Re: Brook amplifiers, sliding bias.

On Mar 28, 9:31 pm, Patrick Turner <i...@xxxxxxxxxxxxxxxxxx> wrote:
Googling Brook amps brought up

http://www.tubecad.com/2007/12/blog0128.htm

Food for thoughts.

Patrick Turner.

Certainly is (I've saved the .pdf for future study.) But I'm
wondering about the benefit if the signal source amplitude is truly
random. As I understand it, there is a roughly 6 Hz lag on the new
bias settings (for good reason!), so the higher, Class AB, bias would
actually "not be there in time" for a fast, large transient after a
quiet passage - the transient would just hit the low bias class A
setting which would then start to move slowly higher towards AB.
Philosophically, you'd need a predictive feature for this to work - no
such thing in real world signal processing. Of course, today we could
use a high-bandwidth, aperiodic delay line so that the bias could be
set up in preparation for the coming signal level, thus delaying the
whole program (no problem, it's a recording so just look ahead on the
CD!) No chance of doing that in 1947... but now we've little need to
do it as we have "infinite voltage and/or power thick straight wires
with gain", i.e. large solid state amplifiers. I did say this is
philosophical! What have I missed?
Cheers.
Roger

.