Bias control opinion question

[mlp-mx6]

One control for both sides of the output, or one for each side (or each tube)?

Submitted for your consideration, here is a snip from a poster on the Hoffman board named PRR. This guy often gives what I consider to be excellent advice to the newbies on that board. He is the first person I've ever seen make a logical case for a single bias control. What do you think about what he says?
---------------------------------------------------------------------------------
(beginning of PRR snip)

Balancing the idle or no-signal currents balances the amp for no-signal.

Balancing the idle currents may balance the small-signal gains, though if you need un-equal voltages to get equal currents, then it is unlikely the gains could be equal.

Hi-Fi amps normally use "same type and lot" tubes, and rarely get much outside the small-signal region. For this use, matching idle current may match gains.

You don't play a guitar amp at no-signal, or small-signal. You play at medium to BIG signal. Approaching or exceeding the maximum undistorted output.

For nearly all audio amplifiers, we aim for maximum peak output at zero grid volts. So the maximum peak input signal is equal to the bias voltage minus zero volts, or just the bias voltage.

So if you have to bias one tube to -35V, and the other tube to -45V, to get no-signal "balance", the side with -35V bias will overload at 35V peak input, the side with -45V bias will overload at 45V peak input. They are NOT balanced up at the power level that guitar amps are normally played at. It's about 28% out of balance.

Yes, you can also offset the driver outputs to give 35V on one side when the other was doing 45V. That is another adjustment to make, and it can only be made at high power (or approximately with measurements and math), not on stage.

If you have four tubes it gets really messy.

We can see it different ways. I would strongly argue for ONE bias pot. Set all two/four tubes to the same grid voltage. That balances their large-signal overload. Check that each tube has "acceptable" cathode current. Not equal, but not over the dissipation limit and not much-much less than Ipk/4, where Ipk is the peak current per tube at maximum output.

If a tube doesn't bias "acceptably" with the same grid voltage as the others, you can't get "acceptable" balance. It will either be unbalanced for small-signals or unbalanced for large signals.

(end of PRR snip)
----------------------------------------------------------------------

On the other side, many (notably Kevin O'Connor) favor individual controls for each side, or even for each tube in a 4-output-tube setup. KOC seems to deem it very important to balance the idle currents.

[mlp-mx6]

Anyone?

[MarkB]

I don't have any expert knowledge to share, but I have some thoughts. I'm not sure how much of a concern his "large signal balance" is. Fender silverface amps had bias balance controls, and I don't know of anyone saying they sounded bad because of it. People change them out to bias level controls but not because the balance control is bad, but because the level control is seen as more important. In any case, tubes that are matched for current draw aren't necessarily matched for transconductance, so the signal produced at the plate can be out of balance even with the same bias voltage on both tubes.

The good news is that tubes are very "generous". You can run them under some funky conditions and they still sound good.

[Bob-I]

Well... I don't claim to be an expert but here's my thinking.....

Most output xformers I've tested are balanced on AC but not on DC. When a DC voltage is applied, as in during idle, you get unequal current flow. When you apply an AC voltage on top of this imbalamce you get an imbalance in the AC ouput because you started at an unequal current. By equalizing the current at idle you start at the same point on the load line on both sides so as you move up you're balanced.

When going to 4 tubes you'd still want 2 bias pots because what you're balancing is the output xformer not the tubes.

Someone jump in and correct me if I'm way off base here.