Neg Bias From HT

[C Moore]

I searched through a lot of the 'Bias' archives, but cannot find what I am trying to learn. I know this has been asked and answered before. I am very happy to read a link for the formula/explanation. I just can't seem to find it.
There is no bias tap. I have to take it from one side of the PT secondary.
I can go Cat Bias if need be, but I would rather do fixed bias if possible. How do I go about determining the pot and resistor value that make up the bias circuit.?
I am in process of building an amp that uses 2 x 6K6 power. Guess the book suggest about -20 VDC on the grid.
So how do I calculate a starting point for the bias circuit values.? I am using the PT from the schematic below.
Thank You

http://scottbecker.net/tube/sheets/127/6/6K6GT.pdf

[tubeswell]

Just copy a BF princeton bias supply, but make the DC voltage adjustable

[martin manning]

Use the bias circuit from the TW express, but increase the resistor right off the HT to get the range you need. I think 330k should get -17 to -25V from 270VAC. https://tubeamparchive.com/download/file.php?id=6854

[Super_Reverb]

If you connect the cathode of a Si diode to either phase of your PT HV secondary and then connect a 2 resistors in series from the diode anode to ground, you will have a simple divider with applied voltage (275VAC*1.414) --> caps charge to peak voltage in power supplies.

If R1 is the bottom resistor and R2 is the top one in the divider, then the tap between the two resistors Vbias is calculated in your example PT out =+/- 275AC

Vbias=(275*1.414-0.7) x R1/(R1 + R2)

You are simply multiplying the applied voltage (rectified, filtered DC -diode drop) by the resistor divider ratio.

Let's assume you have a 20K Ohm pot as the bottom resistor R1, then calc above says you need R2= 368K with your 275AC PT.

But, this config will allow you to adjust the bias voltage to a smaller value, it won't let you have a larger (negative) voltage, so let's say you want to have -30V max adjustable toward zero, then use calc above and solve for R2 with Vbias = 30V.

My calcs say R2=238K. For this, you need your R2 about 12x the value of R1, so divider is 1/(1+12) = 1/13. Multiply that times applied voltage 388VDC and you get 29.85V. So, it passes a sanity check.

You should use several resistors to make up the large dropping resistor and place filter caps in between resistors for clean grid bias voltage.

rob

[C Moore]

Would you guys shoot me if I told you I forgot to mention the amp uses a Bias Vary Trem.? Sorry.

[tubeswell]

Trem tube stages need a gain of more than 29 for a 3-stage RC network, so that's something to consider in your choice of tube.

[C Moore]

Trem tube stages need a gain of more than 29 for a 3-stage RC network, so that's something to consider in your choice of tube.

Oh...OK.
Looks like the SN7 is only 20, but the SL7 is 70.
Thank You

[martin manning]

...I wonder if I would be better off to copy the bias AND the Trem of the Princeton. Can I make that Trem circuit work with one half of a 6SL or 6SN type tube?
Doing so would save me A LOT of room.
Thank You So Much

The trem oscillator and bias circuit are separate- you need a DC bias vltage and then, if you're doing bias-modulating tremolo, you need a sinusoidal voltage with a peak-to-peak close to the bias voltage to put on top of it. If you have a circuit for a 6SL7 oscillator, use it. If not, with a little work one could be designed around that tube.

[C Moore]

...I wonder if I would be better off to copy the bias AND the Trem of the Princeton. Can I make that Trem circuit work with one half of a 6SL or 6SN type tube?
Doing so would save me A LOT of room.
Thank You So Much

The trem oscillator and bias circuit are separate- you need a DC bias vltage and then, if you're doing bias-modulating tremolo, you need a sinusoidal voltage with a peak-to-peak close to the bias voltage to put on top of it. If you have a circuit for a 6SL7 oscillator, use it. If not, with a little work one could be designed around that tube.

Are you saying the Princeton Trem circuit will not work with a 6SL7.? I do have a circuit (from Ampeg) that use the L7, but it requires both triodes.
That is the big attraction of the Fender style Trem..... it will save me A LOT of much needed room. Is the L7 a bad choice for this.?
Thank You

[martin manning]

6SL7 is close to a 12AX7, but the mu is a little less, and so is the ra. I think a large plate load and matching cathode resistor, maybe 270k and 3k3 might be a good idea to keep the gain high. The phase-shift c's and r's should be okay. You could marry this oscillator with the TW bias circuit (connected at the bottom of the intensity pot, just like the Princeton), which is a little better IMO. Put a really big bypass cap on the Rk too, like 220uF.

[C Moore]

Thanks Martin, sounds good.
My next problem will finding housing for the fixed bias circuit. When I get that far, it will probably be time for pictures and more help.
Again, Thanks To All
Chip

[C Moore]

One more question.....
Can I just add a second bias pot, for the other tube, at the junction of the 47k-10uf.?
And then run each of those to one of the 220k bias split resistors.?
Thank You

[martin manning]

For separate bias controls, the best thing would be to duplicate the 15k, 10uF, 47k, and 25k pot. Then yes a 220k from each network's 47k/15k junction to the power tube grids.

[C Moore]

OK..... I did not see that coming.
Not sure I have the room for that...... but I will see.
I cannot help wondering, with the second bias pot, what necessitates the duplication of the half wave.?
Thanks Again

[martin manning]

You only need one rectifier and reservoir cap but you need to duplicate the entire divider for each bias sulply. The cap to ground from each supply is there so that both tubes see the same resistance to AC ground. i.e. the 220k bias feed.