What would cause a cathode bias-set resistor to blow?

[chuckrock]

Hey Guys, I'm currently using a 10 ohm 1W resistor from the cathode (pin 8) to ground on each of the power tubes in my Twin Reverb for reading the bias current, and recently one of them drifted all the way up to 1.5M.

Plate voltage in this amp is just under 440VDC, and I biased the tubes at 70% max dissipation at 48mA per tube.

I replaced the resistor and everything seems to be fine, but I was wondering if any of you guys had any thoughts as to what caused this?

Thanks!

 

[Wally]

Chuck, I was under the impression that a resistor in that spot for that purpose should be a 1 ohm resistor?????

As to why one 'went away', feces occurrum.

 

[chuckrock]

You can use a 10 ohm resistor instead of a 1 ohm, you just carry the decimal point an extra space when reading your DC voltage.

 

[printer2]

Could have just been a bad part, it is not like you are dropping a lot of power across it.

 

[Wally]

Okay, so puzzle me this...still learnin' --I am trying.

Why would we want to place more resistance from this cathode to ground than is necessary to make this measurement? What effect if any does the 10ohms make versus the 1 ohm in the measurement? And....at what point of resistance would we enter into a partial cathode-biasing circuit?

 

[ThermionicScott]

Why would we want to place more resistance from this cathode to ground than is necessary to make this measurement? What effect if any does the 10ohms make versus the 1 ohm in the measurement? And....at what point of resistance would we enter into a partial cathode-biasing circuit?

Helps with the accuracy of the reading, for one thing. A reading in the hundredths of volts in the 2V range is more accurate than a reading in the thousandths of volts. (I use 10-ohm resistors myself.)

There's no reason a 10-ohm/1W resistor should have blown -- my vote is for a bad part.

- Scott

 

[Wally]

Thanks for the explanation, SCott. I don't use this method, but it is good to know in case I ever want to do it. Now the question....why have I always read/heard to use a 1 ohm resistor, and why do the manufacturers use
1 ohm instead of a 10 ohm? Is it the simplicity of the 1 to 1 relationship of the millivolts to the milliamps?
With the 10 ohm resistor, would the OP's reading have been 4.8mv???
.048V???

 

[printer2]

Thanks for the explanation, SCott. I don't use this method, but it is good to know in case I ever want to do it. Now the question....why have I always read/heard to use a 1 ohm resistor, and why do the manufacturers use
1 ohm instead of a 10 ohm? Is it the simplicity of the 1 to 1 relationship of the millivolts to the milliamps?
With the 10 ohm resistor, would the OP's reading have been 4.8mv???
.048V???

1 ohm is easier for people, they can not complain to manufacturers if they are out by a factor of 10.

 

[ThermionicScott]

Thanks for the explanation, SCott. I don't use this method, but it is good to know in case I ever want to do it. Now the question....why have I always read/heard to use a 1 ohm resistor, and why do the manufacturers use
1 ohm instead of a 10 ohm? Is it the simplicity of the 1 to 1 relationship of the millivolts to the milliamps?
With the 10 ohm resistor, would the OP's reading have been 4.8mv???
.048V???

I'm sure it's this. You would read 48mA as 0.048V (or 48mV). The reading on a 10-ohm resistor would be 0.48V (or 480mV).

- Scott

 

[Wally]

Thanks, Scott. I was in a circle yesterday, wasn't I? If I had been dropping bombs, I might have stunned someone...but no direct hit with that decimal point, eh? LOL

 

[celeste]

........ And....at what point of resistance would we enter into a partial cathode-biasing circuit?

Any resistance in the cathode circuit results in "partial cathode biasing". the question is where does it start to make a difference. I build amps like this and call it Hybrid Bias. As for where it starts to make a difference, I would say if you are not adjusting plate voltage to maintain Vak, then when the resistor generated bias is somewhere around 10% of total bias voltage, it will start to make itself known. If you maintain Vak, you can use a lot more

 

[Wally]

thanks for that insight, celeste. I"ll see if I can digest that.
Fender did something like this in the '68 circuits, didn't they? They used a 150ohm resistor coming off of each cathode....along with the bias voltage balancing act. That didn't last long, did it?

