Bias keeps changing on Hiwatt

[frank57]

This is sort of related to the Hiwatt custom 20 problems I was having in another thread here,
but the problem I find is the bias keeps changing when I measure it so I thought I'd post it separately.
I seem to be getting a different bias and voltage on this hiwatt Custom 20 korean amp.
I had it set at 379 19ma, but a few days later it's at 371 16ma.
Before that I had it at 369 20 ma but it shot up to 384 the next day or so.
This tech put the bias pot on the outside  glued on the chassis near the power tubes and covered with some kind of black material, but I have to lock it so to speak with loctite glue.
But is this normal? If so how are you supposed to bias it?
I don't remember the amp changing its sound so drastically.
Hot one day cold the next.

[RicharD]

You've listed a total fluctuation of 4mA.  What's your wall voltage doing?  Here at my house I see changes from 117 to 127.  Did you let the tubes warm up completely (like greater than 5 minutes) each time you measured bias?

[loogie]

I did a Marshall 1959 recently and I had to wait ten, fifteen minutes before the thing stabilized.  Even then I probably didn't wait long enough.

[frank57]

Well at least 15-20 minutes today.
I did it twice turning the amp off briefly but it was the same thing.
Maybe I should check it tomorrow for longer?
I didn't touch the pot at all and it's locked pretty good. You really have to twist it to make it move.
When it went to 384 the other day, I could not get that voltage down to 369 where I had set it at one point.
No way no how. You would have to get in the 90 percent of the plate dissipation to even get there.
So I set it at 379 19 ma about 60 %.
Then today it started at 371 16ma and stayed there.
So far it's swung from 10 to 20,23 percent.
I don't know if the bias probe is part of the problem or the location of the bias pot perhaps.
I thought the amp sounded biased hot the other day and sure enough that's what
 the amp head bias probe readings indicated.
How would I check the voltage in the wall?
Does your voltage and bias more or less stay the same?

[HotBluePlates]

How would I check the voltage in the wall?

You could make a "suicide plug", but that's dangerous, so we won't mention it again.

The wall voltage has to come into the amp at the cord or IEC receptacle (which your amp probably has). Measure from hot to neutral where the cord enters the amp.

Does your voltage and bias more or less stay the same?

Not really, and it doesn't really matter. Wall voltage drifts up and down depending on season, time of day and local usage/demand, among other things. Further, tubes drift. The plate current with a given bias voltage can and will meander up and down depending on the tube sample.

It is less obvious with cathode biasing, because the cathode resistor times the changing current produces a voltage drop (which is equal to tube bias in 99% of cases) that offsets the effect of the current change. You don't get the automatic compensation in a fixed-biased amp. That is, unless you use some kind of servo-bias arrangement, but the added complexity doesn't give much in return in a guitar amp.

I can't say that I've ever cared about the small drifts you cite. If the tube drifted from 10mA to 100mA, that would be a problem and you'd have to wonder if there is a problem in the bias circuit. In your case, the bias circuit could be rock-solid and it could just be a matter of simple tube drift.