Single supply, three-pole compensated class-B retro amp

[LKA]

I never saw 3P compensated amplifier so I tried to build one. The basic design is come from the local forum's guru, Danhard. I made fine-tuning to adapt it to the real conditions. Three boards were built, the biggest problem was the correct placement of components and routing so as not to cause distortion degradation. The measured 1kHz/4R thd is under 0.0001%, the 10kHz/4R is about 0.0005%.



As you can seen the loopgain slope falls 60db/decade View attachment 1044059

 

[MarcelvdG]

Does it recover well from clipping? Did you need to do anything special for that?

 

[LKA]

not perfect but quite good

 

[BSST]

Very impressive!

Would you post a schematic? Any before-the-mod data?

Thanks.

 

[LKA]

9 transistors, no CM/CCS/CASCODE, only topology elements known from 60's
omitted emitter resistors at the output, idle 3mA, overall 33mA
world champion

 

[LKA]

IMD 19k+20k 1:1 (CCIF), load 4R

 

[BSST]

Thanks, LKA!

I recall building something with a similar single-transistor input stage, from Popular Electronics, circa 1970. I believe that design came from an RCA app note. I've no idea re its distortion performance. This amp is way more impressive.

I see the 3rd order compensation network. Do you have a copy of the original schematic, or a Link? I'm curious about other compensation subtleties as well.

Again, impressive work!

Many thanks,

Steve

 

[Elvee]

Truly impressive and daring. Bravo!
Fortune Favours the Bold

 

[Bigun]

I love it. I got as far a singleton input, buffered VAS and two pole. Three pole is out there!

 

[dreamth]

Cool stuff! Will you , please provide the asc file?

 

[gannaji]

Subscribed.

 

[Nelson Pass]

Nice piece, especially the pcb layout.

Are you available for adoption?

 

[LKA]

THX

When I wrote "basic design" I did not mean the "root design". I don't know the origin of the circuit, but these are similar
https://sound-au.com/project217.htm
https://www.diyaudio.com/community/threads/retro-amp-50w-single-supply.236256/

I am preparing a pcb for a 60-75V power supply, which also includes an output coil. This board will be available, gerber or bare , I haven't decided yet.

 

[dadod]

I never saw 3P compensated amplifier so I tried to build one. The basic design is come from the local forum's guru, Danhard. I made fine-tuning to adapt it to the real conditions. Three boards were built, the biggest problem was the correct placement of components and routing so as not to cause distortion degradation. The measured 1kHz/4R thd is under 0.0001%, the 10kHz/4R is about 0.0005%.



As you can seen the loopgain slope falls 60db/decade

I see you are using simple voltage probe to simulate Loop Gain, better to use Tian probe it will give you more accurate result.
If you prefer simple voltage probe try to plot -V(out)/v(b), it simpler to read the phase margin.

 

[LKA]

The presented loopgain captures the EF output at a quiescent current of 80 mA (linear region). BTW, the circuit can operate in AB mode with emitter resistors 0.1R-0.15R. (the main goal was not to use emitter resistors). This is the second version of the PCB, assemled with those resistors. (for testing and proving the AB mode)

 

[latala]

Love it its so nice to see something different I like the capacitor output the buffered vas both on input and output

Trev

 

[LKA]

ltspice simulation files

 

[dreamth]

Thank you Sir!

 

[dreamth]

I thought that 10khz distortion isn't really that important...

 

[dreamth]

Then why not having t3 and t4 at the same temperature as long as we like even harmonics and lower overall noise?