Tube filament wiring question

[Davebassman]

Hi,

I'm building a Bassman aa864, and I choose to wiring the filament close to the chassis instead "flying" over the tubes like Fender did.

The other question is that I'm thinking to wire first the power tube instead the pilot light. First power tubes with awg 18, preamp tubes and pilot with awg 22.

It's a good idea? What do you think?

Any advice will be welcome.

[loudthud]

Your main objective has to be to keep the filament wiring away from the signal wires. With that strategy in mind, I'm not a fan of the way wires are routed across the top of the preamp tubes. I would favor keeping the filament wires laying on the chassis and the signal wires routed up in the air until they are some distance from the tube sockets, then close to the chassis and up over the edge of the eyelet board. Another way might be to mount the eyelet board on standoffs perhaps 1/2 an inch above the chassis, then just keep the signal wires in the air to the board. This adds a restriction on any large diameter parts on the board.

The way Fender did it, all the signal wires are routed laying on the chassis and the filament wires are up in the air.

That's a good strategy for the wire gauge. Will the wires from the transformer go to the first power tube ? You don't need 18 gauge to go from the power tube to the pilot lamp.

[Justin Thomas]

I'll usually put a terminal strip somewhere near the transformer to run the heater winding to & then I'll run one set of heater wires to the pilot light & another string to the tube heaters. Really depends on the geometry of the amp.

Like loudthud says, put more effort into keeping heater wiring well away from signal wiring. I'm also a big fan of keeping the heater wiring laying on the chassis & having the signal wires or components flying in the air. Even in Fender-style amps.

Jusrin

[Chuck H]

Yep. Filament wires on the chassis and signal leads up. As to wiring the power tubes first, I like Justin's idea with the terminal strip. I think I did it this way once too. If you can locate the terminal strip en route to the power tubes then that would virtually achieve your goal of going straight to the power tubes first. But now you can wire the pilot from the terminal strip instead of multiple leads on the power tube sockets. And I also wire 18ga to the power tubes and 22ga for the preamps. As loudthud mentioned you should be able to use smaller than 18ga for the pilot.

On one of my designs I locate the first preamp tube close to the input jack at the front of the chassis with the other preamp tube sockets to the rear of the chassis like usual. Rather than run filament leads to that first socket up in the air or on the chassis going under the preamp board I actually ran them from V1 to V2 on the outside of the chassis through a couple of holes with grommets.

[mozz]

I usually do the terminal strip also. You can easily put your 100 ohm resistors there too if using that method. If its a homebrew/conversion/diy I try to use a 120v indicator lamp. Terminal strips also come in handy if I have to add filament dropping resistors.

[J M Fahey]

Real *sensitive* wires are those carrying high impedance low level signal, so it pays to place them close to chassis which offers crude shielding, better than nothing.

Filaments are not that critical.

[Chuck H]

Real *sensitive* wires are those carrying high impedance low level signal, so it pays to place them close to chassis which offers crude shielding, better than nothing.

Filaments are not that critical.

Maybe I've been lucky doing it the other way around.?. But somehow this rings true with my experience and makes sense. Some of this ideal is demonstrated in older Marshall layouts. I'll be reversing my scenario on future projects and paying attention Thanks Juan

EDIT: I'll still be running the filament leads outside the chassis for my one design rather than spanning or going under the peramp board. Of course.

[Davebassman]

Thank you very much for the advice. I now have the filament wires routed close to the chassis, and I’ll keep it that way. If I encounter noise issues, I’ll make adjustments. Regarding the 100R CT resistors, they’ve always been soldered to the pilot light. What do you think about using the filaments from the first power tube and soldering them to pin 8 of that tube?

Thanks.

[Chuck H]

Thank you very much for the advice. I now have the filament wires routed close to the chassis, and I’ll keep it that way. If I encounter noise issues, I’ll make adjustments. Regarding the 100R CT resistors, they’ve always been soldered to the pilot light. What do you think about using the filaments from the first power tube and soldering them to pin 8 of that tube?

Thanks.

If you mean to change the location of the false CT resistors (not sure I understand the question) then you'd need one 100r to pin 8 on EACH power tube. Not just the first.?. Anyhoo... I wouldn't prefer this. I'd leave them on the pilot or on an intermediate terminal strip only because I try to keep extra stuff off the power tube pins so as not to complicate later repairs. And I usually like to add 1R resistors from pin 8 to ground for easy bias reading. JM2C.

[R.G.]

Wow, talk about topical with me. Central Texas is locked in one of its periodic Ice Ages, where the temps drop below 32F/0C every single night at about 5AM, and it never gets over 40F during the day. This has afforded me the time to just think and write about things, and this time it was grounding and shielding.

