Adjustable cathode bias?

whoaru99

Epic Member
In my forthcoming build cathode bias is the plan. However, I'm wondering if it might be a good idea to make the cathode bias adjustable?

560 ohms is the planned cathode resistor, but what about changing that to, say, a 500 ohm resistor and 100 ohm pot in series? That would put the pot at roughly midpoint.

Thinking to wire the potentiometer wiper in shunt configuration rather than use it as a rheostat.

I plan to use matched pairs of KT88s, but wouldn't this scheme facilitate final dial-in on the amp?

Drawbacks (couple extra parts and labor notwithstanding)?
 
you can do it that way, just needs to be a stout high wattage pot. If you're going with a single pot it won't allow you to dial in the tubes, unless the plan is one pot per tube. That would work fine.

The early Williamson amps worked sort-of similar but the pot was used between the cathodes. basically the cathodes went to either end of the pot, and the wiper went to the cathode resistor. The idea was you used the pot to set equal current draw but you could not actually adjust how much current it pulled.
 
Since the bias target is 85mA, across 100 ohms that's only about 3/4 watt.

3W to 5W pots are not too expensive so that's sorta where I was thinking. 3W, 100 ohm pot in theory handles 170mA, right?

And, if the pot ends up at 60 ohms, then at 85mA, it's dissipating about 1/2 watt.

Does that pass the red-face test?
 
Last edited:
class A or AB? AB draws a lot more current under load than idle.

Sounds right otherwise

Normally I think of pots being rated for 1/4 to 1/2 watt, so 3-5 watts would be "high wattage" for me.
 
In my forthcoming build cathode bias is the plan. However, I'm wondering if it might be a good idea to make the cathode bias adjustable?

560 ohms is the planned cathode resistor, but what about changing that to, say, a 500 ohm resistor and 100 ohm pot in series? That would put the pot at roughly midpoint.

Thinking to wire the potentiometer wiper in shunt configuration rather than use it as a rheostat.

I plan to use matched pairs of KT88s, but wouldn't this scheme facilitate final dial-in on the amp?

Drawbacks (couple extra parts and labor notwithstanding)?
Hi, 99 it seems that if you are going to do that you should consider using fixed bias. Removing the cathode resistor and more importantly the bypass capacitor from the signal chain is an audible improvement. Plus, that will also give you more power which is not insignificant since you are running single ended. The hardest part is sourcing the trans for the bias voltage if your PT doesnt have a bias tap. But all the parts are small enough to fit under the chassis.
 
The power transformer voltage and all that is sized to account for cathode bias and the operating point. I have all the parts already ordered and paid for, except now maybe $15 worth of potentiometers and different cathode resistors if I decide to go that way

Maybe something to keep in my back pocket for future change.
 
Last edited:
In my forthcoming build cathode bias is the plan. However, I'm wondering if it might be a good idea to make the cathode bias adjustable?

560 ohms is the planned cathode resistor, but what about changing that to, say, a 500 ohm resistor and 100 ohm pot in series? That would put the pot at roughly midpoint.

Thinking to wire the potentiometer wiper in shunt configuration rather than use it as a rheostat.

I plan to use matched pairs of KT88s, but wouldn't this scheme facilitate final dial-in on the amp?

Drawbacks (couple extra parts and labor notwithstanding)?

I did this in my SE 6L6 amp. Schematic attached for reference.
 

Attachments

  • ST-12-SE Mk I schematics.pdf
    46.9 KB · Views: 84
Should work fine. Harmon Kardon did it too, but they used something like a 25 ohm pot on their A500 (pot was used for DC balance between two output tubes, but same idea). I worry about pots if they heat up at all. Could have a tendency to compromise the wiper contact. I'd use the smallest pot you can get by with, to give you the bias adjust range you want. I'd probably derate pot heat dissipation by a factor of 6 to 8, just to be on the safe side, considering some heat from the output tubes will seep in under the chassis where the pot is installed.
 
I would use only a wire wound pot , a carbon track pot will degrad with DC on it , and become intermittent .
 
Since the bias target is 85mA, across 100 ohms that's only about 3/4 watt.

3W to 5W pots are not too expensive so that's sorta where I was thinking. 3W, 100 ohm pot in theory handles 170mA, right?

Yes (in theory),... and NO (in real world). 170 mA through a 100 ohms resistor is equivalent to a 3W dissipation but the max. wiper current of the potentiometer is usually much less and this spec is not allways given by the manufacturer. I would not use a potentiometer rated for less than 5W in this position, and good quality wirewound pots aren't cheap.
 
