As John points out, if you can't be in Class A all the time, then
you try to have the current through the gain devices approach
0 asymptotically so that the transition is gentle as possible
and occurs at as high a level as possible. In this way you can
avoid high-order artifacts.
The 800A accomplished this asymptotic approach by bending
the transfer curves with a circuit that resembles the later
Hawksford / Cordell error correction circuit. Instead of
seeking to correct error through feedback directly, it adjusted
the bias in a continuous curve so as to maintain forward
conduction in real time.
By contrast, the Krell circuit was step-switched to higher
discrete values as needed, and then maintained at that
level for a period of time. As it is abrupt, it is detectable,
although it doesn't occur very often in the course of a
piece of music.
😎
If you read the Krell blurb it states that as long as the heatsink (effective transistor temperature with quite a few C extra) isn't too hot it'll supply a class A signal. However, if the heatsinks (transistors) become too hot, IE you are attempting to deafen yourself, they'll cut back the plateau bias to a level that isn't injurious to the power transistors. In other words it goes to AB biasing 😉 Once things cool down it'll allow the amp to use one of the higher plateau bias again.
I know you stated that deratives from Japan mainly (I had a pair of Denon POA-8000's in the eighties that were marketed as 'dynamically biased NFB designs') spoiled it for you to continue with this concept. Do you think it had technically spoken any potential of being explored/developed further? Would you consider in the future to go from static back to a further refinement of the dynamic bias system? Theoretically such a system would never have to transfer from A to AB because the biasgenerator will supply as much current as needed (and can of course, nothing is limitless)to keep the transistors in their linear Class A region. Sounds great!
Thanks for the explanation. Should read their white papers better!
I think the problems with the dynamic bias clones is that
the idea allows for amplifiers with a low bias to refer to
themselves as "Class A" because the output stage retains
a nominal forward bias. The engineering and marketing
departments created cool running Class A amplifiers. With
the low bias, I don't think they sounded much like Class A,
regardless of the technicality.
The Threshold amps were biased quite high - typically they
idled at approximately their rated output. It was a short
jump to "pure" Class A.
😎
the idea allows for amplifiers with a low bias to refer to
themselves as "Class A" because the output stage retains
a nominal forward bias. The engineering and marketing
departments created cool running Class A amplifiers. With
the low bias, I don't think they sounded much like Class A,
regardless of the technicality.
The Threshold amps were biased quite high - typically they
idled at approximately their rated output. It was a short
jump to "pure" Class A.
😎
So what I gather from this is that your current line of amps with their high static biasconfigurations is the logical conclusion after exploring other techniques to get a satisfying linear response from semiconducting devices.
I always thought that those dynamically biased techniques were developed to free the Class A designs from their inefficiency and high thermal dissipation and accompanying expensive hardware. But in your case it was just a bridge to those full Class A designs.
Do you see in the near future any new ways (techniques) or devices that can be implemented by audio highend manufacturers to develop products that run relatively cool but deliver the same soundquality as let's say your current line of XA amps? Do you see any attempts now in your field that could be succesfull in this context?
Thanks for the enlightment so far.
What the world in these global warming days needs is low bias class A amps!
But here in Oslo it is -20 degrees C, so I am running all class A amps all the time!
But here in Oslo it is -20 degrees C, so I am running all class A amps all the time!
Aren't we a Class of Dinosaurs?
If you were living in the Netherlands the police, alarmed by the power company, would be investigating if you were running a homegrown Marihuana (Cannabis Sativa) plant. ;-)
They don't call it Class 'A' for nothing.
American educationratings run from A through F, A being the highest rating!
To be serious:
I'm not targeting this particular amplifier topology for environmental reasons.
I do believe in the principle that to fulfill a legitimate desire, the solution should tax our natural resources as low as possible.
I accept that fullfilling the need for high quality musical sounding poweramplifying devices can only be realized in optima forma with the deployment of Class A enrichment.
Curious as ever and not hindered by any knowledge I wonder if the experts in the field can envision with the current state of technology another solution for producing linear operating circuits without swapping them with loads of current.
Would a "Current Displacer" (loke fig 5 of your article "Leaving Class A") would achieve that "asymptote"?
Thanks
Fab
Last edited:
- Status
- Not open for further replies.