My first push-pull vacuum amplifier

[DangerMan]

Hi Miguel,
When you connect the phase splitter to the output stage the amplifier starts to oscillate.
One or two things I don't know: Is the rest of the amplifier already built and connected up, that is, the input stage? And is the negative feedback already wired in?

If this is so, I am wondering whether the feedback is actually positive and not negative. An easy way to put that right with the least amount of trouble would be to cross over the feeds to the output valve grids from the phase splitter outputs.

Pete

[Miguel Lopez]

Originally Posted by Ed Dinning

it would save a resistor

That is very good

Originally Posted by Ed Dinning

and up to a certain level of HT the valve was rated for it

Sorry Ed, not sure if I understood this part. Can you explain a bit more? Thanks mate.

Originally Posted by DangerMan

Is the rest of the amplifier already built and connected up, that is, the input stage?

Yes, it is. I did this test with the input shortcircuited.

Originally Posted by DangerMan

And is the negative feedback already wired in?

No negative feedback yet. I always try to make it work first, and then apply the negative feedback.

[Diabolical Artificer]

Think what Ed was saying Miguel is that in a lot of designs 6V6's were run over their recommended max specifications and that's cool, but obviously up to a point. They wouldn't be happy with 1kv HT for instance but a bit over rated max HT is ok. Bit like running 807's with 400v on the screens.

The amp didn't oscillate in the first layout, what has changed?

Apart from changing your layout or putting ferrite beads on wires on the valve base the only thing I can think of is to put a 100 ohm R right up tight to the anode.

Andy.

[Miguel Lopez]

I understand that they could be run at higher HT than specified in the datasheet, what I don't understand it's why do not connect the screen to HT via a resistor. I understand that saving a resistor is a good reason, but that is an economical reason, not technical.

In my opinion if you use a higher HT you should take care with other parts of the circuit, so a bigger resistor in the screen good be an option for me to protect the valve, so I don't understand why they connect it directly to HT+.

Originally Posted by Andy

The amp didn't oscillate in the first layout, what has changed?

The fact is that I tested the circuit in the breadboard without the OPT, which wasn't winded by that time. I connected a couple of 56 ohms resistors in the anodes and it worked fine. It seems that the PL36 has a personal problem with the OPT
I don't think that it could be the layout, as in the breadboard it was really messy, and it did not oscillate.

By the way, yesterday I tested the 6V6 and they worked very well. I used them with HT+ = 230V, but I want to run them at 300 or 320V. I got about 4W of clean signal in the speaker. Maybe I will have to rewind the OPT, but that won't be a problem. I'm getting closer.

[Miguel Lopez]

I'm really puzzled with the amplifier.

I spent the whole weekend working on it and the nasty oscillation did not stop. I tried diferent configurations, including several phase splitters and outputs. The fact is that if I include three stages, it oscillates, and I don't know why. With the output and the phase splitter only it works fine, but there is not enough gain, then I included another stage and it oscillates.

Oscillation occurs without signal, so it is inherent to the circuit, I guess. I tried several pre-amplifier and phase splitter configurations, including a concertina, and it oscillates all the time, when the whole circuit is connected together.

The pre-amplifier and the phase splitter alone do not oscillate (two stages). The problem is when I connect three stages. I really don't know what is happening there.

Maybe if I connect the negative feedback it would stop oscillating.

Help!!!! I need somebody!!!!

[Diabolical Artificer]

Any chance of a schematic Miguel? It might help diagnose the problem, but I suspect it's your output transformer. Something I read rings a bell, which I can't put into words, something about positive feedback and phase, something about a loop being formed somewhere. Sorry, I can't be specific, I'll look through all the articles I have, it may take me a while.

I guess the thing to do is scope the oscillation, at what frequency is it oscillating at?

Don't despair, you've overcome worse problem's, Andy.

[Radio Wrangler]

If it's unstable without intentional feedback, Miguel, it's still going to be a problem with feedback applied.

Oscillation is caused either by unintentional feedback, or sometimes by the creation of negative resistance.

I'd go hunting for some way that some of the output leaks back into the input stage.

David

Power supply decoupling (or the lack of it) springs to mind.

[Diabolical Artificer]

Had a look at what N Crowhearst has to say last night about stability. Niquest diagrams were involved and phase angles, which I didn't quite grasp. But the upshot is, that when some of the output is fed back to the input,and these two signals are out of phase with each other, you get oscillation or instability.

So as others have said, stop the output signal returning to the input by de-coupling your power supply at V1/input- if you know what frequency amp is oscillating at, an RC filter found by 1/(2*pi*f*C) to filter it out or just make f=1hz , try direct coupling of V1 to PS or change value of coupling cap and reduce gain of V1.

I'm still learning, so apologies if I have some of wrong and I know you probably know all this anyway mate, but hope it helps.

Andy.

[Miguel Lopez]

See the attached files. Those are the circuits that I have tried. See the arrangement of the power supply. I also tried an ECC83 with one triode as amplifier and the other being a concertina phase splitter. Oscillation always occurs when I connect three stages.

I'm using bigger capacitors than those in the pictures.

Oscillation is about 2kHz

[peter_sol]

GEC in their amplifier book recommend a resistor and capacitor combination between each output valve anode and the centre tap in cases of low grade output transformer
Try 0.01 caps only or higher and see what happens.

[Herald1360]

Try decoupling the Screen feed to the output valves between the common feed resistor and the individual stopper resistors?

[Miguel Lopez]

Originally Posted by Herald1360

Try decoupling the Screen feed to the output valves

A big value capacitor to ground? Let's say 150uF?

[Herald1360]

No, just something reasonably low Z compared with 1K at the frequencies of interest, though if you've got the 100uF at the voltage rating it won't harm to try it.

No guarantees, but it's easy to do.

[Miguel Lopez]

Hi, I'm here again

I have been away from the forum due to work. On the other hand I had nothing to share on this thread as the amplifier is not finished and I decided to give myself a break, in order to learn more about this issue. I wasn't able to supress the oscillation, and this project has taken me a lot of time (more than foreseen), so I want to make a stop to go on with some other projects. I will resume this one, but not soon (maybe next year). Mods please, do not close the thread if possible.

I will open other threads soon, to share my other projects.

All the best from sunny Cuba

[Diabolical Artificer]

Hola amigo, good to see you back.

All the best from the sunny UK. : )

Andy.