AF Osillations (~3kHZ)

[rocket]

Hello,

I have build a 2 stage preamp with pentode-triode (ECF80) and it keeps oszillating at rather low frequenzies. Due to the bad lead dress and tight space I would understand RF oszillation but 3000 Hz is very low for parasitic effects isn't it?

What can I do (apart from lead dress, see tight space).

The amplification factor is a little more than 100 for the pentode and another 10 for the triode (measured). Oszillations start when I connect the pentode's plate to the triode's gate (with a cap).

[Enzo]

Disconnect the NFB and see if it stops. If it stops, then you need to swap the plate leads on the push pull power tubes - your NFB is actually PFB. If it is simpler, you could swap the speaker winding leads.

[rocket]

There is no NFB, and no power stage either, the preamp oscillates by itself.

[Enzo]

I am very sorry, I was thinking small combo amp yesterday, and today when I opened this thread the word PREAMP left out at me. Having another senior moment.

The freq will be whatever it will be, all the result of the feedback path and the components in it. When you couple the preceding plate to the following grid, you completed the loop.

What comes to my mind is power supply first. Are the two plates running right to the same B+ node? Are the two cathodes sharing ground? I might consider adding some more decoupling to the pentode - a few thousand ohms and an e-cap.

How about the screen grid of the pentode? How is it served with B+ and how much bypass does it have to ground?

ANd consider the amount of gain you have strapped together. I had a guy bring me in a GK250ML once and it sat there screaming. That amp has a gain adjust control for about every stage. The guy had gone in and turned all of them all the way up. You can only have so much gain before the input starts to notice the output and the circuit howls. He got there. I adjusted him back into the range of useful gain. If you have a gain of 1000, what is coming out of the amp if you plug 100mv of signal in the front?

If the preamp is pentode first, then don't connect the pentode plate cap to the triode grid, connect the cap to the output of the device. Now listen to it. How does the pentode preamp alone sound? If that is unstable too, then the pentode circuit needs work. But if it works, then pad the signal down on the way to the triode. FOr example, what is the value of the grid resistor on the triode? Let me say 470k for discussion. COUld have been 1 meg or whatever. If you connect the pentode plate to that grid with a cap, you get a very hot signal. If you added another 470k resistor in series with that cap, it would form a voltage divider with the grid and cut the signal level in half at the grid itself. Half sounds drastic, but it is 6db. 6db is the difference between the high and low jacks on a typical Fender amp. It might make the difference between stability and a lack of it.