Sovtek 5881/ 6L6WGC drift

[LarBal]

One of my main gigging amps that I built years ago into an old SF Bassman chassis has an interesting issue I'm kinda stumped on.

The amp is fixed biased, and itself perfectly fine(I least I think it is based on the hundreds of gigs I've played on it).

2 years ago, I tried some Sovtek 5881, the wafer base that most people seem to hate (Russian 6n3c-E), and they sounded fantastic in the amp. It's basically a vibroverb w 15" speaker in a combo.
After about a year of sonic bliss, and probably 50 gigs, I noticed something changed with the tone. It had been sounding especially good, then it seemed to lose a little power...something wasn't quite right.
I had originally biased them around 30-32mA at the normal 450VDC ish.
When I yanked them, they read 45mA . I thought that was odd, as usually dying tubes seem to go the otherway.

Anyways, I put another set of the sovteks in, biased them up to 30-32mA, and went along with life, and have continued to use the amp regularly. Side note, I think it must be the combination of the 15" speaker with these tubes as to the magic combo.
Together everything is very warm , and kinda crunchy, in a great way, even at lower volumes.
So it's been a year, and I plugged in today and the amp seemed to have lost a little power. I put a read on the tubes and same thing, drifted up to 42-43ma...both of them.

Any insight into what's going on here? Is this something that tends to happen that I have somehow just missed all these years?? Should I crank the bias back down and go on with life? Leave it? Put in new ones?
I just ordered another set, and I think I will check the bias weekly this time!

[The Dude]

Bias alone won't effect max power output. Bias is an idle condition. Tube wear is another story all together. Was line voltage the same for both bias measurements? How often do you use the amp? Did you let the tubes warm sufficiently for the initial bias adjustment.

[Helmholtz]

Is the bias supply stable?

[Mick Bailey]

New tubes require burn-in time to stabilize the bias and a lot of tubes just get a matching test and nothing more, so when fitted can initially drift before settling down. A 10mA increase would be unusual, though. I'm with The Dude - bias doesn't affect power output (unless there's a serious fault causing massive current draw). Within a normal range from very cold to hot there's no significant difference when scoping the output.

[mozz]

Sounds like the tubes are going leaky.

[Helmholtz]

I understand the same thing happened with 2 different sets of tubes.

When I yanked them, they read 45mA .

What exactly do you mean here?

[LarBal]

Sorry, didn't mean to ghost this thread,...busy week...
The Dude Yea line voltage is steady, I always run new tunes for 30 min then re-adjust, amp's probably got 100hrs on this set of 5881s
Helmholtz I assume it is, but that could very well be an issue. I'm putting in a new set in a minute, I'll check it . When I pulled the current set, they were pulling 45mA each at idle. I'm positive I set the bias around 32mA when I buttoned it up.

The amp has definitely lost high end, bass, and all it's character. Of course this is common for failing tubes, but I usually experience this with HRDs that are 15 years old with the original tubes, or whatever.

I'm not that concerned with it, other than the fact that this amps has sounded fantastic the last 6 months or so, and I just don't know where the bias was during that point! Somewhere between 32 and 45 ha.
But if these tubes are in the habit of drifting, I certainly don't want to start a set of what I'll still assuming is a ~23W tube in the 40s, when it will end up in the 50s+.

I'm going to install these at ~32, and monitor them every month, till I'm back to this place again

[mozz]

Bad coupling caps?

[g1]

I was wondering that too, but if the power tubes are staying fairly matched, I don't think the coupling caps could do that?

[Helmholtz]

I was wondering that too, but if the power tubes are staying fairly matched, I don't think the coupling caps could do that?

Only if both coupling caps show the same (increase of) leakage.

Leaky coupling caps should show in increased (less negative) grid voltage.
Best test is disconnecting them to see if this drops the idle current(s).

[LarBal]

I didn't suspect the coupling caps to the outputs bc I let the amp run for a while both times before I buttoned up, so I guess I assumed if they were leaky it would have happened quicker.

I just ran this thing for 90minutes after installing new Sovtek 5881/WGC (NOS 2013), and the bias more or less stayed steady where I set it, 35mA, the whole time

It sounds amazing again, and all the grunt is back. I built it in May 2021, and while I have played probably 100 gigs with it, this is the 3rd set of tubes.

I had used one of those tiny pihers for the bias pot, it turned rather easily, so I replaced that with a more robust trimmer. Maybe all the rockin knocked it outta place??
Nonetheless, even at 45mA that's not crazy hot for these ~26 watt (not 23 as I prev stated) tubes, so the short life is still a bit of a mystery. They were definitely on their way out, if tone can be an indicating factor.

It's my favorite amp for anything besides small rooms, so I'm reluctant to start changing anything