want 4 bias adjustment pots for BF Showman

[deci belle]

I still haven't found what I'm looking for (searching= 4 bias pots).

My pwr tubes' bias numbers are getting farther and farther apart. I'm having to juggle pairs more often to keeps things somewhat balanced as they drift more than they used to. I'd like to be able to install an individually adjustable circuit for each tube before I have to get new ones (also so I can mix tube pairs).

My bias supply (–61v) feeds a 10k pot and a 27k resistor going into two 220k grid-load resistors (one to each 6L6 pair).

Should I use four 10k pots and four 27k resistors to feed four 220k grid Rs? Will I be able to use just the two .1µF PI coupling caps the way it is now?

Could someone please post a schematic of such a circuit for me, or describe what is required for me to implement the change I want to make?

Thanks!

[defaced]

http://music-electronics-forum.com/t16704/

My quad bias schematic is in post #8 of that thread. You will need four coupling caps between the phase inverter and the power tubes to keep the bias supplies separate. As an aside, so you understand the freely hanging 100k resistors (R16-19), I have included those bias feed resistors (or as you're calling them, grid load resistors) in this schematic because I made this into an SMD board with those parts on board.

[Enzo]

And when you find tubes "drifting" make sure to check the operating conditions. For example the mains voltage. For every volt of change there, your B+ will change 3-4 volts. Your bias will vary too, but only half a volt or less for that same mains volt. If your mains changed 5 volts between session A and session B, that 15v or more change in B+ can easily cause your familiar tube current to change as well. And not be the tubes' fault. Besides, bias is not all that critical anyway.

Your search might work better looking for independent bias or individual bias.

[deci belle]

It's a beautifull thing, defaced— ok, R1 is the 100k master bias pot with a 22k R15, saving the tubes from bias failure meltdown.

My guess is the four coupling caps aren't on the schematic.

What kind of caps are C1~C5— 10µF 100v electrolytics? Those must be isolated power nodes providing negative bias voltage to each of the tubes' bias pots.

I get the 100k 1w bias feed resistors (R16~19).

This is a lot for me to grasp, but I really appreciate your response— I'll have to turn it into a layout to find out more things to ask about.

Thank you!

[deci belle]

Ya bias isn't all that critical (heck, I was blissfully cranking away with only 3 tubes in the other day! haha!— and thanks for that search tip, Enzo!

[defaced]

Correct, the coupling caps are not shown. In my build, I have those stashed on a different board.

Yes, C1 -5 are 10u, 100v electrolytic caps.

The nodes are not isolated. During initial adjustment you may see some interaction, but once you settle around the final adjustment point, interaction should be so small it doesn't really make a difference.

All that said, I agree with Enzo, bias is not that important. This project was more for playing with mixing tube types as opposed to achieving perfectly matched bias for all of my tubes. Now that I've reaffirmed that I like KT-88s pretty much exclusively, I'll be using standard boring single bias supplies in future builds.

Attached is an old through hole layout I did for this. I don't think the part numbers or some of the values (caps particularly) match the schematic in the other thread, but the general layout of the schematics are the same.

[deci belle]

Cool, defaced!

I'll be ready to just stab in a set of matched tubes and call it good (after I figure this out, of course)!! I want to try out different tubes, as you have done, l to see what works best for me.

As I study your schematic and pcb layout and draw this out, I find I don't recognize how each pair of resistors work, and I'm not clear about the ground traces. I know where the caps (+) go to GND, but it isn't clear where the GND traces are on the thru-hole layout— and the pots aren't making sense yet for the same reason.

But— thank you for the study material!

[defaced]

They do a couple of different things. If you go back to the thread I linked to originally, Enzo and Paul worked through the methodology of how to do this based on the original Marshall bias supply. Because that is a single supply, it's a little easier to see the arrangements of things, so use the two schematics in tandem to see the similarities and differences between the two.

Terms:
First resistor - the 10k resistor between the two 10 uf caps (R2, 4, 8, 11)
Second resistor - the 10k resistor that the bias is taken from that also connects to the pot (R3, 5, 9, 12)

The first resistor and the two 8 uf caps form what is called a Pi filter, or CRC filter. Basically, this arrangement provides very good smoothing of the rectified DC voltage. This resistor, in combination with the 22k/100k resistor pot arrangement, provides a limit for the maximum negative voltage the bias can supply. I think to some extent this cap and resistor arrangement also provides some level of isolation/decoupling of the branches, but I've never bothered to really test this by removing this filter stage.

