Hacker News new | ask | show | jobs
by mdorazio 1760 days ago
Other commenters here haven't really provided any good data to show the issue. If you want to run, say, 10A at 24V through DC wire on a circuit you end up needing stupidly thick wire (~6AWG) to avoid voltage drops over just 100ft. There are charts available to show you, for example [1]. If you're thinking of running normal hobby wire for 12V circuits around a house, the resistance losses are going to be huge.

[1] https://www.boatus.com/expert-advice/expert-advice-archive/2...

2 comments

many post like your's turn around those formulas

  V=RI and W=VI 
combined to give

  WireLossInWatt = RI²
assume a 120W bulb and a 2.5mm diameter 1km long copper conductor (around 7ohm).

  at 120V you need 1A so you lose 1²A² × 7ohms = 7W

  at 12V you need 10A so you lose 10²A² × 7ohms = 700W
That example with a small diameter long cable is a bit extreme but I hope it's illustrative
What? 12AWG Romex is rated for 20A in normal residences in the US and it's already in a lot of homes, so the retrofit would be easy. And for LED lighting you wouldn't need nearly that much current at 24V, so voltage losses would be negligible.
"Rated for 20A" means that putting 20A through the cable will heat it to the maximum rated temperature, which is ~90C for that type of cable.

Heating all of your house wiring to 90C is a) dangerous, and b) expensive. Even if you're doing less than 20A, decreasing the voltage by a factor of 5 still increases energy loss in cabling by a factor of 25.