First Post - Bias voltage dropping until red-plating - follows the tube

[ThermionicEmissions]

Hello everyone, this is my first post to the MEF. I'm not an electronics tech, but I'm slowly learning more about tube amps as I find it all very fascinating.
What brings me here today, is my Mesa Boogie Express 5:25, which has been giving me fits due to some intermittent red-plating of one of the two EL84 output tubes. The red-plating occurs (at least, when it occurs) after playing for some time.
Today I took the chassis out, and used my looper pedal to continually play through the amp while I monitored the bias voltage on pin 3. On both tubes the bias voltage started around -12v, but started dropping slowly. After about 15 minutes, one of the tube's bias stabilized at around -11.4v, but the other kept slowly and steadily dropping, until after 30 minutes or so, it was down to -5.8v and the plate was starting to glow orange - at that point I shut 'er down.
A couple of hours later, I reversed the positions of the tubes, repeated the test, and saw the same results with the same tube (not the same position, the red-plating occurred on the same tube).
Again, after a couple of hours, I reversed the positions of the tubes again, and repeated the test. Again, I saw the bias voltage start dropping again, however this time, it went down to about -9.5v, and then climbed back to -10.5v and seemed to stabilize (at least for another 10 minutes).
Anybody have any thoughts as to what would cause a tube to behave as such?

Thanks all!

[Leo_Gnardo]

Anybody have any thoughts as to what would cause a tube to behave as such?

Yes. The tube is rubbish. Throw it out.

[ThermionicEmissions]

Oh for sure. I'm just curious as to what could condition in the tube could cause it to pull down the bias voltage as such. And, FWIW, old/bad tubes make great Xmas tree decorations!

[Chuck H]

Oh for sure. I'm just curious as to what could condition in the tube could cause it to pull down the bias voltage as such. And, FWIW, old/bad tubes make great Xmas tree decorations!

Why? You can't go into the envelope and fix the tube anyway! Tubes simply aren't made to the same stringencies as they use to be and often exhibit odd behavior that is not easily explained relative to their function. But, rather, relative to their construction affecting function. Which has proven to be difficult at times. Rather than trying to "understand" what is wrong with the tube (since it's almost surely a construction flaw that no one can do anything about) I think it's easier to just accept that $h!t happens and replace the tube/s. Now the question is whether you'll want to replace with a matched set or just the one tube.?.

[Justin Thomas]

And, FWIW, old/bad tubes make great Xmas tree decorations!

One, I'dimit this to BAD tubes, because OLD tubes can be perfectly fine. Two, it's all good til the cat knocks that 6L6GC off the tree, and the bulb shatters, sending all that cathode coating into the air & my lungs...
I'm not saying don't do it, just be careful!

Justin

[Mick Bailey]

Oh for sure. I'm just curious as to what could condition in the tube could cause it to pull down the bias voltage as such. And, FWIW, old/bad tubes make great Xmas tree decorations!

One condition could be the grid connection within the tube going high-resistance and causing loss of bias voltage. I would expect the sound to be affected, but if you're playing at low volume you may not notice so much until it reaches a critical level (by which time the tube will be red-plating). Tubes rely on a number of construction techniques - spot-welding, slot-and-tab, rolled seams, crimps. Generally the electrical connections are made using spot-welds and these can sometimes fail. I have carried out post-mortems on tubes and have occasionally found a hard fault, but usually there's nothing visually detectable.

[ThermionicEmissions]

Thank-you, Mick, this is the kind of information I was looking for.

[ThermionicEmissions]

"I love the smell of cathode coating on the morning....smells like...victory"
jk, thanks for the warning

[Leo_Gnardo]

"I love the smell of cathode coating on the morning....smells like...victory"
jk, thanks for the warning

A tube full of air might make a poor substitute for a snow globe too. It's best to avoid inhaling any of the white flaky barium oxide coating inside the glass. The doctors would go crazy looking at your x-rays.

[ThermionicEmissions]

Why? You can't go into the envelope and fix the tube anyway! Tubes simply aren't made to the same stringencies as they use to be and often exhibit odd behavior that is not easily explained relative to their function. But, rather, relative to their construction affecting function. Which has proven to be difficult at times. Rather than trying to "understand" what is wrong with the tube (since it's almost surely a construction flaw that no one can do anything about) I think it's easier to just accept that $h!t happens and replace the tube/s. Now the question is whether you'll want to replace with a matched set or just the one tube.?.

To clarify, my motivation for understand the "why" is to better understand how everything works as a system.
I already got a "matched" set of Mesa tubes to replace this pair. The quotes are because when I put them in and measured the plate current draw (using a bias probe), one was drawing ~35ma, and the other ~28ma. Rotated them and checked again to make sure it's the tubes. It is. I know there's lots written about this, but ~7ma difference doesn't seem all that well matched to me.

[ThermionicEmissions]

A tube full of air might make a poor substitute for a snow globe too. It's best to avoid inhaling any of the white flaky barium oxide coating inside the glass. The doctors would go crazy looking at your x-rays.

But...but...they're so pretty


Besides, I don't live in the state of California, so it's all good! /s

[Chuck H]

I burn them in, match for idle current and then crank things up into clipping to check for any gross differences in current. There usually isn't a significant difference, but sometimes there is. So, is it better to match for idle current or for when the tubes are actually amplifying a signal? This is a real stretch, but perhaps Mesa is matching tubes in a way not specific to idle current, when the tube isn't really doing anything but cooking at it's ready state. Maybe if you test the tubes at close to their max dissipation you'll find that they are closer. Maybe not.

[ThermionicEmissions]

I burn them in, match for idle current and then crank things up into clipping to check for any gross differences in current. There usually isn't a significant difference, but sometimes there is. So, is it better to match for idle current or for when the tubes are actually amplifying a signal? This is a real stretch, but perhaps Mesa is matching tubes in a way not specific to idle current, when the tube isn't really doing anything but cooking at it's ready state. Maybe if you test the tubes at close to their max dissipation you'll find that they are closer. Maybe not.

I put an email in to Mesa about it to see what they say. Their amps are cathode-biased, so there's no way to adjust the idle.
Edit: no "simple" way to adjust the idle

[Chuck H]

But you can plug in one tube at a time and put a signal into the amp. You wouldn't want to drive things for too long like this, but it's pretty safe and it would allow you test each tube for current under operating conditions.

[ThermionicEmissions]

But you can plug in one tube at a time and put a signal into the amp. You wouldn't want to drive things for too long like this, but it's pretty safe and it would allow you test each tube for current under operating conditions.

Not following as to why I would want to do this with only one tube at a time. Wouldn't both tubes in be a better representation of operating conditions?