Heater wiring - Coaxial?

[DF96]

When wiring heaters there is no advantage in using fancy wire. The twisting is far more important that what is twisted. Also important is the route followed by the wire, and how you untwist it at each valve holder.

Regarding coax vs. twisted for magnetic field cancellation, I am surprised if someone has found that twisted is better. The net magnetic field summed in a circle around a current-carrying wire is proportional to the total current. Two wires carrying opposite currents (coax or twisted) give zero field sum around the circle, because the total current is zero. This does not necessarily mean zero field everywhere, but I would have thought that cylindrical symmetry would guarantee this for coax but not for twisted. Was there some special feature of the setup in the cited article which changes this?

A magnetic 'shield' which broke the cylindrical symmetry might make things worse, not better? Just a speculation.

 

[trobbins]

DF96, I'm going to have to pull my head in regarding that referenced paper - I misinterpreted the test set up and results table - too many beers! So I think I'll take a step back and agree with your comment - apologies.

Ciao, Tim

 

[jcx]

ideally coax with exactly equal shield, core currents will have no external magnetic field

twisted pair just has a field that falls off fairly quickly at multiples of the wire separation distance, and at a higher order at multiples of the twist pitch

the principle is to have a "common centroid" for the +/- currents, twisted pair are always separated by their insulation thickness

with individual wires star quad can have the common centroid property if each of the currents divides equally in each of the diagonal pairs, still some nearfield flux but falls off quicker than twisted pair for similar wire separation

 

[DF96]

OK, so coax is best for magnetic cancellation but twisted is best overall (of simple wires). That is why we all use twisted. You could get more complicated, as jcx mentions, but I doubt if it is worth it - use DC for low level inputs if twisted AC is not good enough.

 

[Rundmaus]

So many comments on my late-night idea...

As my current project does not allow to put the heater wires in the chassis corners (space limitations, position of tubes etc), the test setup will give me enough chances to experiment with different heater wiring approaches - hopefully finding a hum-free solution...

Greetings,
Andreas

 

[payitforwardeddie]

Me? I will stick with the bog standard twisted pair kept well away from other wiring and components. Everything in balance.

BTW, Rundmaus..the only stupid question is the one you don't ask.

 

[DF96]

Don't worry too much about chassis corners - this helps, but most of the hum is likely to come from wiring near the valveholders. Playing with different solutions is made more complicated by the requirement to put the heater wiring in first (close to the chassis).

 

[Rundmaus]

The circuit (well, one channel) will be bread-boarded in the same geometry it later has on the chassis - if it will not hum on the piece of wood, it hopefully will be quiet when put into a chassis afterwards...

Greetings,
Andreas

EDIT: As the enclosure for the amp will be quite high (6-8 inches), perhaps it is a good idea to go into third dimension - components and signal wiring directly below the chassis, heater wiring vertical down from the sockets (4-5 inches) and then to the heater transformer. Distance might be a good shielding against AC heater hum.

 

[DF96]

Normally heater wiring is next to chassis, other components a bit further away. This gives a measure of electric field shielding, as the heater wire voltage has an opposite image reflected in the chassis. I suppose you could swap over, but you can't mount most components right next to the chassis. Remember that over the life of the amp you are more likely to need to change components than the heater wiring, so make sure the components are accessible.