Rocket red-plating on one phase leg...

[MysteryFever]

I'm stumped. As the subject reads, only one leg glows and it happens gradually over about 5 mins after B+ is switched on. It's not the coupling cap and it's not the grid-leak resistor, those have been checked /replaced and the problem persists. I have swapped tubes as well, only the tubes connected to the bad leg glow, regardless of which ones they are. I have one clue, which is that the grid voltage on the bad side gradually climbs from zero and the red-plating occurs at ~4-5 volts positive at idle. I feel like this should be a dead giveaway... but for what I don't know!

Drew

[Zippy]

Is this a new build or a new issue on a previously stable amp?

[MysteryFever]

It's a stable build but I recently changed out the power transformer (I was worried about the previous one's ability to supply adequate current since it used to be a two-bottle rocket). It worked fine for a short period after the swap but now this is happening. I do not suspect the PT but I can't explain the problem so I'm open to suggestions. I've double checked all PT connections and all voltages are nominal save for the climbing grid voltage.

[fishy]

Did you check the screen resistor......what value are you using?

Not a fan of 100 ohms - doesn't provide much protection and a rocket is biased hot already....... A bad screen resistor will plate one tube.

Can we assume voltages are comparable to previous?

[MysteryFever]

Value is 120r. I did check it when B+ was off and it was correct. Is it possible that those are drifting? I'll try to find some replacements.

[fishy]

Let us know how that goes...

I like 1k on the screens because the Rocket is biased so hot.

[fishy]

Let us know how that goes...

I like 1k on the screens because the Rocket is biased so hot.

[MysteryFever]

No luck, replaced 120s with 1ks and the same thing happens. Something interesting is happening now though... I lifted the wire bringing voltage to the screen on one tube on the bad leg. I ran the amp with no signal and monitored the voltage at the grids, of which there was none. Then I replaced the wire and ran it again, and the problem has shifted to the other leg... This has got to be a clue for somebody in the know! Right now I think it's possible that these tubes have a problem at their current bias and I'm going to bring up the value on the bias resistor to cool them off a bit.

Drew

[passfan]

Have you looked closely around the tube sockets and made sure that no solder or metal has fallen from the tranny swap?

[LarryLarry]

When you say "lifted the wire bringing voltage to the screen" what do you mean? Do you have wires running from your screen resistors to pin 9 on your EL84s?

[passfan]

Check bias cap on the output as well. What do you have in here for bias cap and res. ? Also the 220k's to ground and the cut circuit , everything running back to the coupling caps.

[MysteryFever]

Larry,
I mean I disconnected it and your assumption is correct. The resistors are board mounted with short wires running to their pins on the sockets.

passfan,
I used what I had on hand for the cathode resistor. It was 2 100r 5w in parallel, but in my attempt to cool the bias a bit, I replaced one with a 150r, bringing the effective cathode resistance up to 60r. The capacitor is a 35v rated 220uf electrolytic. Also, the physical condition of the socket was the first thing I checked.

[passfan]

Should be fine thewre. Some of us raise the resistor as high as 75 or 80 ohms to cool the output down , no harm.