Biasing combination bias

[EFK]

If an amp uses some kind of combination bias - let's say it has a typical fixed bias setup like a bassman or jtm45 - but in addition there are cathode resistors (or a shared resistor) between the cathodes and ground, how exactly does one bias it?

Is it as simple as 'normal' cathode bias? i.e., voltage drop divided by resistor value = current in mA?

[jrfrond]

You don't see combination bias too much, and on a personal note, I can't see any practical use for it. Perhaps someone out there can shed some light on this, because quite frankly, I've never understood the reasoning behind it, and since you only see it once in awhile, I am guessing I am not alone.

Fender did combination bias on some later SF models, and in every instance, it caused weird sine wave distortion upon testing, causing the middle of the arc to form a sharper peak, like the top of a triangle wave. Increase in odd-harmonic distortion perhaps? Never analyzed it, but rather pulled the cathode resistors out, tied the cathodes straight to ground and rebiased normally to solve the issue (which was also Fender's tech bulletin at the time).

[EFK]

Hmmm. I had read a couple of old threads somewhere where it was praised to some extent as splitting the difference in 'feel' and/or harmonics between cathode bias and fixed bias. Was just thinking about trying it out (already have fixed supply) but not sure how to measure bias if I remove those little 1 ohm resistors currently in place.

[mooreamps]

Hmmm. I had read a couple of old threads somewhere where it was praised to some extent as splitting the difference in 'feel' and/or harmonics between cathode bias and fixed bias. Was just thinking about trying it out (already have fixed supply) but not sure how to measure bias if I remove those little 1 ohm resistors currently in place.

Bias is measured between the control grid with respect to the cathode.

-g

[jmaf]

on a personal note, I can't see any practical use for it.

At most, you can use it for some tone shaping on the power stage if you bypass it with a capacitor of your choice. Without the bypass cap you can add a bit of compression / neg. feedback to the power stage if you have a resistor there.

Did not know that Fender had done it and the distortion it caused, interesting bit there.

[tubeswell]

If an amp uses some kind of combination bias - ..., how exactly does one bias it?

Is it as simple as 'normal' cathode bias? i.e., voltage drop divided by resistor value = current in mA?

Yes - the current through the cathode resistor is the tube current (and you have to multiply this by the P-K voltage to get tube dissipation etc). The effect of the extra -ve bias at the grid is just to lower the tube current some more in this case.

[Chuck H]

Bias is measured between the control grid with respect to the cathode.

-g

I'm not getting this. How does the grid/cathode relationship indicate the bias current? I understand that this is the bias voltage, but that tells us nothing about current through the tube.

BTW Gary, I need to know if there are any beef dishes or recipes your especially fond of.

[jmaf]

Yes - the current through the cathode resistor is the tube current (and you have to multiply this by the P-K voltage to get tube dissipation etc).

Noting also that if the cathode is shared, you gotta divide by the number of tubes.

[jrfrond]

At most, you can use it for some tone shaping on the power stage if you bypass it with a capacitor of your choice. Without the bypass cap you can add a bit of compression / neg. feedback to the power stage if you have a resistor there.

Did not know that Fender had done it and the distortion it caused, interesting bit there.

1968 chassis. Here's a schematic example: http://www.webphix.com/schematic%20h...b568_schem.pdf

OK, so you are saying that you can balance between the cathode and fixed bias feels. Makes sense, but from what I've seen, on the Fender examples anyway, it does introduce some odd distortion. Has anyone analyzed this at all?

[Enzo]

Chuck, Gary didn't say that was how to measure bias current, he said the definition of bias. Bias is the voltage difference betwen control grid and cathode. 50v bias is 50v bias, regardless of how much current flows. We generally set the tube bias to result in a certain amount of idle current through the tube, and by extension the idle dissipation of the tube. Calling it "bias current" is really a verbal shortcut.

And that really is the point. The tube has no idea in the world how the bias got there. Ground the grid, and get +50v across a cathode resistor, or ground the cathode and put a -50v supply on the grid, all the same to the tube. Either way, the grid is 50v more negative than the cathode. 50v of bias.

The plate voltage itself is not enough, we need to know the voltage across the tube. In the case of cathode bias, the cathode voltage is subtracted from the plate voltage. In fixed bias, the cathode is grounded, so the plate voltage is the voltage across the tube.

Lest someone point it out, if you convert fixed bias to cathode bias, the bias voltage is then subtracted from the plate voltage. SO the same circuit results in a lower voltage across the tube for cathode bias than fixed.

Either way voltage across the tube times current through it yields dissipation.

WIth combination bias, two methods are used to bias the tube. Nonetheless, the result is still some amount of voltage difference between grid and cathode. Want to adjust it? Measure the voltage across the tube and the current through it. A bias probe still works, because all it does is insert a current monitor in series with the cathode. It doesn't care if that current is also generating a bias voltage or not.

You can also just measure voltage across the cathode resistance and calculate current, and as was mentioned by jmaf, if the resistor is shared, divide the current by the number of tubes. All that matters is the voltage across the tube and current through it. If you check the voltage across a cathode resistor, it has no idea at all whether you also put negative bias voltage on the grid to help control the current. All it knows is how much current there is.


All the above discussion is ignoring the small contribution screen current makes to the cathode current.

[jmaf]

1968 chassis. Here's a schematic example: http://www.webphix.com/schematic%20h...b568_schem.pdf

OK, so you are saying that you can balance between the cathode and fixed bias feels. Makes sense, but from what I've seen, on the Fender examples anyway, it does introduce some odd distortion. Has anyone analyzed this at all?

Thanks for the schematic link. Yeah, I've experimented with adding a small resistor to the cathode on a fixed bias amp. I recall that I didn't make much of the result, but I didn't do any serious study of it, just ear tests.

[Chuck H]

Chuck, Gary didn't say that was how to measure bias current, he said the definition of bias. Bias is the voltage difference betwen control grid and cathode. 50v bias is 50v bias, regardless of how much current flows. We generally set the tube bias to result in a certain amount of idle current through the tube, and by extension the idle dissipation of the tube. Calling it "bias current" is really a verbal shortcut.

Sure. Right. The important numbers indicate current. I was just pointing out the relevance of the the statement as it applies to the thread. Gary can be a little obtuse sometimes.

FWIW, having worked a lot with el84's in the last couple of years, I never neglect screen current.

I would like to know more about Gary's favorite beef dishes or recipes though. I'm trying to develope a recipe for "Gary's beef with Merlin" and I'd like to be honest about my reference.

[mooreamps]

I use combination bias in my 36 watt amp, and I'll rather be boiled in oil than to teach "merlin" how I do it.

-g

[pdf64]

Vox used combination biasing on the AC50 (all but the very earliest anyway), see http://www.webphix.com/schematic%20h...mps/ac50_1.pdf and they sound pretty darn good. Unbypassed individual 47 ohm cathode resistors.
Similarly for the Sound City 50 plus MkIV http://www.webphix.com/schematic%20h...city50plus.pdf though their tonal merits are more controversial.

[jrfrond]

I use combination bias in my 36 watt amp, and I'll rather be boiled in oil than to teach "merlin" how I do it.

-g

I sure do hope that was a tongue-in-cheek response.

No one here is interested in anyone else's "proprietary magic". Perhaps you can tell us WHY you use combination bias.