Electrical engineers, feel free to chime in here!

I'd like to use a short section of two conductor shielded wire (conductor pair + shield) for internally routing through my front rack tubes. I need three conductors total for headlight, taillight and ground (I am not using the fork as ground, even though the dynamo grounds to the fork). There are lots of wire options available, but I'm restricted to using 3mm thick cabling for the internal run, anything fatter won't work. The only two conductor + shield cable I've found that's only 3mm thick uses 28AWG conductors for the pair. It's basically line-level stereo cable. Are these too narrow gauge for this application? Will they heat up? Will there be significant voltage drop? If it matters, the internal run will be no more than 1 foot long. (Voltage drop is proportional to length, right?) All other wiring throughout the system will be 22AWG. the standard for dynamo lighting.

I've bench tested my dynamo lights using about three feet of the 28AWG conductors and they light up fine, but I'm only concerned about potential resistance issues and wires heating up with extended use.

Thanks!

the search term you are looking for is "ampacity"

I would be more worried about breaking wire that light. I would go 2 conductor and split after the bottleneck.

Here is a chart:

https://www.powerstream.com/Wire_Size.htm

There is also a voltage drop calculator on the page.

I'd be more worried about durability. Will there be any flexing? Any wiring from the fork to the frame?

Yup - mechanical durability should be a bigger concern than current carrying capacity. Bicycles produce a lot of vibration and and the wires will slso be exposed to condensation, heat from exposure to the sun etc. Use larger wires.

Thanks for the replies. I'm not worried about it breaking; it's stranded, not solid core, and the sheath is pretty tough. This is the stuff that is used for headphones and such, so flexibility is not a problem. There is no flexing or strain expected in this installation; it's going from the rack leg to the light, that's it. No chance to go 2 conductor for this short 1 foot stretch, I absolutely need 3 conductors.

I'm not recommending it, but I'm testing army surplus telephone wire for my dynohub setup. 23awg 2 conductor zip line with 4 tinned copper strands and 3 galvanized steel strands per conductor. The insulation is very tough as well, I'm able to split it and run a conductor down the rolled seam of my Honjo fenders. I've also installed it in a Peugeot, replacing the original internal downtube wire. it used the frame as chassis ground, I'm running twin conductor, for an independent ground.

On one bike I've looped it loosely once around the head tube-fork crown area. On the other I've turned the wire around a rod to make a coiled section from the hole in the down tube to the fork mounted light. I think they will both last a long time.

Coiling wire is a nice touch between the fork and frame. But I discovered that coiling wire around a rod generally gives poor results; the coil always unwinds as the insulation doesn't have "memory" in the new coiled state. After coiling on a wooden rod (usually a pencil) I heat my insulation up to close to melting temp, either in an oven or with a heat gun. This "erases" the original memory and remodels the insulation to remember the new coiled shape. Years later, my coiled wires continue to snap back into a nice coil even if they're repeatedly stretched:

Update: I've tested the light (a re-wired hanging Schmidt Edelux) on my morning commute. I disabled my current dynamo light and plugged in the Edelux with ~3 feet of 28awg wire. I rode about 15 minutes with the light on (short commute). At the end of my commute, the wiring felt as cool to the touch as the unused wiring to my disabled light. I suppose I should test the light more thoroughly, say for an hour or longer with continuous use. But so far it's very promising!

The reason for rewiring the Edelux is because when Schmidt introduced their "hanging" or "inverted mount" Edelux, they disabled the taillight lead because its location would have allowed water penetration. Their solution was to remove the taillight lead altogether. They expect buyers of their $200 light to install a separate taillight switch elsewhere on the bike (inelegant IMHO) or just use a battery taillight. They have the 3mm coax cable that goes from the dynamo to the headlight entering the inverted light through a 3mm drilled hole, sealed with glue from the inside (a crude but effective technique, IMHO). I wanted to keep the look of the svelte 3mm coax cable but needed an extra conductor in there to tap into the taillight circuit (which is still there on the circuit board, just not used). I had wanted to ground the circuit board to the aluminum headlight housing and re-purpose the coaxial braided shield as the main dynamo hot lead, and repurpose the 22awg single conductor for the taillight. But internally grounding the circuit board to the light housing was not possible. My challenge was to find 3mm round cable with three conductors. This limited me to the twin conductor 28AWG with braided shield, sold as line-level audio cable.

I'll post photos of the rewired light soon...

Sexy!

Seems the size restrictin is for the length of the wire that is going through the hollow of the rack pipe...however the rack is made of metal so it could serve as the shield there, from a shielding standpoint. Thinking you could go up a gauge or two on the wire to gain higher ampacity by just removing the shield where the cable goes through the rack???