Bad power tube, filter cap, resistor?

[joulupukki]

In my newly-built 18w TMB I cranked it up last Saturday to near max and played for a while (with ear muffs). It was incredible. But ... at some point I must have done something bad. Now when I take it off of standby I get a few seconds of terrible noise. I'm pretty sure it's not the preamp tubes because I swapped those out for the ones I have in my micro amp and it still has the same problem.

What does this sound like?

 

[corliss1]

Try, in this order:

-pulling the preamp tubes, one at a time. Not swapping them, just removing them completely. See if the issue stops after any of that. If so, that's the stage with the problem.

If not....

-pull the power tubes and see if the issue stops. It likely will.

Then...

-replace power tubes with new ones and see what happens.

 

[Lowerleftcoast]

When it calms down can you tap the tubes and get microphonic similar noises?

It might tell you which tube is wonky.

And look at the tubes in the dark when it makes the odd noise. See anything?

 

[joulupukki]

Good thinking guys. Thank you for those tips for quickly confirming what I thought might be going on. No noise with the power tubes out. With the lights dimmed and the tubes swapped, the one in V4 was red plating. Switching it back to V5 it still red plated there too. But the other one sounds a little loose when I wiggle it next to my ear as well.

 

[joulupukki]

Looking back on my voltage measurements...

Is it normal to have that much difference in resistance between the two sides of the output transformer? Everything else seems pretty close. Seems like it'd be nice to have the plate dissipation for V4 be close to 100% like I've got going in V5. Could this have been why one (or both) of the EL84 tubes failed so soon?

Also, to fix the disparity between the two sides of the OT would you add a resistor in series with the OT wire going to V4 to bring it closer to 339.7R that V5 has? Or would it be better to switch the cathode resistor from 130R to something higher?

 

[joulupukki]

So this tube on the right red plates whether it’s in V4 or V5. :-|

 

[tubejockey]

I would bias it a bit colder, or make an adjustable bias for each tube.

 

[corliss1]

Is it normal to have that much difference in resistance between the two sides of the output transformer?

Yes, that's only a few percent, really, when you're dealing with numbers in the hundreds. Remember, it's a guitar amp, not some specialized scientific device. Close Counts.

 

[King Fan]

Remind us of your bias setup, bias resistor value, which tubes you’re running? The fact the redplating moves with the tube tells us that tube is either wonky or just sensitive to bias. And remember to recheck voltage drop when you move or swap out the tubes — the OT resistance is fixed of course, but also of course it’s not the problem here, or the redplating would stay with the socket.

Did I already link to this in your other thread? EL84s can be sensitive little guys.

Biasing El84's in 18 watt TMB - 18Watt.com

www.18watt.com

 

[Wally]

It would be helpful if it was known if the redplating tube is the high dissipation tube, wouldn’t it? That is a 13% increase in plate dissipation over the tube that was working at 100%, which is where one would like the circuit to be, imho. Those two tubes were not a really good match, imho. If they were a good, stable match; then there is an imbalance 8n the circuit somehow.
If these tubes were bought new for this application, did you buy burned-in tubes? The advantage of pay8ng for burning in is that tubes‘ operational parameters can drift in the first hours of use. This drift can undo the match if that match that was made without a burn-in period.
This situation supports the thought that anytime an amp changes sound…for the better or for the worse….one should immediately take a look at the tubes…while the amp is processing signal…immediately in other words.

 

[joulupukki]

Remind us of your bias setup, bias resistor value, which tubes you’re running? The fact the redplating moves with the tube tells us that tube is either wonky or just sensitive to bias. And remember to recheck voltage drop when you move or swap out the tubes — the OT resistance is fixed of course, but also of course it’s not the problem here, or the redplating would stay with the socket.

I’m using the JJ EL84 tubes that came with the Mojotone 18w TMB kit. No idea if they’d been burned in but I highly doubt it.

I think that tube just didn’t like when I cranked it up and something went awry.

Did I already link to this in your other thread? EL84s can be sensitive little guys.

