Magnatone MP1 trouble

[pontiacpete]

This amp has a cathode resistor that is heating up and smoking after 30 seconds to a minute. It has an unusual bias circuit. There is suppose to be -17vdc on the grids and 4.5vdc on the cathodes with a 100 ohm resistor. The voltages look good on the plates and screens. They are higher than the schematic at 380 and 380vdc. The voltages are pretty close in the phase inverter, plate is a bit higher at 250vdc and grid is about 75v and cathode 130.
As the amp warms up the cathode voltages start to rise and the smoke comes from the 100R resistor.
I have new tubes in place 7199s, I tried another OT. Change bias caps, 100uf and 22uf, resistors are good 22K and 12K. Also replaced the .047uf on the grids of 7199s
Does anyone have an idea as what to do?
the schematic is below. The power supply is not the same as the drawing but I just put together the same arrangement for the first 2 nodes. Schematic is a photo of the one attached to inside of the cab.

[Helmholtz]

Does the bias voltage at point E significantly change during warm up?

Also check with power tubes pulled.

[pontiacpete]

With the tubes pulled the bias voltage stays -17.20. With the tubes in and as it warms and the cathode voltages rise up to 20v, it drops to -16

[Helmholtz]

Could be a conductive power tube socket.

[pontiacpete]

Forgot to mention I put in new power tube sockets

[Helmholtz]

I saw you replaced the coupling caps but nevertheless disconnect one end of each 0.047ĩ coupling cap and check again.

[pontiacpete]

With the .047s disconnected the voltages still rise on the cathodes

[Helmholtz]

So we can also eliminate oscillation and a bad coupling cap.
Running out of ideas.
I would not completely eliminate a bad new tube, though.
Did you replace with 7189A , not 7189?

Please post pics of the socket wiring.

[pontiacpete]

I’ve tried, at first, a new pair of JJs el84s and then just got these new tung sol 7189s. I guess I could try the 84s again.

Still rising cathode voltage, even faster with the 84s

Oh well, thanks for helping. If you think of anything let me know!

[Helmholtz]

Schematic calls for 7189A, which can be used with a socket wiring different from an 7189 or EL84.

https://frank.pocnet.net/sheets/106/7/7189.pdf

[The Dude]

Tweaked schematic:

[J M Fahey]

This amp has a cathode resistor that is heating up and smoking after 30 seconds to a minute. It has an unusual bias circuit. There is suppose to be -17vdc on the grids and 4.5vdc on the cathodes with a 100 ohm resistor. The voltages look good on the plates and screens. They are higher than the schematic at 380 and 380vdc. The voltages are pretty close in the phase inverter, plate is a bit higher at 250vdc and grid is about 75v and cathode 130.
As the amp warms up the cathode voltages start to rise and the smoke comes from the 100R resistor.
I have new tubes in place 7199s, I tried another OT. Change bias caps, 100uf and 22uf, resistors are good 22K and 12K. Also replaced the .047uf on the grids of 7199s
Does anyone have an idea as what to do?
the schematic is below. The power supply is not the same as the drawing but I just put together the same arrangement for the first 2 nodes. Schematic is a photo of the one attached to inside of the cab.

Hybrid biasing always struck me as a particulaly bad idea, itīs either one or the other.

Suggest you increase negative bias (by acting on the 22k/12k divider) until you read expected +4.5V at cathodes.

Nothing "sacred/immovable" about old circuits, I much suspect nothing is "broken" but current tubes are quite different to original ones, nothing further.
Adaptt to the "new normal"

[The Dude]

What is bias current? That's where I'd start. You may need to up the cathode resistor value or modify the bias supply to get tubes running in acceptable range.

Edit: Just saw your post saying cathode voltage is 20V. That makes 100mV per tube idle current. No wonder things are getting hot.

[Helmholtz]

Edit: Just saw your post saying cathode voltage is 20V. That makes 100mV per tube idle current. No wonder things are getting hot.

We can't know if it's 100mA/100mA or 10mA/190mA. At least one of the tubes should be severely redplating.
Even more astonishing that this happens together with the bias voltage being 15V more negative than normal.

Does the 100R resistor still measure good?

[pontiacpete]

Some really good observations and suggestions. Thanks. Thanks for the schematic makeover.
THis schematic does not correspond with what I have been seeing in the power supply so I think I have the power supply wired like the schematic. Probably not a huge deal.
Anyway I did try grounding the cathode and raising the value of the resistor that is parallel to the 22uf cap to 22K from 12k. The tubes red plated a lot. I will look at this differently tomorrow, and see what happens.
It’s the weirdest thing. I hope my tubes aren’t shot. They only took one red plating episode, but mostly I’ve been turning it off before they glow and going by my meter.