Juan is of course right - the really sensitive wires are high impedance (on both ends) and carry low level signals. They are prone to picking up signals from wires that have low impedances on the driving and receiving ends, and high voltages. Other than inputs (guitar input, reverb return, FX loop return) most signals are not all that low level; but practically every wire connected to a tube grid is high impedance. Conclusion: Be careful of every wire that connects to a grid.

For just heater wires, you can improve things by twisting the heater wires fairly tightly between tubes. It turns out that twisted pair tends to confine the fields between the wires to the spaces right between the wires. The fields then tend to cancel outside the pair. I consider it good practice to use 4-5 twists per inch on heater wires, and best practice to use shielded twisted pair with the shield grounded on one end. True, this is overkill for all but the most lovingly constructed amps ... like mine. One scheme I've seen in over-the-top Lah-De-Dah(r) hifi amps is to solder copper tube or brass tube to the chassis and run heater wires in the tubes, but I think that guy was nuts.

If you think about it, "high impedance" and "low impedance" are relative. Other than currents running through the chassis metal itself, the lowest impedance conductor in a typical tube amp is the rectifier return wire, connecting the first filter cap negative terminal to the PT CT. That wire carries pulses of current at 2x line frequency, and peaks of two to six times the whole DC currents out of the power supply. Its impedance is really, really low. It's so low that it can yank around the ground reference for every single stage in the amp, and the spike waveform has many, many harmonics. Really think about how that one gets routed.

[Davebassman]

Wow, talk about topical with me. Central Texas is locked in one of its periodic Ice Ages, where the temps drop below 32F/0C every single night at about 5AM, and it never gets over 40F during the day. This has afforded me the time to just think and write about things, and this time it was grounding and shielding.

Juan is of course right - the really sensitive wires are high impedance (on both ends) and carry low level signals. They are prone to picking up signals from wires that have low impedances on the driving and receiving ends, and high voltages. Other than inputs (guitar input, reverb return, FX loop return) most signals are not all that low level; but practically every wire connected to a tube grid is high impedance. Conclusion: Be careful of every wire that connects to a grid.

For just heater wires, you can improve things by twisting the heater wires fairly tightly between tubes. It turns out that twisted pair tends to confine the fields between the wires to the spaces right between the wires. The fields then tend to cancel outside the pair. I consider it good practice to use 4-5 twists per inch on heater wires, and best practice to use shielded twisted pair with the shield grounded on one end. True, this is overkill for all but the most lovingly constructed amps ... like mine. One scheme I've seen in over-the-top Lah-De-Dah(r) hifi amps is to solder copper tube or brass tube to the chassis and run heater wires in the tubes, but I think that guy was nuts.

If you think about it, "high impedance" and "low impedance" are relative. Other than currents running through the chassis metal itself, the lowest impedance conductor in a typical tube amp is the rectifier return wire, connecting the first filter cap negative terminal to the PT CT. That wire carries pulses of current at 2x line frequency, and peaks of two to six times the whole DC currents out of the power supply. Its impedance is really, really low. It's so low that it can yank around the ground reference for every single stage in the amp, and the spike waveform has many, many harmonics. Really think about how that one gets routed.

R.G., you have no idea how much of an honor it is to receive your feedback. I've been reading your website, schematics, etc., for years... thank you very much for the response. I will take into account all the suggestions from the others. I'm interested in the shielding of the filament wires; I actually thought about that but it seemed excessive. From what I see, it's not that unusual. Could you upload a photo of one of your amps to show the shielded wiring? And what type of cable have you used? I would really appreciate it!
Thank you.

[Justin Thomas]

Holy crap, if there's no reason why that copper pipe trick wouldn't work, I'd consider trying that! Just, because it's me.

I have a feeling I'm going to end up with a collection of pasted notes from R.G.'s posts on my phone.

Jusrin

[Tom Phillips]

...One scheme I've seen in over-the-top Lah-De-Dah(r) hifi amps is to solder copper tube or brass tube to the chassis and run heater wires in the tubes, but I think that guy was nuts...

Wouldn't a steel (Magnetic) tube be needed to be really effective?

[mozz]

I have seen 3 wires twisted together, might have been a Bogen PA. amp. The third wire was grounded on 1 end.

[Chuck H]

Wouldn't a steel (Magnetic) tube be needed to be really effective?

Unless I'm mistaken a plain copper tube (NOTE* not even grounded) will mitigate radiant fields effectively because the copper reacts effectively to magnetic fields even though it's not magnetic.?. Where a steel tube would need to be grounded (so it wouldn't magnetize and just pass the field emission?) for the same affect. Random readings. Don't quote me.