Blues Jr III Universal issue

[lowell]

Guys, the last 3 blues JR III's I've had in have the same odd issue happening. When I try to measure plate volts on one side (blue) of the OT I get audible hum, and the voltage is quite a bit higher (20vDC) than the other side of the OT. The other side is 335Vdc and doesn't cause a hum when the probe touches it. This is with a brand new pair of working tubes. Both grids have -14vDC on em. Both screens are good. Doesn't matter whether I use chassis or main filter ground as common. So everything SEEMS to be fine, amp works, hum is minimal, seems fine. This only seems to happen when I'm repairing Blues Jr III amps. Even if I don't connect the black lead on my meter to anything, just touching the red probe to the plate of that side of the OT causes an audible hum, and even occasionally a very very small sparking if I watch closely.

What could this be? I doubt that the plate volts are really off that much when my meter probe is not touching the TP because the hum goes away when I stop measuring it. (double-slit experiment lol - in another universe this isn't happening)
.
Instead of measuring to common, I take the "good" plate Vdc measurement, then measure from there to the other plate, and only 3mv difference so I know it's all good.

* Originally it came in with a blown Rscreen 100ohm but I've repaired that.

I've also found on multiple occasions that the two EL84 don't seem to bias correctly (not matched) even with matched sets of tubes, and that some physical rerouting of some wires can solve the bias issue. Not sure if it's the gray ribbon cables next to the OT primary leads or what, but some finagling usually fixes that. No issue with the bias in this particular one.

Any theories?

Thanks!

[Enzo]

Your meter is reacting to the inductance of teh transformer. Pretty normal for many amps.

[Helmholtz]

HF oscillation caused by parasitic effects of the meter leads.
You may verify by pulling the PI tube or by shorting the speaker output during the idle DC measurement (no signal!). This should prevent oscillation.

[nickb]

HF oscillation caused by parasitic effects of the meter leads.
You may verify by pulling the PI tube or by shorting the speaker output during the DC measurement. This should prevent oscillation.

^^^^^ This is what is going on. Seen it many times on an FBJ (and other amps).

[Pedro Vecino]

The solution that I use is to install both tubes, measure the tube on the good side, invert them and measure again on that same side the other tube. Itīs better to forget the left socket.

[lowell]

BRILLIANT... thanks guys! I knew it was something more obscure. Perfect solution, if pulling the PI works.

Question though, since global feedback is NEGATIVE feedback, how is this being injected back into the PI and causing an issue. I'd think the issue would null itself.

[Helmholtz]

Question though, since global feedback is NEGATIVE feedback, how is this being injected back into the PI and causing an issue. I'd think the issue would null itself.

The two halves of the OT primary are 180 degrees out of phase with respect to AC ground. So one side may produce negative feedback with the meter leads and no oscillation while the other gives positive FB and oscillation.

[lowell]

Ahhh… right. Hence the disaster that breaks loose when you connect primaries in reverse when replacing an OT.

[drewl]

Yup, been there.
If you re-do the wiring, twist the pairs real nice and push the ribbon cables back towards the chassis you won't get the HF oscillation and can measure the plate voltage/bias as normal.

[lowell]

Twist the OT primaries?

[drewl]

As much as possible and position them away from other PT leads.

[Chuck H]

This is such an obscure piece of very useful information for most DIYers (including me) and even some experienced techs that I chose to "sticky" it for a time to keep it visible. Right on guys!

[J M Fahey]

Agree that directly measuring plate voltages can often lead to unwanted and dangerous oscillation.

For a long time now I have shied away from directly measuring plate volts ... lost a couple cherished digital multimeters in that task, clearly amp oscillates and 1500V peak HF voltage smashes my meter input to death.

Usually measure centertap which is +B, stable and well decoupled so no oscillation at all, and trust plate voltage is a few V below that.

I really do not need to exactly know whether plate voltage is 453 or 449V, itīs close enough (and much safer) to know itīs "a few V less that centertap 460V" or whatever you have there.

And if I ever needed to really know it, in the extreme case I suspect an open primary or something, I pull an old needle meter and go at it.

[Dave H]

To measure bias current I put a shorting plug in the speaker jack and use the transformer shunt method. It doesn't oscillate when it's shorted.

If it oscillates when measuring plate voltage try using a "stopper" resistor. Clip a 10k to 100k resistor to the end of the meter lead and probe with that (and short the speaker)

[lowell]

Have 3 Blues Jr III's in the shop from a local church. All have that same obscure symptom. All have the same repair as well - open Rscreen. Pulling the PI cures the oscillation/measurement issue! Thanks guys! Awesome thread thanks for the input.

JM, do you think after pulling the PI and mitigating the oscillation, that it still may kill your meter?