• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Hook Up Wire Gauge?

As has been indicated, there is no simple answer to this, and safety considerations dictate you take into account both the current draw and the voltage rating of the wire.
 

Attachments

  • Heater Wire Gauge.gif
    Heater Wire Gauge.gif
    5.9 KB · Views: 735
here's something you rarely find in tables: AWG vs volts-per-foot or per meter, per amp

AWG … … volt/amp-ft … volt/amp-m
________________________________________

10 … … … 0.00100 … … … 0.00328
12 … … … 0.00158 … … … 0.00520
14 … … … 0.00251 … … … 0.00824
16 … … … 0.00398 … … … 0.01306
18 … … … 0.00631 … … … 0.02070
20 … … … 0.01000 … … … 0.03281
22 … … … 0.01585 … … … 0.05200
24 … … … 0.02512 … … … 0.08241
26 … … … 0.03981 … … … 0.13061
28 … … … 0.06310 … … … 0.20701
So basically you take the length of both conductors (assuming a twisted filament wire, as an example), and multiply it by the gauge factor above, then multiply by the expected design current flow.

1600 mA on a 6550, running nearly ⅔ foot from the power transformer:
1.6 A × 2 × 0.67 ft • (18 ga = 0.00631 v/ft-amp) → 0.0135 V drop, round trip.​
This would be about 0.2% of 6.3 VAC. Not worth worrying about, since the variation of line voltage coming thru the power transformer easily varies by ±5% from whatever is nominal at your house.

You even could get away with 22 or 24 ga.

The recommendation of 22 to 24 gauge wire is just fine for almost everything in most conventional tube amplifiers.

Except the output wiring from output transformer to speakers. The currents are much higher, so the gauge thickness needs to be larger (lower number).

⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
 
Last edited:
See the opening on the lug of the tube socket?, that's your size of wire you should use...& solid core of course. Some small rounded nose needle-nose pliers so you can make a radius according to the nose of the pliers. You only get one shot at bending the wire to turn it, if you mess it up & try to bend it back because you went too far, consider the wire compromised & start over. That's my take, but that's just me.







------------------------------------------------------------------------------Rick........
 

Attachments

  • Bottom-Chassis-Mount-9pin-Ceramic-tube-sockets-Valve-Tube-Base-for-6DJ8-12AX7-font-b-ECC83.jpg
    Bottom-Chassis-Mount-9pin-Ceramic-tube-sockets-Valve-Tube-Base-for-6DJ8-12AX7-font-b-ECC83.jpg
    247 KB · Views: 327
The point about the use of wire which loosely slips into a socket hole is spot on, actually. Which is why I am fond of smallish solid conductor wiring, with relatively high-temperature insulation. I absolutely hate it when I nip a clean, tiny bit of bare copper, bend into a hook, attach onto a socket, crimp the hook end, apply solder, and have the insulation pull-and-melt back down the wire. Grrrrrrrrrrrrr!!!!!

I tend toward 22 GA for everything. Filament wiring is pre-made (a moi) with two different colored 22 GA wires; Its a nice skill, but if you take a helper and an electric drill, you can cut a couple of 3 ft ÷ 1 meter long pieces of insulated wire, fit them to the chuck, have the helper hold the end pointing toward you. Wearing cloth gloves, you just pinch the wire near the tip. As it spins, the individual wires rotate axially, which yields a twisted pair cable that doesn't unravel.

Hookup wire never needs to be very long, so once you get the hang of it, just make a half dozen of these things and hang them on a wall or inside a cupboard door. Always useful.

Since I 'star source' my filaments (i really don't like chaining them), having only 22 GA filament runs, even for my 6550s is just fine. Max, 0.5% losses.

Its also pretty wise to invest in a pretty good wire insulation stripper. A 'good enough one will last a lifetime, or until lost or “liberated” by a friend from your workshop. Mine is on a little chain, attached to the workbench. Keeps me from misplacing the darn thing.

Tho' many an old-timer and purist will ward against pre-tinning one's wires, I am quite fond of it. It really does hone one's skills with the soldering iron, and make for much, much less likely 'cold solder' connections.

Lastly, although perhaps not entirely necessary, i do advocate wire-wrapping with waxed nylon strap your filament wires, to run in directions and away from the signal business.

As to whether one uses 22 GA for everything (like I do, except for speaker connections), or not … its really more hocus-pocus voodoo to ascribe all sorts of wonder-and-magic to specific recommendations. Truth is, there is so little current flowing through the signal pathways of one's amplifier that any series resistance from 22 GA over say 20 or 18 … is nearly unmeasurable. Seriously so.

Pass along the good info, goats.
Its just high end audio!

⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
 
Why solid core only? Skin effect at 60Hz is pretty null no ?

I think the reasoning is mostly “solid core stays put” once bended the way you want it to go. The other is “stranded tends to solder very erratically” if you have even slightly 'old' wire. A little oxide builds up on the copper, and it doesn't like taking a good wetting of solder.

(However, this is why I keep a once-ordinary denatured alcohol lamp on my bench. Bare the stranded, light the alcohol lamp, then heat the wire in the alcohol flame. The -OH business on ethanol or isopropanol (C2H5OH and C3H7OH respectively) at high temperatures does a great job reducing copper oxides back to copper metal!)

So, technically, if you like stranded, stranded is just fine. Just have to learn how to deal with its idiosyncrasies.

⋅-=≡ GoatGuy ✓ ≡=-⋅
 
Last edited:
Got it, thanks. I agree.
Just the solid core sometimes snaps, when stranded never does.
I wasted a week on an amp with an intermittent problem, couldn't find the problem.
In despair I started to change all components in that stage... to find a solid core wire open in it's sleeve so I could not see it. Bummer.
 
We can calculate the skin-effect at Ness Engineering Inc. Home Page - Ness Engineering Inc. And yes, skin-depth is negligible even up thru 20K.
Solid core is for structural rigidity of keeping wires away from EM fields via proximities to other wires. Of course PC boards throws that all away, I'm focusing purely on point-to-point construction, turret boards & the likes.







------------------------------------------------------------------------Rick...
 
Last edited:
I use 22 or 0.5mm³ for everything except high current heaters which get #18. Solid stays where you put it, and stranded is frankly a pain in the butt to work with when the wire size gets small.

As GoatGuy said, twisting wire in a drill makes a very nice twisted pair indeed. If the wire is shorter than the span of my arms, I can twist it myself using just the drill - no need for an assistant unless the wire is longer than about a metre.. I never use a glove though. I get a much tighter twist that way 🙂 YMMV.

Also, if it's signal, you can use solid ethernet twisted pair for an easy way of getting twisted wire from common CAT3 up.

If you plan on letting your chassis run hot and it's a tube amp so that's almost a given, you might consider non PVC wire. The heat bakes the plastisizers out so when you go to move the wire, it cracks. Some PVC is made for higher temps though. Just something to keep in mind.