What is best way to have individual bias on 4 power tube amp?

[plumcrazyfx]

I've seen plenty of schems for individual adjustable fixed bias in 2 tube amps but not for 4.  Would it be adjustable for each side and then balance for the tubes pairs?

[PRR]

Usually simplest to just have one pot per tube.

[silverfox]

One Pot per tube
When I tried that I could never get the bias to stabilize. Attempting to adjust it would also cause all the other tube bias values to drift all over, sometimes resulting in a momentary Red Plate. I do believe the right design would work but I couldn't figure that one out. Basically, my design was: A bias voltage source with 4 outputs supported by a final filter cap on each output.
Maybe it needed a regulator, but then you get too clean of a bias source. Could also have been related to the wrong size dropping resistor as I never got the power supply calculation done right.
silverfox.

[plumcrazyfx]

So, I need some advice.  Can this work - the pot is for global bias and the trimmers are for fine tuning.

[PRR]

Ohms Law.

The 6L6 grids suck "ZERO" current.

0 to 33K, passing zero current, is zero volts dropped.

(If you try this with a mter, it _will_ drop; only from meter-loading, thus only while you are poking at it.)

What you propose will change the high-frequency roll-off, and the overload dynamics, but not the idle bias.

[jjasilli]

Ditto to one pot per tube & Ohm's Law. IOW, to drop bias voltage you need to bleed some voltage to ground = voltage divider.  A simple series resistor doesn't attenuate bias voltage, because there's no current flow to cause a voltage drop. 

Also, re your 1st pot.  Best practice is to always have a fixed resistor between the bias pot > ground.  If bias voltage is dialed to -0-, or too low, or the pot fails, the tubes will fry quickly due to loss of adequate bias.  The value of the fixed resistor should be chosen so that bare minimum bias voltage is always supplied to the tubes, without redplating or worse.

And 10K pots are easier to fine tune than 50K pots.

[plumcrazyfx]

Okay thanks.  So the 4 would look something like the global bias with a diver to ground on each tube?

[sluckey]

Might look like this but X4 instead of X2...

http://sluckeyamps.com/sunn/sunn_sceptre_1971.pdf

[2deaf]

Maybe start with a basic architecture like this?

[plumcrazyfx]

Those help.  Thanks.

[92Volts]

Most bias circuits present a changing resistance with adjustment, which messes up many shared negative voltage supplies. That's especially true with more robust circuits where a pot is used to shunt voltage to ground, so the failure condition (pot loses contact) is more negative voltage, instead of lost contact/bias.

2deaf's circuit mostly uses pots as traditional voltage dividers which wouldn't result in a changing resistance with adjustment. The only change "visible" from the other tubes is through the 2.2M failsafe resistors.

Another way to make each supply fully independent is to duplicate the upper-left resistors (and/or diode and caps depending how you approach it) for every one of the tubes. This gets annoying if you need to burn off a lot of voltage. This might be easier and more fault tolerant if you have a fixed negative supply maybe 2x larger than the required bias voltage. Not every transformer has a separate bias tap, but if yours does, I'd take advantage of that.

[plumcrazyfx]

Is this better?  Looked at Slucky's and compared it to a Twin Reverb bias arrangement.  I think I figured out the bias arrangement for 4 tubes but you keep having to add parts for the bias and the PI, too.  Getting a huge board for something that most here said was kind finicky to adjust and that may not be noticeable.  Went with 4 tubes but 2 bias adjustments.  You guys are all so helpful and helps me learn.  I have to go back a reread the electronics correspondence course books I got at a garage sale.

[92Volts]

That looks OK and I think it's similar to some Marshalls.

The "risk" is if the wiper of a pot loses contact, you have zero contact with the bias circuit and the tubes could go hot as a result. A large-value resistor to the negative end of the pot (or an even more negative voltage source) is best so the failure condition is cold bias.

Anybody who's dealt with old equipment where the pots need cleaning before they'll pass sound knows they do fail this way, though I've never encountered a new pot with any sort of flaw. Whether this design is acceptable depends whether you plan to mod/rewire this in the near future or leave it like that "forever".

[plumcrazyfx]

Is there an amp schematic that has the example for this kind of bias?  I'm looking at the one from the Fender Dual Showman and I basically see what I have?  I like doing something with a failsafe.

[sluckey]

What you have drawn will work just fine. If you want to add the failsafe resistors just connect a 2.2M resistor between each pot wiper and the top side of the pot (referring to your drawing). Please note, those resistor values work just fine in the Sunn circuit you copied and will give a proper adjustment range for 6550s and that Sunn PT. If you adapt this circuit to a different PT and/or use different power tubes you will need to change resistor values to give a proper adjustment range for your particular tubes.

I just looked at all the Dual Showman schematics and none of them have dual bias pots. The AB763 has a single bias adjust pot. All the later models have a single bias balance pot.

[2deaf]

This one?

[sluckey]

Oops. Missed that one.

[2deaf]

I own one of those.  Otherwise I would have missed it, too.

[2deaf]

I've seen plenty of schems for individual adjustable fixed bias in 2 tube amps but not for 4.  Would it be adjustable for each side and then balance for the tubes pairs?

Is this going into any particular amp?  Do you know the AC voltage going into the circuit and the bias voltage needed?

[plumcrazyfx]

This is a conversion of an Acoustic 164 to an Ampeg B-15 upped to 4 6L6s changing the 8 pins to 9 pins.  It's a complete gut job only retaining the Schumacher iron.

[2deaf]

Suggestions

[plumcrazyfx]

Thanks.  I'll add the 2.2m resistors and tweak the values.

[sluckey]

I'd tweak the resistors to give an adjustment range of -45v to -60v for 6L6s.

[plumcrazyfx]

Thanks everyone.  I think I'm comfortable with the bias design.  I usually go about 60% on my bias.  I always think in voltage and forget current.  Probably due to coming from the effects side where current is negligible.

[jjasilli]

Cold bias does not make for best guitar tone.  There's a wide range of bias that works operationally.  By "operationally" I mean more than cutoff (too cold) and less than redplating (too hot).  Within the operational range is a sweet spot sounds best. I suggest you play your guitar through the amp @ different bias settings to find the tone you like best.