Is this an oscillation problem? -broskie CCA linestage

[spopepro]

Reading through as many threads as I can find through search, I think I have an oscillation problem with my linestage. It's Broskie's CCA design, using his board, with all values as indicated in the manual. I am using 6n23p tubes, and used the 6dj8 specs with the lower B+ and extra capacitance.

So the problem is sometimes a low hum, sometimes a buzz, in one channel, but not always, and not always the same channel. It doesn't follow tubes when I switch them, it doesn't go away when I short inputs, it does not change with adjusting the volume. And sometimes it doesn't happen at all. I was listening today for 5 hours with no problems, turned the power off, came back 2 hrs later and had a hum problem. Less frequent, but also observed is a significant (10+db) difference in the two channels. When I shut the power off the volume on both channels increases and is hum/buzz free until the caps fully discharge and then there is, of course, silence.

So from what I have read the 6dj8 family tubes can be susceptible to oscillation. Since nothing is happening regularly, maybe this is what is going on? I don't have a scope and wouldn't know the first thing to do with it if I had one. I'm currently running it into a cheap T-amp with a volume pot on the front of it (not nearly ideal, but it's my backup for when I'm working on other things, and right now have no other functional poweramps) and suspect that maybe the input impedance with the pot where it is might be causing problems? I guess I'm hoping someone can help me out with if I might be on the right track, and what I might be able to do to stop the "good sound" lottery. Thanks!

[DF96]

What have you done with pin 9, the internal screen? This should be grounded. The manual says a capacitor is OK, but this does not establish the DC voltage and some valves may do strange things if the screen is left floating. If you will only ever use valves with a pin 9 screen, then put a wire link instead of a capacitor. Otherwise, put a high value resistor in parallel with the cap - 1M should be fine.

This might not solve your problem, but it does address a shortcoming of this design.

[pedroskova]

You could try upping your grid resistors to 1k from 300R. Is your wiring from inputs to switch shielded? Do you have any digital equipment on your AC circuit ... cable TV ... light dimmers? Is your PCB in a metal enclosure or on a breadboard?

[spopepro]

Thanks for the help--I really appreciate it.

DF96- You know, I remember trying to decide if I wanted to use caps or a jumper for pin 9 to ground, and it turns out I did neither. Jumper is now in place. Unfortunately, no change.

Pedroskova- Amp is in a case with input leads properly grounded. I don't know how the house is wired, but I ran an extension cord to a couple of different outlets and didn't have any luck solving the problem. I could up the grid resistor values, but I'd prefer to do this only if I'm pretty sure it's an oscillation problem, which I'm not... yet.

[wa2ise]

Grid stoppers may help. That's what the CCA's R2, R7 and R10 appear to be for. I had a strange hum problem with a 12AX7 driver stage that was cured by placing a 68K resistor right at the grid terminal at the tube socket, the other end of that resistor connects to the coupling cap and resistor to ground that previously was connected to the grid terminal of the tube socket. The 68K value is not critical, but that's what many guitar amp folks often use. The 300 ohms in the CCA design may be too low a value, try 2.7K, and if that doesn't do it, 27K. This grid stopper resistor, combined with the Miller capacitance of the 12AX7, forms a low pass filter at around 100KHz, high enough to not affect audio, but enough to suppress oscillation.
Example grid stopper as used in a tube AM radio: . This driver is a high gain circuit, lower gain circuits should require lower value resistors.