David Houpt wrote

Anyone looked at this concept from Douglas Self and Cambridge
Audio?

Comments?

I don't know anything about the technology but I am interested to
see how the new expensive 840 range that includes this technology
will be marketed by Richer Sounds.

Having no demonstration facilities (lost to make way for TVs in many
stores) and a mean policy on home trial (overnight only) may be OK
at the 'normal' Richer Sounds hifi price points but not at nearly
£800 a pop. Or do you think that punters will pay out on the basis
of the positive reviews that 840 range has got so far?

I must be honest and say that I am very skeptical about high priced
solid state kit. When down sizing from a huge house to an apartment
a year ago I changed my hifi to smaller units and listened to a lot
of CD and SS amp combinations in the process. There seemed to be
little difference between £500 and £1500 total spend in sound terms.
Speakers, well that's another matter...

Regards


Yes, if I bought an SS amp for more than a few hundred, I would expect
the extra to be reflected in build quality rather than sound.

There is an article in Elektor written by himSelf, so I thought maybe
plenty of ppl would be aquainted with this class XD thing.

AFAICS, class XD is achieved by drawing current from a standard PP
output stage optimised for class B operation, using a current source.
This has the effect, as Self puts it, of displacing the crossover
point, hence XD, "crossover displacement". This puts the stage into
class A for a proportion of its operation, claims Self. The difference
between this and ordinary class AB is that the class B part can be
optimised, and the crossover distortion can be minimised. Further, the
current source is varied by the signal, so the proportion of available
class A tracks the signal. There is not much detail about how this is
done.

Seems to me there are a few issues he avoids mentioning. He justifies
the design by comparing its distortion to the same output stage
operating in class B and AB, showing that it always remains below
either. He doesn't show the characteristics of the stage operating in
normal class A, though, which is an obvious omission. Neither does he
ever mention that the class A part of his design is single ended.

Not knowing much about transistors and the circuits they inhabit, I'm
not sure that it's class A *is* SE, but it seems a necessary outcome
of shifting the crossover point?

Thinking of how the idea might be applied to a valve amp, you could
set a class B operating point, and then offset the bias so that one
valve is operating at its optimum class A point and the other is
turned off. You then have SE class A up to the point where the signal
reaches the offset, beyond which the operation is class A + class B.
All very well, except the impedance matching can't be right for both.
In the case of valves, the transformer would need to be gapped too.
And the valves would age at very different rates, and....

I don't trust Self much. He is very economical with the truth. Same
with his articles on class G.

cheers, Ian