Krell FPB600 can it be bridge to mono

Also, regarding "how would it know a transient is coming to switch it to a higher mode?," while I don't quite understand the reference by Mr. Colloms to "high-speed current-feedback circuitry," presumably the delay from amp input to adjustment of output stage bias is less than the propagation delay through the signal path from amp input to the input of the output stage.

There have been a few designs that have appeared over the years from other manufacturers which also changed output stage operating conditions in anticipation of signal requirements, except that what was being changed were the DC voltage rails supplied to the output stage. In those cases that I am aware of the purpose was to reduce power consumption and heat, and consequently cabinet size, weight, and cost, rather than to maintain what might be considered to be class A operation. Some of Bob Carver's older "magnetic field power amplifiers" were one such example.

Best regards,
-- Al

"There have been a few designs that have appeared over the years from other manufacturers which also changed output stage operating conditions in anticipation of signal requirements, except that what was being changed were the DC voltage rails supplied to the output stage."

That sounds like how my old Hitachi sr804 Class G receiver worked if I recall correctly.

It was a very nice receiver for its day, but the sound quality paled compared to many more modern designs. I tended to attribute that to the Class G operation as described, but there could easily be more to it than that.

Class G never seemed to take off though. Newer Class D amp technology seems to have taken over that niche. Sound quality is quite good with Class D I have heard and own already but I tend to think there is still useful untapped potential still to be realized down the road with that particular technology.

Krells class A plateau biasing network, is not as good as normal high biased or pure class A, to spell it out simply.

1: The network senses the transient and raises the class A automatically to keep the transient in class A.

2:It then keeps that class A level for a few seconds just in case there's another transient following or not, if not it dies back down to Class B

3: The trouble with this system of sensing and raising is that the initial transient that comes along has long gone through to the speakers before the network has had a chance to raised the Class A. So it or the first part of it still in class B

4: So your not hearing the initial transient in class A it's B, and that's why this system will never be as good as normal class A biasing but then you get the continuous heat problem with that, and much larger heat sinks,

I use to water cool my heatsinks jackets, it was the only way for me to get 150w pure class A without the size and weight.

Cheers George

Perhaps it is possible to respond fast enough with the bias to peaks at the input, since signal is subjected to small delay passing thru the amp?

I don't think you could delay the signal so the initial plateau biasing can stay in front of it, without recording and playback either by digital or tape means, this would be even more detrimental to the sound quality.

Back when I was building those water cooled class A monsters in the 1970-80's. My then boss and mentor Steven Deratz invented and patented the first electronic sliding class A bias system, it was good but never out performed the real thing.
After a few years he let the patent run out. Then I believe the first commercial system I saw of the same, came from Technics and many years after Krell bought theirs out.

Cheers George