Bias circuit problems...

[Colonel_Sanders]

Hello again!

I'm having trouble with my 100V zener shunt regulated bias supply. When I swing the 25K pot from left to right, the bias range only changes from -97V to -96V which is giving my 5881's just under 40mA @ 416V .

Can anybody point out where I've gone wrong? I'm thinking I should decrease the 47K resistor? But I'm really not sure...

Cheers,
C_S

[lowell]

Hi I think you need to put that 47k in series w/ the negative bias line and the 25k pot after it for that voltage divider to work.... I think.

[Enzo]

Lowell is close. Without the zener, the voltage at -Vb would be free to change. COnsidering the size of the 47k and 25k resistors, I think the 1.5k resistor would be too small to allow much voltage division. But with the zener, you have 100v ther all the time. Adjusting the pot doesn't do anything but change the load on the zener regulated supply. Lower it enough and it just drags down the supply through current drop. if you frag the voltage below the 100v of the zener, then the zener stops copnducting and disappears from teh circuit - so it would have zero effect.

Not sure why we need a zener anyway, but that is up to you.

The way it is now, you esentially have the zener supply as the bias voltage, with a variable load on the supply. For voltage division to occur, there needs to be two resistances - one above and one below the output.

What is your target average bias voltage? For sake of discussion I will assume -50v. That seems reasonable with 400v or so on the plates..

I will start with the -100v. Let us assume all voltages here are negativ, so I will drop the - sign. We have 100v and we want about 50v, and make that adjustable by some amount. SO using your circuit, move the bias output to between the 47k and the 25k pot. Let us call the 47k, the upper resistor in our divider, Rt, for R top. And the 25k pot, the lower resistor, Rb, for R bottom. It really would be great to asign part numbers on the schematics in the future. Voltage division is real simple to calculate now. Right now Rt + Rb = 72k. Vb = 100v x Rb/(Rb+Rt) = 100 x 25k/72k = 35v.

Too low. What if we made Rt the 25k pot and Rb the 47k resistor? Now Vb = 100v x 47k/72k = 65v. Maybe too high? I kinda wish we were not starting with 100v, I'd be happier with something more like 70 or 65. But that's my problem I guess.

What if we left the Vb tap where it was, atop both Rt and Rb. But lets add a resistor between the zener and Rt in that little horizontal space. Call it Rs for R series. Now Vb is coming from between Rs and Rt, right? Just as a guess, I am going to make Rs 47k, and also as a guess I changed Rt to 33k and Rb to a 50k pot.

Now we have across the 100v a series string of resistors: Rs, Rt, and Rb. At 47k, 33k, and 50k, we have 130k total. And the Vb will be 100v x (Rt+Rb)/(Rs+Rt+Rb), so in this case Vb = 100v x 83k/130k = 64v.

Since Rb is a pot, we can turn it to zero for the other extreme. We now have Vb = 100v x (Rt+0)/Rs+Rt+0) = 100v x Rt/(Rs+Rt) = 100v x 33k/80k = 41v. (Remember with Rb the pot turned to zero, the other two resistors total 80k.)

So by adding 47k Rs, changing Rt to 33k and Rb to a 50k pot, we wind up with a bias voltage that sweeps from 41v to 64v. For just guessing, that is a pretty reasonable bias adjustment range. it would not be hard to fiddle with the values to widen the range or move the whole range up or down.

[Enzo]

And in my example, if you added a 47k Rs, but left the existing Rt and Rb at 47k and 25k, the range would be 50v to 60v. That seems like too tight a range to me.

[Colonel_Sanders]

Great post Enzo, cheers. Aaaaahhhhh got it. Tried to mash the 103 bias supply and a zener supply, but the Zener stays constant and the reverse 1N4001 doesn't. And for some reason it didn't click in my head that the circuit was just a voltage divider, cheers Lowell :-/.

I'm using your values Enzo (but a 100K pot, don't have a 50K) and getting a range of about -38V to -72V. Thanks for the formula as well, big help. And on to the next problem!

Cheers,
C_S

[Enzo]

Slap a 100k resistor across that 100k pot. That will make 50k. it will alter the "taper" but who cares. if it works OK with the 100k, then leave it.

[Colonel_Sanders]

100K's working fine!

Cheers,
C_S