Single - Ended EL34.

[Lord Rockingham]

Here we have the latest build in my ongoing DIY saga. It’s a 5W, single-ended EL34 amplifier.
The EL34s are strapped as triodes, and the amp uses three stages of amplification, beginning with a 6SL7 input stage. From there, the signal is passed to a 6SN7 driver stage before being sent to the power tubes.

Global feedback is employed from the output transformer secondary back to the cathode of the 6SL7 input stage.
The received wisdom is that global feedback is a complete a no-no with a single-ended amp, but because I seem to find it inordinately easy to make life difficult, I decided to learn all about the concept so that I could apply it properly: you know - give it a chance and all that. To a non engineer like me, the maths involved was bloody difficult, especially with regard to the high frequency compensation required in order to avoid the generation of high, odd order harmonic distortion.

I’ve enjoyed this project a great deal and the sound is quite lovely. For me, it vindicates feedback, provided it’s done properly.

 

[Lord Rockingham]

Two scope traces of a 10KHz squarewave. 10KHz squarewaves never look too clever when an output transformer is in the way. Transformer designers go to a lot of trouble to get good squarewave performance, which is why good output transformers are expensive beasts. Mine are cheap Hammond 125FSEs but they sound surprisingly good.

10 KHz squarewave with global feedback applied without HF compensation: Nasty

The same squarewave after HF compensation: Much better.

 

[Lord Rockingham]

1KHz:


100Hz. Slight bass rolloff but these are cheap output transformer:

 

[rabski]

The results (what it sounds like) are all that matters. Conventional wisdom is worth referring to, but often also worth ignoring. Sometimes it ends in triumph and sometimes disaster, but that's half the fun.

 

[Lord Rockingham]

Aye, the sound is what matters. This project was mostly an academic exercise both with the maths and the measurements. TBH I wasn’t prepared for the high standard of sound quality achieved at the end of the process.

 

[Disarmamant]

The results (what it sounds like) are all that matters. Conventional wisdom is worth referring to, but often also worth ignoring. Sometimes it ends in triumph and sometimes disaster, but that's half the fun.

As Jack Elliano once told me “we don’t listen to square waves”

 

[Lord Rockingham]

As Jack Elliano once told me “we don’t listen to square waves”

Yes, but I bet his squarewaves are still better than mine.

 

[toprepairman]

Indeed but they can tell you an awful lot about where things are going wrong.

 

[audio_PHIL_e]

As Jack Elliano once told me “we don’t listen to square waves”

There must be some on Switched-on-Bach

 

[Lord Rockingham]

For the sake of completeness, here are the schematics:

 

[rabski]

Thanks Steve. Nice that.

As a matter of interest, did you try it with no feedback at all?

 

[Lord Rockingham]

Thanks Steve. Nice that.

As a matter of interest, did you try it with no feedback at all?

No.
It was designed from the ground up to use global feedback.
I could have tried it with no feedback, but the gain would have been ridiculous.

 

[rabski]

No.
It was designed from the ground up to use global feedback.
I could have tried it with no feedback, but the gain would have been ridiculous.

There's probably all the relevant bits in the parts pile here, so I may have a go at one for the 'build fun'. Not really enough power for me, but certainly worth it for late-night gentle stuff.

 

[Lord Rockingham]

Yes, working with global feedback in a single-ended context is interesting. I’d gotten fed up with the same old same old, so decided to try out feedback, despite the perceived opposition to it from some of the SET cognoscenti. I mean, what could possibly go wrong?

Having built an SE amp that employs GNFb, I have to say that I can’t see the problem. The sound is great. The main caveat, is that if you are going to employ GNFb in a single-ended amplifier, or a push-pull for that matter, it is definitely a case of ‘ you can’t just.’ It needs to be taken seriously at the design stage if you don’t want to end up with instability, an oscillator and blown tweeters or worse. This means HF phase compensation arrangements, which mean more components, which some might argue, means worse sound.

However, if anyone asked me now, whether I would build a non DHT, SE amp without feedback, I would have to answer, “not on your nelly.” YMMV and all that of course.

 

[toprepairman]

Er, except that it is non DHT???

 

[rabski]

Er, except that it is non DHT???

And Steve didn't not build it without feedback.

So many double negatives, that I also lost the plot Henry

 

[steve 57]

I'm looking forward to hearing this one next week at the audio talk meet up

 

[Lord Rockingham]

@steve 57

@rabski and @toprepairman
I just had another look at the last sentence of my last post and I haven’t a bloody clue what I am rambling on about. And me with an English degree

 

[rabski]

@steve 57

@rabski and @toprepairman
I just had another look at the last sentence of my last post and I haven’t a bloody clue what I am rambling on about. And me with an English degree

Or, a degree of English?

 

[audio_PHIL_e]

English lecturer: "In English, a double negative is logically a positive; in Russian, a double negative is emphatically a negative. There is no language in which a double positive can be a negative"
Voice-at-the-back: "yeah, right"