What determines battery cable size?

enblethen wrote:
You must consider the length of the conductors and their resistance.
The positive would be copper with more resistance then a welded frame. The welded frame will have no voltage drop that you would need to consider into the calculations.
Blue Sea is most likely figuring a boat where both the positive and negative are using copper conductors.

The frame isn't quite that good, but it is usually lower R than the long pos wire run.

(To answer a question earlier, you don't need the pos and neg paths to be of equal R, but they both "count." It is the total R of them both making the circuit that matters. Any reduction in R anywhere in the circuit helps.)

I had a thread about this a while back, where I was using Salvo's method of measuring R on a path, and posted a bunch of different results from things I tried. Don't have my notes anymore on that, but it must all be in the archive somewhere.

First I tried to reduce the R on the neg side (mostly frame) by adding a second neg path using a 20 ft length of copper water pipe with wires at each end bolted on to connect with. That knocked the neg R way down, and made the amps from the converter to the batteries go up quite a bit.

So next I ran a 25 ft #1AWG cu-al jumper cable set from the battery lugs on the fuse panel by the converter inside, out the door, and up to the battery bank and clamped on. This really knocked down the R on the positive side!

With that all set up for both paths, I was able to get 56 amps from my 7355 13.8v converter to the batteries (as seen on the Trimetric)

So yes, you can make improvements and measure the results so you don't overdo it-- On that, once you get the R down to where the converter is doing its max amps, there is no point reducing the R further, since the converter can't do any more amps than its max.

As it happens, I didn't take advantage of all that! I got different chargers mounted up front close to he batteries instead. The 7355 is still there as before.