Yes (in theory),... and NO (in real world). 170 mA through a 100 ohms resistor is equivalent to a 3W dissipation but the max. wiper current of the potentiometer is usually much less and this spec is not allways given by the manufacturer. I would not use a potentiometer rated for less than 5W in this position, and good quality wirewound pots aren't cheap.

That is how the max current is determined, by the full resistance and wattage.

That means 170mA can flow though any portion of the potentiometer but the max permissible dissipation changes with wiper position.

If the 3W, 100 ohm pot is set to 10 ohms then max permissible dissipation can be only 289mW; not 3W.
 
PartsExpress carries a 100 ohm, 5 watt, wirewound pot for $4, which seems perfect for this (this is probably the exact reason they carry it). It can otherwise be hard and very expensive to find one rated for more than 1/2w.

On a smaller single-ended build I use a 200 ohm pot to tweak cathode bias, which I happened to have lying around, as far as I know it's a standard 1/2 watt carbon pot.

I've encountered no trouble, though at 30-45ma it's within the pot's ratings. The highest current is near the maximum current rating for any portion of the pot, however it doesn't present much resistance and doesn't dissipate much at those higher-current settings.

Multiple KT88s would make me nervous about using a standard pot however!
 
That is how the max current is determined, by the full resistance and wattage.

That means 170mA can flow though any portion of the potentiometer but the max permissible dissipation changes with wiper position.

If the 3W, 100 ohm pot is set to 10 ohms then max permissible dissipation can be only 289mW; not 3W.

This is only true for high quality potentiometers from reputable manufacturers (Sfernice, Vishay,...). In most cheap wirewound pots the max. current through the wiper is the limiting factor and is rarely specified (guess why ?). Also, you must keep in mind that the max. power ratings of the potentiometer is usually given at an ambient t° of 25°C which is allmost allways exceeded inside most tube amps chassis: thus you MUST use the derating curves or reliability will be severely compromised. Remember all the DC current of your output stage will flow through this tiny pot wiper point contact and undersizing critical power dissipating parts has never been considered good engineering practice.
For all the above reasons I would not use a pot rated below 5W dissipation in this application but of course you're free to make cheap compromises if quality and long term reliability are not an issue for you.
 
won't be enough to bother anything, and if its bypassed with a cap its effectively not there at all.
 
won't be enough to bother anything, and if its bypassed with a cap its effectively not there at all.
Sounds good, nice straight answer too. I remember searching it before and decided not to after seeing mixed answers. I can tell you've actually tried it in circuits though. I have this 2 watt 10 turn pots from china, pretty nice for a few bucks.
 
won't be enough to bother anything, and if its bypassed with a cap its effectively not there at all.

That is both true and not necessarily true, and here's why.

The inductance of a wirewound resistor of a few hundred Ω in value is a handful of uH, very slightly more or less, depending upon the technology used. (Wire-wound resistor vs. rheostat, number of turns, etc.) In the audio passband (< 20 kHz) this translates into a reactance of a few tenths of an Ω, which is << 0.1%, i.e. negligible. At a few MHz, however, the reactance becomes tens of Ω, which is ~ 10%, i.e. significant.

But the amplifier is running audio, one might protest, so it's all LF and insignificant. Which is mostly the case. But there's a lot of HF floating around, particularly if one has solid-state rectifiers dumping Qrr every cycle (about 450 kHz) or sometimes oscillation at a few MHz because of a lack of stopper resistors or other self-resonance issues. So a wirewound resistor can easily contribute to HF oscillation in an audio amplifier, depending upon the design. A well-designed amplifier should not have HF components and should not oscillate; the definition of what constitutes a "well-designed amplifier" is left as an exercise for the reader, with the caveat that unless one has carefully analyzed the circuit and thought about all the possibilities it may not be as well designed as one might think.

I learned a long time ago that the component does not, per se, lie to us, it is we who lie to ourselves about what the component actually does in a circuit. That lie, in turn, is based upon our delusions or oft-repeated dogma about what the circuit does, instead of knowing what it actually does because of verification using instruments able to show us. But even those instruments can be deceptive, if we aren't looking at the right timescale.

Understanding the actual functioning of circuits is what Bob Pease, Jim Williams, Dan Sheingold, etc. tried to convey.
 
Last edited:
well, I'm figuring that if it has a bypass cap the cap will have low reactance at frequencies where the wirewound has high reactance so unless things just get really bad it ought not matter to any real degree. The C should negate the L. If this is not bypassed with the usual electrolytic cap, a small value cap that is invisible at audio frequencies will serve nicely to take care of RF nonsense. Maybe for some double insurance even with an electrolytic cap in the circuit bypassing it with a small value film cap wouldn't be the worst idea.

Of course stability testing is just a good idea anyway to prove that its going to behave itself.
 
Back
Top Bottom