Fixed bias: Pots near supply or tubes?

[DrewP]

Is it preferred practice to set up a bank of pots next to the bias supply caps and run flying leads at the adjusted bias voltage over to the output tubes, or do you run leads carrying the full voltage of the bias supply from the supply across to near the output tubes and have the pot as close as possible to the tube socket?

[HollowState]

Technically speaking, it makes little difference where the pot is located. But personally I believe in keeping signal wiring as direct as practical (not possible) to reduce extraneous pickup if nothing else. Therefore, I plan my layouts with the bias pots reasonably near the output tubes since they are connected to the grid(s), even if through a resistor.

Victor

[Enzo]

I don;t think it matters, whatever is most practical in your case. After all this is a DC supply, it isn't going to radiate anything, nor is it going to pick up anything. As far as the signal is concerned it is pretty much just ground.

[DrewP]

Having the pots local to the tube and the bias test points would certainly make it more intuitive for the idiot adjusting the bias current (namely me).

[richwalters]

Originally posted by DrewP
Having the pots local to the tube and the bias test points would certainly make it more intuitive for the idiot adjusting the bias current (namely me).

Why not make them more inaccessible ? I make adjustments through the top chassis near the o/p tubes (marked red felt in pic). There are advantages in having presets/pots fairly close to the output stage; i.e less HF pickup.
Note, some push-pull amps do rely on a symmetrical drive layout for optimum HF response and long wiring can provoke instability. This is where I find 6550 circuits and perform better than KT88 at the higher audio frequencies as the 6550 requires a lower impedance driver.
I use 10Turn cermet presets which are very reliable and can take heat. Once adjusted for correct Iquies there should be no reason to tweak.
If you are continually tweaking then something in the circuit isn't correct. Either the tubes and/or the neg bias rail versus B+ are running wild or there is circuit instability.

...Once I came across a rocker who was trying to set o/p stage bias currents by running hard music through the amp:- one learns the hard way.

richy

[DrewP]

I'll be using wirewoundsand yes, I won't be making them readily "twiddle-able", it'll be a screwdriver through a hole kind of affair, same as you have implemented.

[richwalters]

Originally posted by DrewP
I'll be using wirewoundsand yes,.

What's the reason for wirewounds for a few mA ?

(historically, I find them over time these gradually go intermittant ..not my favourite)

richy

[DrewP]

A close friend has used them in his commercial work for well over a decade and found them reliable.

Not that I have anything against cermet...

[richwalters]

I often use this configuration for my parallel 6550 power stages: don't copy it value for value. It requires cast iron driver stages: the 68K+ preset res loading on output is quite high and conventional stages will run into distortion.
The most important component is the 100K across the wiper to neggy in case if wiper goes open it will save the tube.

richy

[ChrisA]

Originally posted by richwalters

What's the reason for wirewounds for a few mA ?
richy

If I may answer. Wire wounds are the most stable. Their resistance value tends not to change with age, temperature or applied voltage. That said I don't think you need milliohm level precision here and any decent trim pot would work fine.