Need Help with New Amp Build, Ugly signals

[kevinkr]

How is lead dress in the area of the first stage - nice tight short leads? Try adding a 1K resistor between pin 7 and the 0.47uF capacitor, right at pin 7. I think you'll probably find this kills the oscillation.

[LinuksGuru]

SansyG, I may be wrong in your particular case, but I had similar issue with similar amplifier. The real culprit was output transformer. Probably it had some arching between adjacent turns of winding (enamel coating is defect or scratched).
After I replaced transformer, everything went OK.

Do you have 2nd output transformer?

[LinuksGuru]

Originally Posted by kevinkr

How is lead dress in the area of the first stage - nice tight short leads? Try adding a 1K resistor between pin 7 and the 0.47uF capacitor, right at pin 7. I think you'll probably find this kills the oscillation.

You mean V6B 12BH7A grid and C16? Kevin, what is purpose of this resistor?

[SandyG]

Ok, tried a 200pf cap across C15, did nothing.

Tried the 9.4Meg resistor across C15 (Pin 7 to ground). Signal is pretty clean. So something clear that is odd with the input stage.

I wonder if R20 at 1Meg is too high a value, but I really don't know the theory on how this circuit works.

Kevin - The leads and other parts are pretty close on the circuit board, but could be something like that. And for your test was that to place the 1K resistor before Pin 7 and the 1M and 0.47uf cap?

LinuksGuru - Don't think it's an output transformer as the signal is clean when I play a bit with the input circuit as mentioned. The output transformer is new, and I think I would see some odd intermittent signals (and hear it for sure).

Sandy

[SandyG]

Originally Posted by LinuksGuru

You mean V6B 12BH7A grid and C16? Kevin, what is purpose of this resistor?

No C15, on V5B Pin 7

Sandy

[DF96]

No, I think he means on the first stage. I was going to suggest this, but kevinkr beat me to it. It could well be that the original caps were sufficiently lossy at RF that they didn't need stoppers.

As a general rule, if a circuit is behaving strangely and touching it stops it or changes the behaviour then there is a good chance the stage you touched is oscillating at RF.

PS putting a resistor to ground will changes the bias of the top triode and probably reduce its gain, so killing the oscillation. It might also reduce the gain of the bottom half, so that could be the culprit too.

[kevinkr]

Originally Posted by SandyG

<snip>
Kevin - The leads and other parts are pretty close on the circuit board, but could be something like that. And for your test was that to place the 1K resistor before Pin 7 and the 1M and 0.47uf cap?

<snip>

Yes, just lift the connections at pin 7 and solder a 1K resistor from pin 7 to the lifted connections. (1M and 0.47uF cap) - this can be a permanent fix if it works. Since this is a pcb perhaps you can just cut the etch near pin 7.

Note that the 1M resistor provides grid leak bias for the upper triode in the cascode so there may not be a huge amount of room for tinkering with the value.

Sandy[/QUOTE]

[SandyG]

Originally Posted by SandyG

I wonder if the old time paper capacitor that was used in the 50's was loss-y enough resistance wise to keep things stable?

Hey toss the new kid a bone Well I guess if I keep guessing I would have got the right answer eventually!

I'll give that a shot, might have to hack up a bit to try it, nothing a dremel and small drill can't fix!

Back to the lab... well the garage and take it apart again.

[SandyG]

Originally Posted by DF96

PS putting a resistor to ground will changes the bias of the top triode and probably reduce its gain, so killing the oscillation. It might also reduce the gain of the bottom half, so that could be the culprit too.

The output clipped a bit earlier with the 9.4Meg resistor across the C15. The RMS voltmeter showed about 1 volt lower before clipping without the resistor.

This could also be the voltmeter getting a cleaner waveform that is causing a lower reading but hard to say.

Sandy

[SandyG]

Back from the Lab

Adding in the 1k resistor before Pin 7 of the 12AU7A had no effect that I can see. Signal was bad at 100Hz and 1kHz as in the original with the odd oscillations.

Sandy