Bias Test for 4 Tubes?

[ViperDoc]

How do you like building bias test points for amplifiers with FOUR power tubes? Do you place:

A) Two test points, one for a single tube on either side of the OT?
B) Two test points, one connected to both tubes on either side of the OT?
C) Four points, one for each power tube?

[pompeiisneaks]

I'd say c is best, you BUT in theory, if you don't have a separate bias adjustment pot per tube, it's somewhat 'neutral' as a solution, because you can basically bias base on the hottest or coldest tube and not the other in each side of the Push pull pair.

I've seen some setups with each having a bias pot, or a bias pot and a balance pot per side etc... it's honestly probably best only in HiFi if you ask me. A little bit of imbalance actually adds that 2nd order we love so much.

~Phil

[ViperDoc]

This isn't drawn as a layout, just a proposed circuit option. Is this how you would do "dual bias" on an ODS, for example? Or would having two 100 uF 100V capacitors in the bias supply make it stiffer? Slower?

DUMBLE ODS DUAL BIAS.pdf

[sluckey]

Since your dual bias circuit shares the 3.3K the pots will be interactive. Use two separate 3.3Ks to correct.

Here's my only dual bias circuit. No interaction...

[ViperDoc]

Since your dual bias circuit shares the 3.3K the pots will be interactive. Use two separate 3.3Ks to correct.

Here's my only dual bias circuit. No interaction...

OK, very good. Your mod is slick, I like how you designed it. I notice yours uses three 50 uF caps, essentially sharing the first off the bias diode and then one each for the parallel bias circuits. How does the capacitance end up per side? 75 uF?

And is each side of the bias circuit blocked by the pot itself?

Thank you.

[R.G.]

Why not put in one pot per tube?
Yeah, it's more circuitry, but it increases the amp's tolerance for less well matched tubes, and different drift rates per tube.
One per tube also can be set up to make biasing a breeze. If you can set up the post and bias pots so you can get at them from outside the chassis, you can rebias with just a meter and a screwdriver.
The bias scheme on the Workhorse amps had an adjustable bias current indicator and one pot per tube. The indicator was a bi-color LED, where off indicated "below the target bias", green indicated "within a tight range of the target bias" and red indicated "over the target bias". The indicators and pots were accessible from holes in the back panel. You could literally check and re-bias between sets if you were worried about it.

[sluckey]

OK, very good. Your mod is slick, I like how you designed it. I notice yours uses three 50 uF caps, essentially sharing the first off the bias diode and then one each for the parallel bias circuits. How does the capacitance end up per side? 75 uF?

No idea. The important thing is each circuit has a CRC filter rather than a simple RC filter. The bias voltage ripple is much less than a single 100µF cap would provide. If you stare at my schematic for a moment you'll see that all I did was simply duplicate the original single bias circuit for the "Bias 2" circuit.

And is each side of the bias circuit blocked by the pot itself?

Not sure what you mean???

[ViperDoc]

And is each side of the bias circuit blocked by the pot itself?

Not sure what you mean???

I wondered about both bias pots being connected to the 1K in regards to them not being interactive.

[sluckey]

I wondered about both bias pots being connected to the 1K in regards to them not being interactive.

Okay. They are not interactive because of the way the pots are wired. The current flowing through the 1K and other resistors always remains constant. Turning a pot wiper simply picks off a different voltage level to send to a tube grid without changing any current in the bias circuit. Therefore, there is no interaction.

Your circuit has the pots wired as a variable resistor. When the resistance is changed by turning one shaft, the current through that shared resistor changes, causing the voltage to change. Since this "changing" voltage also feeds the other pot, it's voltage will also change, causing interaction. You would probably have to go back and forth between the two pots to achieve balanced current through the tubes. Heck, you may never get there! Adding another resistor should cure this.

[ViperDoc]

How about this? How does the dual 3K3 voltage divider affect the available bias voltage on either side? Half? Or unaffected?

ODS DUAL BIAS DRAFT 2.pdf

[sluckey]

Now ain't that what I suggested real early this morning?

[ViperDoc]

I see it now. I like your single drop resistor/dual CRC design better, though. It seems that instead of the stock design having two caps that supply bias during both sides of the AB operation cycle, you have one 50 uF that supplies through the entire cycle, and then individual 50 uF caps that each supply on only half the cycle, either A or B. This gets back to my question of how that changes either the capacitance or more likely the faster speed at which the bias supply can recharge. Thanks for the explanation.

[Reeltarded]

RG is a NERD! He also might work for BIG POTENTIOMETER!!

[sluckey]

This gets back to my question of how that changes either the capacitance or more likely the faster speed at which the bias supply can recharge. Thanks for the explanation.

You're spending too many brain cells on stuff that don't matter. CRC is smoother than CR. Unless you're driving the output tubes so hard that the grids are drawing current (not a good thing), the only current flowing in the bias circuit is due to the components in the bias circuit. Not like the plate supply node where you must be constantly pumping the caps back up because of the heavy current demand. The bias supply only needs an occasional small pat on the back to keep the caps fully charged.

[ViperDoc]

You're spending too many brain cells on stuff that don't matter. CRC is smoother than CR. Unless you're driving the output tubes so hard that the grids are drawing current (not a good thing), the only current flowing in the bias circuit is due to the components in the bias circuit. Not like the plate supply node where you must be constantly pumping the caps back up because of the heavy current demand. The bias supply only needs an occasional small pat on the back to keep the caps fully charged.

Thank you, Steve. If you were to apply this circuit to an ODS build, using three 50 uF caps would be appropriate, then.