the same 8-Ohm rating, but they will not sound more musically accurate, I guarantee it. And that is the whole point of designing amps. It's all about the music.
I hope the above is helpful, but please let me know if you have any other questions. Thanks for thinking of Bryston!
Chris Russell