Biasing El84's in 18 watt TMB - 18Watt.com

www.18watt.com

I don’t think I’d seen that before. Thanks for sharing. Good info in there.

It would be helpful if it was known if the redplating tube is the high dissipation tube, wouldn’t it? That is a 13% increase in plate dissipation over the tube that was working at 100%, which is where one would like the circuit to be, imho. Those two tubes were not a really good match, imho. If they were a good, stable match; then there is an imbalance 8n the circuit somehow.

The red-plating tube is the one in V5 which, as far as I can tell before the issue started, was running at 100% plate dissipation.

If these tubes were bought new for this application, did you buy burned-in tubes? The advantage of pay8ng for burning in is that tubes‘ operational parameters can drift in the first hours of use. This drift can undo the match if that match that was made without a burn-in period.

They were part of the Mojotone 18w TMB kit (doubt they were burned in tubes).

This situation supports the thought that anytime an amp changes sound…for the better or for the worse….one should immediately take a look at the tubes…while the amp is processing signal…immediately in other words.

Thanks for that tip – I’ll definitely do that from now on.

 

[Wally]

IF the tube that has redplating was the one that was drawing less current, that would be an interesting situation. Things happen, though.

 

[Ten Over]

The current through the screen resistors is 45.2mA according to your measurements. This is not very likely given that only 11.9mA is going through the 2.7k dropping resistor. Try measuring the voltage across each 100R resistor.

What do you measure across the cathode resistor?

 

[joulupukki]

The current through the screen resistors is 45.2mA according to your measurements. This is not very likely given that only 11.9mA is going through the 2.7k dropping resistor. Try measuring the voltage across each 100R resistor.

What do you measure across the cathode resistor?

I'm assuming I need to take these measurements with a good EL84 tube? I need to check with a local shop to see if they have a pair of them.

 

[Ten Over]

I'm assuming I need to take these measurements with a good EL84 tube?

No. I'm interested in the voltages with the tubes that you have.

 

[King Fan]

It'd be interesting to know with the two tubes you've got.

EDIT: Oops, gotta type faster.

 

[Lowerleftcoast]

Is it normal to have that much difference in resistance between the two sides of the output transformer?... ...Could this have been why one (or both) of the EL84 tubes failed so soon?

It depends on how the OT is wound. Many OT's are not designed to have identical windings. This difference is of little concern and you do not want to add resistance on one leg to even it out. I doubt this is responsible for the failure. If there is a concern, separate cathode resistors can be used so the tubes do not share a resistor.

 

[joulupukki]

The current through the screen resistors is 45.2mA according to your measurements. This is not very likely given that only 11.9mA is going through the 2.7k dropping resistor. Try measuring the voltage across each 100R resistor.

What do you measure across the cathode resistor?

Well, it's definitely not looking correct.

13.2 V across the 130R cathode resistor (seems normal?).

Powered off measuring voltage (in-circuit) of the V4 & V5 100R screen resistors gives me: 464R and 465R respectively (same as when I first had it built).

V4:
Voltage across the 100R screen resistor: 0.49 V
Plate voltage (pin 7): 325.7 V
Screen voltage (pin 9): 293.1 V

V5:
Voltage across the 100R screen resistor: 4.02 V
Plate voltage (pin 7): 303.2
Screen voltage (pin 9): 289.4

 

[2L man]

Now looking those numbers there is way too high voltage loss on other half of OT primary. Perhaps tube (anode) socket is bad? Or solder? Or tube pin has corrosion?

 

[peteb]

V4:
Voltage across the 100R screen resistor: 0.49 V
Plate voltage (pin 7): 325.7 V
Screen voltage (pin 9): 293.1 V

V5:
Voltage across the 100R screen resistor: 4.02 V
Plate voltage (pin 7): 303.2
Screen voltage (pin 9): 289.4

4.02 V looks problematic.

0.49 / 100 is 5 mA

4.02 / 100 is 40 mA

no bueno