Bad power tube or oscillation?

[6267]

I am working on a BIG amp chassis, it's actually kind of stereo, in that there's a 6v6 power amp and a 6L6 power amp on the same chassis. Screens are all fed the same voltages, with a separate rectifier. The 6V6 section sounds great, but the 6L6 section seems to be giving me some problems. First off, I don't have any 6L6s to test with! I have been using some fat boy 6CA7s I have on hand (nothing on pin 1). I am using a 4-ohm on the 8-ohm tap, to better match the loading required by el34s. The cathode bias resistor is 250 ohm shared, the screen voltage is around 340, and the plate voltage around 370. It not only works on power up, but it sounds quite good. However, when gain is increased, a very high frequency oscillation begins, slowly reducing in frequency until it motorboats the amp. After twiddling with gain controls and removing feedback, the amp seems fine, sound is clear -- no apparent oscillation. After 5-10 minutes of music, the 60 hz hum starts increasing, inversely proportional to volume. After about 15 min, no signal gets through, and there is a loud 60hz. One of the 6CA7s is putting purple glow on its glass (it does this regardless of which tube which socket, and the glow follows the tube). No redplate. Is this a bad power tube, an oscillation too high to hear disabling one side of the p-p, a 60hz oscillation?@!??! or what? Gassy tube & runaway?

Ultimately: any guesses if this is an amp problem or a tube problem or both?

[Enzo]

What do you mean "nothing on pin 1?" 6CA7 must have pin 1 connected to pin 8 or otherwise terminated to ground or low volts. The 6L6 doesn;t care what you do with pin 1, but the 6CA7 does.

Pin 1 is the suppressor grid, and it won;t be happy with that grid floating.

[6267]

Hmmm...oops! I tamed some of the oscillation by shortening the grid circuit wiring, and there was something strange going on where one grid circuit resistor would be grounded and the other would go to the cathode of the preceding stage...some type of feedback, maybe. The inputs are RCAs, which make it a pain because they're not shorting jacks. Plugging & unplugging causes oscillations at very high, stable frequency on both the 6v6 and "6L6" outputs. With no input connected, it will sometimes hit this frequecy and oscillate...I think that's related to the entire lack of grid stoppers, so as soon as I get some carbon comp non-inductive resistors found for the power tubes and preamp grids, I'll be putting those in.

I didn't know I was running the 6CA7s in "tetrode"...thanks Enzo, luckily the bus bar is right near that pin...somehow I got the pin 1 thing reversed . I should really print some of these tube data sheets out, instead of running from room to room!

[Enzo]

Actually you can just strap it to pin 8 and it will be just fine.

[6267]

Cool...got it tied to cathode. Oscillations were gain related, and the offending source was a 16-ohm speaker lead that the original designers had shortened by cramming excess wire in the corner. By moving the input lead around the amp, I was able to detect the signal source...I moved it out of the way, not sure if I can just ground it (was floating).

The phase inverter is a really strange paraphrase, I think...first power tube's grid resistor (170k) doesn't go to ground but to the top of the grid resistor of one the inverter stage. I guess the grid resistors become the voltage divider, kinda nifty. Sounds ok, haven't been able to really overdrive it without worrying about my test speakers. No other connections between the two driver tubes, separate cathode bias resistors and all.

The strangest thing of all is the "Gain" control...it connects a pot wired as a variable resistor between the preamp output + 80k fixed resistance (it is the load for the last preamp tube) to the cathode of the second preamp stage. I guess this is gain control via variable feedback, but it seems ungainly. Doesn't seem to give a very good tone either...

Thanks for the help Enzo!

[6267]

Well, the oscillations are back, and I can't chopstick 'em. Sounds like the preamp has way way too much gain, starts hissing and howling. I put resistive loads on the transformers and grounded the input and low and behold, the oscillation is audible from vibrations in the transformer lamination (I guess?). The pre-amp is nothing special, kinda fenderish but with lotsa weird feedback loops...I removed these and tone opened up really nice. To clarify, this is a Conn organ chassis I am turning into what I hope will be a nice 25w/40w dissimilar stereo head. 10 octal sockets, 5 noval, one hell of a giant power transformer. 2.2k Rk at V1, not 1.?k, 100k plate resistors, 1 meg grid resistor...tone stack is bypassed, replaced with 150k load (any higher and you can hear this thing on the moon). 6v6 side has 2x 12ax7, "6L6" side has one 12ax7 and one 12au7. Both are funky paraphrase inverters, never seen these before. Installed grid stoppers at inputs & 6CA7s. 6v6s at 349 plate & screen, dissipation ~ 12 (sure, a bit high, they're beater tubes, and can take it). 6CA7s at 411 anode, 349 screen (shared supply...is this a possible culprit?). Dissipation @ 23 watts, they sound & look like they're doing well. Squeals worst with guitar volume at half, squeals anyway with input grounded, frequency & intensity often but not always respond to volume control. I *think* this is happening in the input valves, but there's no nasty lead dress or parallel wires etc. up there that I can see. I wonder if maybe it starts somewhere else and the high gain just starts amplifying it? Any recommendations?