🚀 go-pugleaf

In article <11900019@hpmwnpd2.sr.hp.com>, allana@hpmwtd.sr.hp.com (Allan Armstro
> 
> 
> 
> P. S. -- This brings up an interesting question. Consider a speaker that
> has a very low impedance at the cabinet resonance (1 Ohm at 100 Hz, for
> example). A solid-state amplifier would provide the speaker with
> constant voltage. The lower impedance would mean that the amplifier
> would supply more power at the resonance (if you're not loud enough to
> be current-limited). Would this flatten the frequency response or make a
> big peak? If driven by a tube amplifier with an output transformer, the
> power match would suffer at resonance and less power would be delivered.
> I believe this would make for muddy bass, no? Would it be better to
> design an amplifier which monitored both output voltage and current and
> design feedback so it delivered constant power with frequency despite
> changes in the output load? Would it sound better?
> 
I cant offer a practical view here but it seems that most speakers today would
be designed and tested to have the flattest possible response with a constant
voltage style amp so anything else (assuming it could be easily produced) would
not produce the response expected.



> P. P. S. -- I may be out to lunch here. I'm just assuming that they're
> still designing tube amps with transformer-coupled outputs. Would it be
> practical to design an output transformerless tube amp? Since there are
> only one-channel devices available, it might be tricky, but the real
> problem would seem to be the very low transconductance of tubes. Is this
> it? If you can't get enough gm, can you make it up with a lot of loop
> gain?
> 
Output transformerless tube amps are possible and are being made, generally
with a pseudo complementary symatry or bridge output and a number of low
 impedance tubes in parallel for each arm.
 



> P. P. P. S -- That brings up another question. If you design a class A
> amplifier, whether it be tubes or transistors, how do you couple the
> output. I assume you're forced to transformer couple it. Does this mean
> that you get a lousy damping factor or are there practical ways to add
> enough feedback around the transformer?

Class A only means having current flowing through the output device(s) all the
time. A plain class B stage can go to pure class A by increasing the bias
current (and power supply and heat sinks) until it is greater than the peak
output current.