 

[celeste]

thanks for that insight, celeste. I"ll see if I can digest that.
Fender did something like this in the '68 circuits, didn't they? They used a 150ohm resistor coming off of each cathode....along with the bias voltage balancing act. That didn't last long, did it?

Always hard to tell what Fender was thinking at the time, it could be simply they were trying to burn up some voltage because the B+ was higher then they wanted with the PT's they had in stock, next order of PT's they adjusted the voltage and did away with the 150 ohm resistor. I am not saying that is what happened, just suggesting how it MIGHT have come to be.

I do it because I use parallel output tubes and I want each to be biased very differently. One is deep class AB, and just standard fixed bias, the other is deep class A, it clips before cut off. Because the share a common OT, to control dissipation in the A biased tube by burning up voltage in the 5k cathode resistor. Because both are DC coupled to cathode followers so I can drive into sub2 operation I have to add fixed bias to the cathode biased tube as well.

 

[ThermionicScott]

Always hard to tell what Fender was thinking at the time, it could be simply they were trying to burn up some voltage because the B+ was higher then they wanted with the PT's they had in stock, next order of PT's they adjusted the voltage and did away with the 150 ohm resistor. I am not saying that is what happened, just suggesting how it MIGHT have come to be.

My theory has been that the bias tap on a run of 2x6L6 PT's didn't offer enough voltage, so the cathode resistors were a cheaper/faster way to do it than throwing out the PT's and waiting for more. On the couple of schematics I've seen that had the measurement, there was +10V at the top of the 150-ohm resistors, indicating 67mA. At 460-10=450 plate volts, that means each 6L6GC was still idling at 30 watts!

I think there is a lot of promise in "mixed-bias" designs -- the partial cathode bias lets each tube adjust for its current draw, reducing the need for tubes to be "matched". And it doesn't rob as much output power as going all the way to full cathode bias! I think CBS/Fender missed a good marketing opportunity there.

- Scott

 

[celeste]

My theory has been that the bias tap on a run of 2x6L6 PT's didn't offer enough voltage, so the cathode resistors were a cheaper/faster way to do it than throwing out the PT's and waiting for more. On the couple of schematics I've seen that had the measurement, there was +10V at the top of the 150-ohm resistors, indicating 67mA. At 460-10=450 plate volts, that means each 6L6GC was still idling at 30 watts!

I think there is a lot of promise in "mixed-bias" designs -- the partial cathode bias lets each tube adjust for its current draw, reducing the need for tubes to be "matched". And it doesn't rob as much output power as going all the way to full cathode bias! I think CBS/Fender missed a good marketing opportunity there.

- Scott

Another very plausible explanation. The problem with mixed bias from a production standpoint is it offers the worst of both worlds. All the parts of both combine and all the required adjustments to get it out the door

 

[Wally]

FWIW, SCott, the Twin Reverb AC568 circuit and the Showman/Dual Showman circuits from the summer of '68 use the same approach....for whatever reason.

 

[ThermionicScott]

Another very plausible explanation. The problem with mixed bias from a production standpoint is it offers the worst of both worlds. All the parts of both combine and all the required adjustments to get it out the door

Yeah, it's easy to see why any company trying to wring maximal profit from their parts bin isn't going to bother with it.

FWIW, SCott, the Twin Reverb AC568 circuit and the Showman/Dual Showman circuits from the summer of '68 use the same approach....for whatever reason.

Y'know, I looked for 4x6L6 '568 schematics on the usual "Field Guide" sites, and didn't see them there, but you're right. So that rules out bad PTs. Perhaps it was that the first-gen bias balance circuit robbed too much voltage, and that's why the resistors were needed? Once I come up with a theory I like, it's hard to let it go.

- Scott

 

[tubeswell]

heat stress can cause resistor values to drift