Control pots issue, help please

[Helmholtz]

Question please:
I have to look at a point where are created conditions for parasitic oscillations or at the point where this condition manifest the oscillation effect ? It is not clear for me where the intervention should be done. Need an advice please to get rid of this parasitic issue. Thanks.

[ATTACH=CONFIG]54423[/ATTACH]

Power amp oscillation is typically a system effect and involves interaction of many components.
As noted above the added gain stage within the feedback loop causes "huge" open loop gain and relatively high loop gain (always critical). Making it stable under varying conditions will require adequate frequency/phase compensation. Your schematic doesn't show this. A first and crude lag compensation measure could be wiring a 200pF(or somewhat higher) cap between the PI plates. Also a Zobel network across the OT primary might help. Or read up on lead compensation which is typically achieved by wiring a cap across the feedback resistor.

For a more systematic approach look here:
http://www.angelfire.com/electronic/...pensation.html

More information can be found here:
https://music-electronics-forum.com/...=amp+stability

[catalin gramada]

The amp without first gain stage looks very stable. Is very hard for me to understand why adding one more gain stage in front it make it unstable, as time feed if from external source shows no issues. Should be a diference how signal generator and first half of 12ax7 react with the rest of circuit, despite ths fact both are decoupled and connections was done to the same points. If I create for the first gain stage conditions to work exactly like signal generator does it will not be necessary any compensations as the amp didn.t shows no issues to be feed directly in second gain stage from generator. But if I leave the first gain stage as it is the conjuncted interaction with the all circuit components do a mess which should be fixed (compensated) exactly where the effect appear meant in power stage. So, my question comes in this sense : to try to do first gain stage to work more like signal generator does (meant will not require any compensation in power stage) ? or to leave it as it is (meant to accept it create conditions for osscilation over circuit) and to try to fix unwanted effects (osscilations) in the point where appear (power stage).
I talk about the same problem as can be fixed from the source or from the point where have effect.

[catalin gramada]

Power amp oscillation is typically a system effect and involves interaction of many components.
As noted above the added gain stage within the feedback loop causes "huge" open loop gain and relatively high loop gain (always critical). Making it stable under varying conditions will require adequate frequency/phase compensation. Your schematic doesn't show this. A first and crude lag compensation measure could be wiring a 200pF(or somewhat higher) cap between the PI plates. Also a Zobel network across the OT primary might help. Or read up on lead compensation which is typically achieved by wiring a cap across the feedback resistor.

For a more systematic approach look here:
http://www.angelfire.com/electronic/...pensation.html

More information can be found here:
https://music-electronics-forum.com/...=amp+stability

There actually good material. Thanks for share

[Helmholtz]

Amp oscillation is hardly ever produced at a single point by a single component. Rather it requires gain and feedback meaning interaction between different stages.
You say, the power amp is stable. This I doubt. Did you test square wave response feeding the 12AU7 from a signal generator via a 100k series resistor?

From what you describe I am quite sure that you don't have a preamp but a power amp problem. Feeding the NFB to the 12AU7 cathode means using the tube in grounded grid mode. But as the grid circuit is high (and frequency dependent) impedance, the grid "grounding" as required for grounded grid mode is probably not sufficient. In result the gain/frequency response of the power amp's feedback path will be influenced by the grid circuit impedance, which changes with the controls' settings. And in an amp with little stability margin this may give rise to oscillation. A CF in front a the 12AU7 should improve things.

[catalin gramada]

Amp oscillation is hardly ever produced at a single point by a single component. Rather it requires gain and feedback meaning interaction between different stages.
You say, the power amp is stable. This I doubt. Did you test square wave response feeding the 12AU7 from a signal generator via a 1M series resistor?

From what you describe I am quite sure that you don't have a preamp but a power amp problem. Feeding the NFB to the 12AU7 cathode means using the tube in grounded grid mode. But as the grid circuit is high (and frequency dependent) impedance, the grid "grounding" as required for grounded grid mode is probably not sufficient. In result the gain/frequency response of the power amp's feedback path will be influenced by the grid circuit impedance, which changes with the controls' settings. And in an amp with little stability margin this may give rise to oscillation. A CF in front a the 12AU7 should improve things.

Thanks That is something I can do. The only problem for now is my amp stopped to oscillate and work without issue even I pushed hard...very confused

[catalin gramada]

Okay, if I put a cf.in front of what is supposed to put into 12au7 grid? I want to know what impedance is supposed to drive,? It will be classic 1M or something more 100-200k around, please ? Thank You

[Helmholtz]

Okay, if I put a cf.in front of what is supposed to put into 12au7 grid? I want to know what impedance is supposed to drive,? It will be classic 1M or something more 100-200k around, please ? Thank You

I would use a 0.1µ coupling cap and a 1M or lower grid (reference) resistor to ground. Not sure if the 100k grid stopper makes sense.The value of the grid leak resistor should not be critical as the dominating grid load will be the low cathode impedance (a few hundred Ohms) of the CF.

Don't forget to do square wave response testing. If the square wave shows pronounced ringing, the amp is most probably not uncondionally stable.

[catalin gramada]

OK, this is what I did:



At first glance The amp did not shows any sign of oscillations even pushed hard. Still have to adjust some voltages. I will do some tests tomorrow and post the results. Thank You

[Helmholtz]

Any new results?

[Helmholtz]

Sorry to be insistent, but I'd really like to know if the CF cured the oscillation problem.

[catalin gramada]

Sorry, forgot to post the results but did not arrived at a final conclusion. With the CF the amp becomes stable and did not record any osscilation issue... but I reworked the preamp and tone stack, moved some components here and there, wiring more carefully, and at a time bypassed stage by stage to check any possible issues. I did not done yet but what can I say is the amp is stable also without CF in circuit and with 12au7 grid referenced to ground through capacitive tone stack. The circuit seems was very sensitive at a point and CF cured those but I wonder if those was caused by design or by careless layout build. I built an identical one few month ago and my buddy didn't report any problem with it.

But , yes the CF fixed the problem in first instance , before to rework the whole preamp.