Do You Twist Your Wires from Bias Resistors to Power Tubes ?

[PCollen]

And do you worry about making these wires equal length, even if the distance is shorter to one power tube vs. the other ?

[David Root]

No, keep 'em short, you never know about parasitics. Not sure if twisting them together would help vs. parasitics but anything is possible with parasitics. I know bias is DC but it ain't perfect there is some AC too.

I have a long run to one pair of tubes in my current build (with signal from the PI caps, not bias) that I am thinking about twisting or using shielded coax on.

[Alexo]

Ever measure the DCR of an OT primary? It's always asymmetrical from the center-tap to either plate by at least a few ohms. Considering the length of copper needed to add up to a few ohms, I don't think a couple extra inches of wire in the bias circuit will make a difference.

...as for twisting, if it isn't curing a lead dress problem, there's no reason to do it. You could even argue that twisting them hampers the tone by inducing negative feedback at ultra-sonic frequencies, if you were an OCD nitpicker or a marketing executive, but in reality, of course it won't make a difference either way.

[PCollen]

Ever measure the DCR of an OT primary? It's always asymmetrical from the center-tap to either plate by at least a few ohms. Considering the length of copper needed to add up to a few ohms, I don't think a couple extra inches of wire in the bias circuit will make a difference.

...as for twisting, if it isn't curing a lead dress problem, there's no reason to do it. You could even argue that twisting them hampers the tone by inducing negative feedback at ultra-sonic frequencies, if you were an OCD nitpicker or a marketing executive, but in reality, of course it won't make a difference either way.

I twist them and try to get them about the same length, but purely for asthetic and wire-run layout reasons. I never really considered the resistance/impedance idea, but once considered phase difference until I learned that electricity propagates at about 210,000,000 m/sec in wire and so a 5-10 inch length of wire is damn short relative to even a 10K wavelength . Twisting will minimize crosstalk..so will keeping the wires apart from and not parallel to each other. Anyway, it looks good and often recommended so I do it.

[Cliff Schecht]

Ever measure the DCR of an OT primary? It's always asymmetrical from the center-tap to either plate by at least a few ohms. Considering the length of copper needed to add up to a few ohms, I don't think a couple extra inches of wire in the bias circuit will make a difference.

...as for twisting, if it isn't curing a lead dress problem, there's no reason to do it. You could even argue that twisting them hampers the tone by inducing negative feedback at ultra-sonic frequencies, if you were an OCD nitpicker or a marketing executive, but in reality, of course it won't make a difference either way.

The imbalance in DC resistance is from the interleaving process they use to wind low-leakage audio transformers. The DC resistance is different but the impedance is the same.


Twisting the bias wires isn't such a bad idea actually. If you look at the typical CMRR of an amplifier, it isn't all that great. Twisting the wires will cause more similar noise signals to appear at the inputs to the power tubes and I think would help with noise cancellation. Then again, just running the two long wires in parallel should cause each wire to pick up about the same noise. I dunno, just thinking out loud as I haven't tested twisted vs. non-twisted.

[Merlinb]

The cathodes are low impedance nodes, twisting will have no significant effect on performance. Not with wires a few inches long anyway.

[Andy Le Blanc]

Separate bias resistors for a push-pull pair....

Seems funny to separate the cathodes and then inductively couple them.

There's a capacitance of a few pf in the wire insulation too.

But it's a short run, mostly DC current, might be a new trend if someone
starts to say they hear a miracle, aesthetically speaking of course.

[dynaman]

Some folks just spin the PI circuit 180 so the grid resistors go right from the caps to the tube sockets. No wire necessary.

[PCollen]

The cathodes are low impedance nodes, twisting will have no significant effect on performance. Not with wires a few inches long anyway.

...we're discussing control grid wiring between PI and power tubes, or at least that was the intent of the OP.

[PCollen]

Some folks just spin the PI circuit 180 so the grid resistors go right from the caps to the tube sockets. No wire necessary.

I've never seen an amp where the PI circuit on the board (bias R's, plate R's, and coupling caps) is close enough to the power tubes to allow that. But I'd like to, and think it would be a great design. Got any pic's ?

[Cygnus X1]

Some folks just spin the PI circuit 180 so the grid resistors go right from the caps to the tube sockets. No wire necessary.

I've never seen an amp where the PI circuit on the board (bias R's, plate R's, and coupling caps) is close enough to the power tubes to allow that. But I'd like to, and think it would be a great design. Got any pic's ?

Not a pi setup, but a single ended.
But I could easily do it with this kind of layout:

[img:680:360]http://img404.imageshack.us/img404/8867 ... tlarge.jpg[/img]

[azatplayer]

Some folks just spin the PI circuit 180 so the grid resistors go right from the caps to the tube sockets. No wire necessary.

I've never seen an amp where the PI circuit on the board (bias R's, plate R's, and coupling caps) is close enough to the power tubes to allow that. But I'd like to, and think it would be a great design. Got any pic's ?

I got some inspiration from being inside an Ulbrick Arena.
Drew heavily from Daves tagstrip ideas. Pretty cool way to build, this was my first attempt, a JTM50 typology , kind of.
The loose wires in there are the OT wires, i was looking for some issues ad had another OT wired in off the chassis.

[IMG:800:533]http://img831.imageshack.us/img831/5755/imag0016m.jpg[/img]

Uploaded with ImageShack.us

[Gaz]

I've done it that way, where the grids of the power tubes were connected by only the grid stoppers coming from the coupling caps on the board. Don't have a pic, but it wasn't that exciting, and not worth the other layout compromises I had to make. I'm realizing with each build where I have to really worry about noise, and that place ain't one of em!

[azatplayer]

Yeah ive ended up using larmars in the builds ive done since that one, as the MV pots on the front, its best to keep the coupling caps standard format.
Did my express with tagstrips, like it a lot!

[dynaman]

Some folks just spin the PI circuit 180 so the grid resistors go right from the caps to the tube sockets. No wire necessary.

I've never seen an amp where the PI circuit on the board (bias R's, plate R's, and coupling caps) is close enough to the power tubes to allow that. But I'd like to, and think it would be a great design. Got any pic's ?

Just click on the myspace link below.