Howling Feedback in Reveb Recovery

[jewishjay]

Hi everyone, I'm working on this 1979 Silverface Super Reverb. Everything works but when you turn the reverb past 4 it rings and howls like feedback. It's not a microphonic tube, I chopsticked, and swapped tubes. No change. It's not an oscillation, I tried a 10pf cap from plate to cathode. No change. It only stops howling if you pull out V4. If I put a 12AU7 in V4 you can go all the way to 8 before howl...so I figure the reverb recovery stage just has too much gain?

I'm thinking of seperating the cathodes on V4, maybe using 1.5k and 25uf for the mix stage and 2.7k unbypassed for the verb recovery. What do you think of this plan? Or is there another way to attack this problem?

[PRR]

Did you try a fresh 25uFd at V4 shared cathode?

Yes, separate cathode networks is a better plan.

[jewishjay]

Did you try a fresh 25uFd at V4 shared cathode?

No, I can try that first. Thanks!

[PRR]

For background: Fender has a grounded-cathode stage into a grounded-cathode stage, with both cathodes tied together. This is Positive feedback. If there is also gain, it is an oscillator. The first triode has good gain, the second cathode is unity gain at most. If it is grounded (perhaps with a large capacitor) it has no gain; but if that cap craps-out, it wants to howl. Here the 0.003uFd interstage blocks bass howl (motorboat) but allows >500Hz to go through. The Reverb (return) pot lets you trim for no oscillator gain (and no reverb!), or enough gain to howl. The change from 12AX7 to 12AU7 confirms the basic diagnosis: you have to turn the pot up further for the low-gain tube. (And you probably have essentially the same reverb-before-howl either way.)

With the big cap healthy, it works. Leo surely knew this would be trouble "someday", but the e-caps he was getting were good for 5 or 10 (or more!) years so who cared? If the amp were still in service in 1967 (transistors looked to obsolete tubes), then a $2 cap repair was a fix; or Leo (or by that time CBS) would be pleased to sell a new amp.

[jewishjay]

Yes, separate cathode networks is a better plan.

In the end, that's what I did. Removing or replacing the 25uf didn't solve it, but seperating the cathodes did. 1.5k & 25uf for the mix stage and 1.5k unbypassed for the reverb. Everything is happy and the world is at peace.

Many thanks for the help!

[GroundhogKen]

I learned this lesson earlier this year.   I tried reducing the bypass cap to 2.2uF on the V4 shared cathode of my Deluxe Reverb.   It began to  howl with the reverb knob above 5. 

It took me a few hours before realized I had introduced a feedback path.   
Separate cathodes is the best solution, otherwise the shared bypass cap has to be 22uF or larger.
Ken