small a-class power amp non global nfb

[padamiecki]

Hello
I would like to introduce my latest projects that I have made.

Those are simple a-class low power amplifiers with only a local feedback, however the amp from the 1st option is almost nfb-free (the natural feedback is only from used parts and is an inherent character oof the used devices).
The amplifiers are enough to home-listening.
The signal from cdplayer was feed through the 47k potentiometer.

I projected those topologies many years ago and last time seen them similar on a French site with mosfets. The world is small! But my first designs used a source resistors (both input and output). Latest do not!

After reading a few posts at diyaudio I decide to minimalise the nfb in fets/mosfets and introduce larger nfb for the bjts.
It was after an article in which there was proven that SE fets suffer from nfb (even a local) and produce high order odd harmonics.
It was shown that the thing with bjts is different: high nfb makes the distortions smaller.

So please look at the attachment.

There are 3 options on it.

1st option: hexfets as output devices: very good sound, fast bass, however I can hear the distortions when higher levels occurred.

2nd option: bjts as output devices: amazing sound but weak damping factor: lazy bass.

3rd option (that I have never build): a mix of above. To build this an experiment with a re value is needed. Probably there is a need to set the currents of hexfet and bjt equal.

I would like to see your comments both of the designs and the sound (30 minutes is enough to build this). How you expect, is there a chance in 3rd option to distortion cancellation and to improve the damping factor (of course yes but how much?)?

For those who wish to build this I wrote a few notes:
1. Output transistors on medium hitsinks,
2. Regulate the potentiometers to get a current through the output transistors at about 1 A and 0VDC at the output, Be carefull during regulating; I have burned at about 10 pairs of hexfets: start tha regulation from 0 ohm on potentiometers!!!
3. The output drift should be at about ±100mV,


Cheers!
[B]

[padamiecki]

hera is a more friendly image

[lineup]

I like it.
The Option 1. with IRF9540/540 MOSFET.
It wilol not have lowest THD and need a well filtered/regulated supply for best result.

[Bigun]

I've never built with MOSFETs but I remember a paper looking at distortion due to local feedback with FETs, Triodes and BJTs. The BJTs in EF configuration were better than MOSFETs and this is the direction I am going with my own ngb amplifiers at the present time.

Option 3 looks like something I've seen from wavebourn and I talk about it in one of my threads on the Valve forum. A friend of mine tells me it was popular in Russia many years ago. The BJT turns on first (0.7V Vbe) and then the MOSFET comes to the rescue when the BJT hits beta-droop.

It's worth checking out the valve forum, they have explored SS power followers to put after tube voltage amplifiers without nfb.

[Nelson Pass]

Looks like it will have a [I]lot[I] of gain, and I think you'll have to watch
the bias drift on those IR outputs.