Wire size substitions

Re: Wire size substitions

You can make the single/parallel wire run based on solar panels (one group of four, 2 groups of 2, 4 groups of one, etc.)... Because each panel (group) is the only source of current for each circuit, you are safe running parallel runs (for what ever reason) to a common destination as there is no issue of improper current sharing.

Just use the Wire Calculator in the sticky and just allocate the current to the wire run (1x4 vs 4x1, etc.).

Personally, I too do not expect parallel wires/contacts to properly share current (from a common source to a common destination). I frequently had to figure out internal wiring for computer systems, and many time it would have been really nice if I could have just used one ribbon cable to bring signal and power to a daughter card of some sort.

From my experience, you can only guarantee current sharing if you have a calibrated resistor (or resistance) in series with each run--that is significantly higher resistance than that of the conact+wire run resistance.

Of course, nobody does that and very frequently I would see systems with (for example) 5 wires/contacts rated at 1 amp per set, running 4 amps, and after a year or three-there would be cascading burned contacts (first one would fail, then heating the next, causing it to fail, etc.)... The problem was that the contacts generally had the highest and most variable resistance--so there was always one contact with a bit higher resistance than the others (but the rest of the contact/wire run averaged low resistance--enough to start heating, but not high enough reduce current flow significantly). So, typically (IMHO) the weak point in paralleling is where the wires join (wire nuts, binding screws, etc.).

My personal design rule was I would limit using parallel wire current increase by a 1/n type series:

1/1 + 1/2 + 1/3 +1/4 +... (first wire capacity + second + third +...)

So, 10 amps for n=1, 15 amps n=2, 18.3 amps n=3, etc...

I would never exceed this ratio and I would almost always way over design parallel connections (try to limit where max current is only slightly over the rating of one circuit in the first place)--and if I could, I would use another solution instead of paralleling in the first place.

-Bill

By the way, the relationship between wire gauge, current, and heating is not a simple/linear relationship... Because:

1. The surface area of wire is proportional to the diameter, but the current conductor is proportional to the square of the diameter (a wire with twice the diameter has 4x the conductor area, but only 2x the surface area--so it can run hotter).

2. Conduit fill affects capacity ratings. A conduit can only safely dissipate so much heat--the more current carrying conductors you stuff into one conduit, the more heat generated--the hotter the wires will run.

3. Where you run the wire/conduit will affect how much heat can be dissipated--Running bare wires through the air--cool well. Buried in dirt, will not cool nearly as well.

Your best bet is to use the NEC (or equivalent for your locality) to calculate safe current carrying (and voltage) capacities for your installation.

-Bill