Blackface Super Reverb PT issue

[Randall]

I have a 66 SR that has had it's PT replaced with a non-stock one with nothing stamped on it. It has the rectifier tube removed with diodes in its place, and the 5v winding leads cut back almost all the way. I can measure 5.5v on them as is, and should be able to splice them back. But it has no tap for the bias supply. Instead it has a wire run from one side of the secondary on rectifier pin 4 to the bias board. Then, instead of the usual 470 1 watt resistor, it has a 220KR with a 82K going between it and the diode to ground. With no power tubes in it, I get -48 to -54 volts at the bias pot center wiper.

I have never seen this done. Is this kosher?

https://www.thetubestore.com/lib/the...-Schematic.pdf

[The Dude]

Yes, it's kosher. It's a common method of deriving voltage for a bias supply if there is no separate bias supply winding.

Edit: Here's a similar example on a JTM45

http://marstran.com/66JTM45Schematic.gif

[Helmholtz]

With the JTM 45 bias circuit, the bias filter caps should be rated at least 350V. Fender circuit uses a 50V cap.

[g1]

With the JTM 45 bias circuit, the bias filter caps should be rated at least 350V.

Did you mean 150V ?

[Helmholtz]

Did you mean 150V ?

No, my JTM 50 uses 350V caps.
If for some reason the shunt resistor goes open, the caps will charge to high voltage and might explode if not sufficiently rated.

[The Dude]

Just for the record, I wasn't saying that the JTM bias circuit is something you should copy for this amp. I was using it only as an example of a bias supply derived from a HV winding- nothing specific as it relates to parts.

[Helmholtz]

What I meant is that the voltage rating of the bias cap(s) should be checked with the mod.

[Randall]

It has an 80/150v in there now. I often like 100/100v for this cap, any reason not to do so in this case?

[The Dude]

Edit: What Helmholtz is saying (I believe) is that when deriving bias from the HV winding, the caps should be rated higher in case the shunt resistor (82K in your case) goes open, so you should go with a higher voltage cap.

[Helmholtz]

It has an 80/150v in there now. I often like 100/100v for this cap, any reason not to do so in this case?

See post #5.
If the 56k shunt resistor goes high/open in the Marshall circuit, the cap voltage may rise to over 320V average.
In theory over 500V are possible at the first cap if there's no load.

A good design always should consider possible single component failures.

[The Dude]

No, my JTM 50 uses 350V caps.....

Wait,...... you have a JTM50?

[g1]

No, my JTM 50 uses 350V caps.
If for some reason the shunt resistor goes open, the caps will charge to high voltage and might explode if not sufficiently rated.

I don't recall ever seeing a 350V cap in a Marshall bias circuit (or Traynor), and most of them have bias derived from the HV winding. If they did so in the JTM50, I would argue it was because they had them on hand, rather than an effort to fail-safe the circuit. But that is pure speculation and you may well be correct.
Over-rating all components to anticipate circuit failure would be very impractical. Where would you draw the line?

[pdf64]

No, my JTM 50 uses 350V caps.
If for some reason the shunt resistor goes open, the caps will charge to high voltage and might explode if not sufficiently rated.

I don't think that sufficient current could flow via the 220k resistor to cause significant overheating / explosion.
Rather the leakage current through the ecaps would rise as the applied voltage rose above their working / forming voltage, and an equilibrium would be reached.
Perhaps akin to the original forming process by the manufacturer?

[Randall]

I settled on leaving the 150v Mallory cap that it came with. It tests good. All I have on hand is 450v, and that just seems a little much for a bias circuit.

[Helmholtz]

I don't think that sufficient current could flow via the 220k resistor to cause significant overheating / explosion.
Rather the leakage current through the ecaps would rise as the applied voltage rose above their working / forming voltage, and an equilibrium would be reached.
Perhaps akin to the original forming process by the manufacturer?

You're probably right, explosion is not likely.
But an extended "forming" process produces gas and is likely to damage the cap.