The second resistor provides a limit for the minimum negative voltage the bias can supply. This is the idiot proof resistor that ensures that if the pot is turned all the way to one of the extremes that the tubes won't red plate.

Both resistor values were determined through simulation, caps were chosen because of their physical size and that they were "enough" for the job. If you really wanted to, you could do the hand calculations for the extreme positions of the bias and range pots (this isn't anything more than a simple resistor network), or just develop the resistor values for your amp empirically. There really isn't anything "magical" in this design per se; the layout/structure is more important than the specific values. Keep the caps in the tens of uf range, the resistors and pots in tens to hundreds of k ohm range, and you'll likely end up with a perfectly fine bias supply.

And if you find you have more questions, keep posting them here. I'll yack for days over stuff like this.

[deci belle]

ok, off to Enzo+Paul's back and forth for a while— that'll keep me busy for a week!

[deci belle]

I think I did this correctly (thanks to defaced) but I'll include the description/exercise of my grasp of how defaced's quad-bias design would apply to my old Showman …also, I'm really weak on how the pots should be wired

My Showman's bias supply consists of the pwr tranny's bias tap, a 470Ω R, a diode and a 100µF 100v cap, (+) to GND.

Between the cap's (-) end and the diode, two 22k Rs branch out; one goes to the 100K-L (master bias) pot's #2 wiper tab; the #2 and #1 tabs are bridged. The other 22k R connects to the same pot's #3 tab. Nothing on this pot goes directly to ground. The lead continues on from the pot's #3 tab.

Next, the lead connects to the (-) end of a 10µF 100v cap; its (+) side goes to GND.

The lead continues past this cap's (-) end and connects to a junction joining four identical circuits, one to each separate power tube.

Since all four circuits are the same, I will describe just the continuation of the lead past the master bias circuit's 22k resistors, 100K-L pot and 10µF/100v cap to the four way junction.

One end of a 10k R connects to the junction and its other side goes to the (-) end of a 10µF 100v cap; its (+) end goes to GND.

The lead continues past this cap's (-) end and connects to a 10k R, which branches off and connects to the #3 tab of a 100K-L pot. The #3 and #2 tabs are bridged. The #1 tab goes to GND.

The main lead now leaves the resistor+pot branch and connects to a 100k bias feed/pwr tube grid resistor. Next in line comes a .1µF 600v coupling cap that branches off; its other end is B+ voltage/PI output.

The lead continues past the coupling cap and terminates at the first power tube's #5 pin via a 1K5 R.

I'm really sorry I couldn't show the drawing.

I think I'm close, except for the bias pots' connections. Would you please comment on this, defaced!

Mar'

[defaced]

My Showman's bias supply consists of the pwr tranny's bias tap, a 470Ω R, a diode and a 100µF 100v cap, (+) to GND.

Keep all of that. Connect the (-) terminal of the 100uf cap to the input junction of R14 and R15.

For certain things, I find it easiest to take the schematic, and layout the parts exactly like drawn on the schematic when I'm building. I built my first long tailed pair phase inverter that way. This would be a good circuit to do like that for the first run of building it.

If you have an image, that would really help. Layouts are hard to describe in words. To post an image from your computer, click the "Go Advanced" button under the quick reply box. From there, click the "Manage Attachements" box, the "add files" button is in the top right of the window that just opened. Follow your nose from there. Hit "preview post" to make sure the image is included in the post.

This should help clear up some things on how pots are wired. Beavis Audio Research
For this case, if you swap which outside lug gets tied to the center lug, all you'll do is change which direction the knob turns to increase/decrease the bias. I actually think mine are backwards from what I wanted, but really I just watch the meter and twiddle my screwdriver till it's right.

[deci belle]

I love Beavis' site! I have a whole file on pots— ha!! Haven't drawn enough of them yet I got the drawing posted— thanks mate!!

[deci belle]

Here's another good site about pots.
Potentiometers (Beginners' Guide to Pots)

[defaced]

How's the build coming?

[deci belle]

Hi! I'm slowly getting the parts and a little eyelet board too! I'm looking forward to nailing this and documenting it for the other 4-tube folks on the forum.

S'body floated me a scope, and I've been investing in some test equipment to compliment that serendipitous boon— so that's why the bias build is on hold!!

Gonna sneak up on it on the cheap, then get some cool little cermet pots I found, when the values are dialed.

Thank you for checking back on this one, defaced!!