I looked into how fat the wire going from stunt batteries to motor speed controllers needs to be.  Results are interesting, but not too astonishing. I'd assumed that AWG 20 wire would suffice to carry the current with acceptable resistance loss, and that people are using AWG 12 or 14 wire for some Eli-the-ice-man* reason. However, it appears that inductance (the pertinent Eli-the-ice-man component) is a function of wire length, rather than thickness, and is maybe not a big deal even for length.. 

Is weight reduction from smaller wire worth the resistance penalty?  Ideally, one would cipher airframe change due to weight and other esoteric stuff, but I took an easy way out.  The upper left table in the attachment shows properties of three comparable battery species: 1) five Molicel P50B 2700-size LiIon cells, 2) six Molicel 18650-size LiIon cells, and 3) good old Thunder Power 2800 mAhr six-cell LiPos.  The upper right table shows a trade between the latter two.  Name-brand stunt fliers have decided that batteries of LiIon cells are better than those of LiPos.  To wit, it's worth  .066 ohms of internal resistance to lose 61 grams of battery: .00108 ohm/gram.  The lower right table shows trades between 18-gauge wire and 14-gauge wire and between 20-gauge wire and 14-gauge wire.  From 14-gauge to 18-gauge costs .00111 ohm/gram, so if it's worth changing from the LiPo battery to the LiIon battery, it's worth changing from 14-gauge wire to 18-gauge wire.  Returns diminish for 20-gauge wire.

There are other considerations, e.g.  1) For a given connector, which size wire would require more solder?  2) It's easier to deal with thinner wire in a narrow fuselage.  3) Heat could be a problem with thin wire.  At 20 amps, the average current for my Impact with a six-cell battery on a cool day at 200 ft. altitude, 18-gauge wire dissipates 2.6 watts/foot.  20-gauge wire dissipates 4.1 watts/foot.  Peak current is more than 20 amps, and remembering the nursery rhyme, "Twinkle twinkle little star, power = I squared R," one might worry about melted insulation or charred paint.  Even worse would be on a multi-motor airplane with high-kV motors and fewer cells.

*E = L dI/dt, I = C dE/dt

Electric planes tend to fly better nose heavy.  No idea why but weight in the nose is not the big deal weight in the tail is. I use the same gauge wire that comes with the motor and ESC which, since I am mostly into 28xx and 35xx which means nothing smaller than 14(.). 

Ken

I drew some pictures.  The first one shows the trade between wire mass and resistance.  Remember we are willing to pay .00108 ohms per gram of mass reduction for the change from Thunder Power 6s 2800 batteries to 6s 18650 LiIon batteries.  That's the slope of the red line (-.00108, actually).  That slope is about the same as the ohms per gram of switching between 14 and 18 gauge wire, or the ohms-per-gram slope at 16 gauge wire.  Considering only the mass-resistance trade, if you would switch from Thunder Power 6s 2800 batteries to 6s 18650 LiIon batteries, you wouldn't want to use any wire fatter than AWG 16.  As for thinner wire, you can see the diminishing returns.  Bob Carver, an electrical person, was at the house Saturday.  I asked him about battery wire gauge.  He said he'd use 22 gauge, but he likes to fly Nelson .36-powered combat planes on .012" lines.

The second picture shows how much power goes into heating the wire for various currents.  Heating is proportional to the square of the current.  The third picture shows voltage drop.  This could be a concern toward the end of a flight with LiIon cells if current or battery internal resistance increases.

Please tell me if I missed something or ciphered wrong before I rewire my dogs.

Unless basic electric doctrine has changed recently, you could run higher volts and lower amps and deliver the same watts through a thinner and lighter wire.

I think we do and we can.

22AWG is used as a fusible link rated at 30A, should be interesting. Nobody ever tried to save weight in wire gauge, even back in the NiCd days of electric RC. Competitive electric RC models keep wire lengths to an absolute minimum, just enough to make the connection.

Interesting since our local electric code requires a 15A fuse/breaker on 14 gauge wire.

Ken

*I am not an electrician nor an electrical engineer.*  From many of the things I've read, wire amp ratings are safety factor rated for enclosed spaces with little to no ventilation or cooling.  In our planes, we generally have some reasonable air exchange happening.  We generally use heat resistant silicone jacket wires, but there is probably a best practice for heat related efficiency loss.  Use what makes you feel safe, but it seems like we might have a tendency to over spec.  The attached silicone wire manufacturer chart is certainly an interesting read for max current ratings. 

Having flown electric since '97 or so I've selected quite a few power systems and well beyond. I can't recall ever being concerned about wire size supplied by an established mfg ever being excessive or inadequate. I always kept things as short as possible, and almost never used a connector between the ESC and motor.

When comparing batteries, I always look at watthours. I don't understand why 5S p50B would be compared to 6S P30B when 7 still weigh less the the P50B pack.

Fusible links aren't really rated like fuses. However, I want video if an attempt is made to fly a 30A setup on 22AWG wire. If my power supply was in the garage, I'd demonstrate it myself. I burned up a Toolkit M6 charger yesterday, did that smell. Don't push these to the limit.

Keeping wires short is more important than wire gage.

Can you show me quantitatively?

The great automotive engineer, Charles "Boss" Kettering invented, along with many other things, the electric starter.  Previous engineers and inventers concluded that (like with model engines today) the electric starting system needed to be bigger and heavier than the car engine itself.

Kettering made a smaller and lighter starter based on the fact that you could overload an electric circuit for a short time but not all day long.  Likewise, a modeler can eliminate the safety factor in order to gain advantage while only risking a model airplane.

I'd exhaust all means of weight saving prior to reducing wire size.

I did, except for paint, of course.  Nontautening dope shrinkage crushed the hollowed four lb/cubic ft ribs on my current dog.  That got me to wondering about all that copper in the nose.

John Hart-Smith told me that his previous company, a now-extinct California airplane maker, was willing to increase primary structure weight by a pound to save $50, while paying $1000 per lb for weight reduction elsewhere in the airplane. (I forget the exact numbers, but those are close.)

While I was typing this, Amazon delivered my 18-gauge wire.

Using a small connector with the small wire might double that.

If connectors are bad, why don't they get hot?

I am switching from Deans to Spektrum IC3 connectors, but for reliability, rather than resistance.  Igor recommends something-Jet connectors, which have super low resistance and are made of good stuff, but I haven't seen any that come in pairs. 

Crist, I think you need some trip strips and vortex generators to give you the aerodynamic oomph you need to carry that extra weight.

Touche'

The lightest connectors I've ever seen were Kontronic, I believe solid silver even. When it matters, I use a bullet type with heat shrink.

Even double of a very low value might be acceptable, just the point you're making. I've measured connector resistance and equated it to wire lengths. If optimizing to this point, you'll have to make the call on mass vs. loss. This application is not high current or power.

That is interesting.  I'll keep the fat wire on the ESC.  I imagine aluminum heat sinks on the ESC wire, being careful to stagger the fins to avoid shorts. 

One should keep wires short.  With short wires the wire gauge will make very little difference in weight.  I know a gram here and a gram there adds up.  But I personally thing there is better places to save weight. Reliability Just saying  -----