Cuircuit for simplistic current amplifier

[xmvlk]

There is no reason why you couldn't take an IC OTA front end and buffer the output with a pair of current amplifying current mirrors from each supply.

But we do not have option of global feedback via one resistor {which power OTA could have
analogically to "emitter" - ground  resistor with OPA660's "emitter"}
BTW I could not imagine purely transilinear mirrors done from one discrete low-power
and one power transistor. I can imagine, that temperature change in P.A. transistor will results
in dramatic current gain change. May be some power HEXFETS have some kind of "current sensing leg", but such that elements are switching types with all problems. And some
experiments with switching semiconductors in linear mode commonly results in kind of
short-wave transmitter.

 

[abbey road d enfer]

I do mean, that circuit like Howland and so on are
not suitable for power circuits. There is sensing resistor and feedback floating over full supply swing.

Why wouldn'it work? What is so different? The Shure Musicmaster PA system of the 60's, based on that topology, delivered 100W. Incidentally, the reason was not to make it sound better; it was their protection system against short-circuits (the loudspeaker connections being on 1/4" jacks).

 

[xmvlk]

Shure Musicmaster PA system of the 60's, based on that topology

Please, can you specify it more (type) or get the link to the schemo?  May be interesting.

BTW I do not say, that IT would not work, i say something about effectivity. And when
signal current trough "emitter" of my hypothetic "power OTA" will be 1/100 of output signal, the
feedback resistor will be smaller.

 

[abbey road d enfer]

Shure Musicmaster PA system of the 60's, based on that topology

Please, can you specify it more (type) or get the link to the schemo?  May be interesting.

I had the name wrong, it was called Vocal master. I had not seen this schemo since about 40 years. They used a transformer for current-to-voltage FB.
http://edmullen.net/manuals/Shure_Vocal_Master_pro_va300_ug.pdf

BTW I do not say, that IT would not work, i say something about effectivity. And when signal current trough "emitter" of my hypothetic "power OTA" will be 1/100 of output signal, the feedback resistor will be smaller.

I don't get it at all. What is the "effectivity" problem? Efficiency as in ratio of useful power to supplied power or accuracy of the actual behaviour as a current-source?

 

[JohnRoberts]

Interesting.. I've seen a transformer used like that a lot more recently to deliver near zero output impedance to deliver silly high damping factor.

Once again the ancients are stealing our ideas...  :

JR

 

[xmvlk]

I don't get it at all. What is the "effectivity" problem?

In fact, that full output current must go trought feedback network.

 

[abbey road d enfer]

I don't get it at all. What is the "effectivity" problem?

In fact, that full output current must go trought feedback network.

Yes, but it's not a problem, a low-value resistor can be used, much smaller than the load, so only a fraction of the output power is dissipated in it. In the case of the Shure Vocal Master, the sensing resistor is 0.12 ohms. The nominal load is 8 ohms. For 100W in the load, the loss in the sensing resistor is 0.12/8.100 = 1.5W

 

[xmvlk]

http://edmullen.net/manuals/Shure_Vocal_Master_pro_va300_ug.pdf

Yes, and here are not problems with amplifier paralel-connection.

Somewhat more elegant solution {which behaves like pure-voltage amp} is here:

 

[abbey road d enfer]

Sorry, I don't understand the text, but according to the drawings, it is standard load-sharing technique as used in RF transmitters. Well, at the time (1949) that may not have been standard...
They rely heavily on transformers for implementation of I-to-V feedback.
Today, this would be done with differential amps (for audio apps).