Another idea

[Bernhard]

Maybe this has already been used in some exotic amp ?

Two current sources, one for positive and another for negative output current are modulated so that current will flow through a connected load, following more or less the input waveform.
No negative feedback is used, the waveform will be slightly distorted.

The output is connected to the output of a SE classA amp.
This amp determines the output voltage because the current sources only deliver current as they are told to.
No fighting between those two, the classA amp carries the low current necessary to correct the errors of the current sources.
Low power & high speed / low capacitance transistors could be used for the classA amp ?

Works well in sim.

What U think ?

This can be combined with constant power for gain device.

[tschrama]

I had a similar idea, and it works for constant loads, but in complex impedances it failes....

[lineup]

Originally posted by Bernhard
Maybe this has already been used in some exotic amp ?

Two current sources, one for positive and another for negative output current are modulated so that current will flow through a connected load, following more or less the input waveform.
No negative feedback is used, the waveform will be slightly distorted.

The output is connected to the output of a SE classA amp.
This amp determines the output voltage because the current sources only deliver current as they are told to.
No fighting between those two, the classA amp carries the low current necessary to correct the errors of the current sources.
Low power & high speed / low capacitance transistors could be used for the classA amp ?

Works well in sim.

What U think ?

This can be combined with constant power for gain device.

The output is connected to an output.

I cant get how this circuit works.
But sounds very interesting

As you have simulated, you might be able to show me a schematic
illustrating this princip ....

thanks

[ilimzn]

Output voltage is controlled by output impedance and the difference in currents from the current sources. Class A SE amp used as an error amp indeed DOES fight the current sources but that is how it effectively modulatess output current by either sourcing or sinking a correction current.
The only problem I have with it is this: what possible advantage does this have over a regular class A or class AB amp? I can only see a disadvantage, namely since the current sources that are responsible for most output current haave no feedback, this will fail miserably into reactive loads unless the class A amplifier can correct with currents on the same order as the current sources - in which case efficiency is lower than even SE class A.

[tschrama]

Hi ilimzn,

indeed, and just to explain a bit...

imagine the reactie-speaker impedance drops to halve it's nominal value

The openloop current sources have no idea of this, so their current output stays the same

However for constant voltage output, the current should be doubled, so the SE-ClassA amplifier suddenly has to diliver 50% of the needed current.


that's no longer a small correction amp, so the advantages of this system are gone.




If you were building a constant current ampifier however... this could work.. but you need a current sensing circuit, not a voltage sensing circuit..