Wire size and fuses: why voltage drop decides your cable, not amps

Most people size camper wire the way they size house wire: look at the amps, pick something that looks stout enough, move on. That works at 230 V because the supply voltage is enormous compared to the losses. At 12 V it does not work at all. A 3 % voltage-drop budget on a 12 V system is 0.36 V — about a third of one volt to spend on the whole out-and-back journey of the current. On van-length runs you will run out of that budget long before you run out of the wire's ability to carry the current. Length, not amps, sizes almost every conductor in a camper.

The formula, in words

The cross-section you need to stay inside a drop budget is:

A = ρ × L × I ÷ V_drop — where A is conductor cross-section in mm², I is the continuous current in amps, V_drop is the volts you are willing to lose, and ρ is copper resistivity. Use 0.0178 Ω·mm²/m, the warm stranded-copper figure, not the 0.0172 lab number for annealed copper at 20 °C — your cable lives in a hot cabinet, not a laboratory.

L is the round-trip length. This is the single most common arithmetic error in van wiring. Current goes out on the positive and comes back on the negative, and both conductors drop voltage. A fan three metres from the fuse block has a six-metre circuit. Forget the return leg and you halve the area you calculate, which typically picks a wire two sizes too small and doubles the drop you were budgeting for. Measure the path the cable actually takes — up the rib, along the roof, back down to the panel — and then double it. It is also a reason to run a real insulated negative back to a busbar rather than grounding to the chassis wherever convenient: with a chassis return you no longer know what L is.

Two limits, and the thicker one wins

Every DC run has to satisfy two independent tests, and you take whichever demands more copper.

Ampacity is the current the conductor can carry without cooking its insulation. The figures used here are ABYC E-11 allowable amperage for 105 °C insulation, outside engine spaces, with no more than three bundled conductors — 35 A for 14 AWG, 45 A for 12 AWG, 120 A for 6 AWG. Those numbers look absurdly generous next to household wiring intuition, and that is exactly the point: ampacity almost never governs a small camper circuit. Bundling wires tightly, running them through an engine bay, or burying them in insulation all derate those figures, so treat them as a ceiling for a well-ventilated run.

Voltage drop is the other test, and on lighting, pump, fan, fridge and charger runs it is the one that decides. Take a 5 A roof fan with a 20 ft (6.1 m) round trip at a 3 % budget: A = 0.0178 × 6.1 × 5 ÷ 0.36 = 1.51 mm². The next real size up is 14 AWG (2.08 mm²), which lands at 0.26 V or 2.2 % drop. Ampacity alone would have accepted 18 AWG — rated 20 A, four times the load. Drop drove the decision, and it drove it two sizes thicker. Drop the return leg from the sum and you would have specified that 18 AWG and lived with 0.66 V, a 5.5 % loss.

Longer runs get worse fast, because the relationship is linear in length. A 3 A light circuit on a 50 ft (15.2 m) round trip needs 0.0178 × 15.2 × 3 ÷ 0.36 = 2.26 mm², so 12 AWG (3.31 mm²) — thicker cable for 3 A than the fan needed for 5 A, purely because the wire is longer. Nothing about the load changed.

The exception is the short, brutal run. A 1500 W inverter running flat out pulls roughly 125 A from a charged 12 V bank (1500 W ÷ 0.9 efficiency ÷ 13.3 V), and it sits close to the battery — call it 0.8 m each way, 1.6 m round trip. Drop asks for only 4.94 mm², which 8 AWG would satisfy. Ampacity asks for 4 AWG (21.2 mm², 160 A), and ampacity wins. Here the current genuinely sets the size — and the resulting drop is a trivial 0.17 V, 1.4 %. If your inverter's manual specifies a larger cable or a particular fuse, follow the manual: it knows its own surge behaviour better than a general table does.

3 % or 10 %: pick the budget before you pick the wire

The drop budget is a design choice, and it changes everything downstream. Use 3 % for anything that matters — battery feeds, DC-DC and solar charger runs, the inverter, the fridge, the water pump, panel feeds. Use 10 % only for genuinely non-critical loads, which in practice means cabin lighting and little else. That 3 A light run needed 12 AWG at 3 %; at 10 % the budget is 1.2 V, the requirement collapses to 0.68 mm², and 18 AWG passes — three sizes smaller and a fraction of the cost. LEDs genuinely do not care. A compressor fridge does.

24 V changes the arithmetic twice over. The same watts draw half the current, and 3 % of 24 V is 0.72 V rather than 0.36 V — double the budget. Both terms move in your favour, so the required cross-section falls to one quarter. That 125 A inverter feed becomes about 63 A, and its drop requirement falls from 4.94 mm² to 1.24 mm². It is why big builds — long vans, big inverters, heavy alternator charging — keep drifting to 24 V. You are buying copper back.

A table of real circuits

All at 12 V, 3 % drop budget, round-trip lengths, warm copper:

Read the last column. Five of six circuits were sized by length, not by load. The only one the current decided is the short fat one — and note the fridge at 2.9 %, sitting 20 cm of extra routing away from needing the next size up. That is how tight 0.36 V is.

Fuses: the fuse protects the wire, not the appliance

Sizing is mechanical. Take 1.2 × the continuous load, round up to a rating that is actually sold — 5, 7.5, 10, 15, 20, 25, 30, 40, 50, 60, 80, 100, 125, 150, 175, 200 A and so on — and check the result against the cable's ampacity. The 7 A pump: 8.4 rounds up to a 10 A fuse, well under 12 AWG's 45 A. The 125 A inverter: 1.2 × 125 lands exactly on a stocked 150 A fuse, under 4 AWG's 160 A. Never specify a fuse above the ampacity of the wire it sits on. If the 1.2× number lands above the wire's rating, you do not fit a bigger fuse — you fit a bigger wire. Push that inverter to 1800 W and the draw goes to about 150 A; 1.2× is 180, the nearest stocked fuse is 200 A, and 200 is above 4 AWG's 160 A. The answer is 2 AWG (210 A), not a 150 A fuse you hope never sees 180.

The sentence worth memorising: the fuse protects the wire, not the appliance. Your fridge, inverter and charger all have their own internal protection; nothing you put in the fuse block will save them. What the fuse is there for is to open the circuit before a short or an overload turns your cable into a heating element.

Which makes the dangerous combination obvious: a big fuse on a thin wire. Put a 30 A fuse on 18 AWG and the wire can sit at 29 A indefinitely — nearly 50 % over its rating — without the fuse ever noticing. The insulation softens, then chars, then the conductors touch. Builders arrive at this honestly: a circuit keeps blowing fuses, a bigger fuse is fitted, the symptom goes away. The symptom was the protection working. If a fuse keeps blowing, the load is wrong or the wire is wrong; the fuse is the only part of the system telling you the truth.

One placement rule that people skip: the main fuse belongs within about 180 mm (7 in) of the battery's positive terminal — ABYC E-11 puts it at 7 in, extended to 40 in (1.02 m) where the conductor runs inside a sheath or conduit. The reasoning is that the stretch of cable between the battery post and the fuse is unprotected by definition, and a battery bank can push thousands of amps into a dead short. Keep that stretch as short as you physically can. Every branch circuit gets its own fuse at the busbar too, sized to its wire.

What undersized cable actually does to you

Nothing dramatic happens on day one, which is why this mistake survives. It arrives later, as a list of unrelated-looking faults.

Voltage-starved motors. A compressor fridge or a diaphragm pump fed through too much resistance sees a lower voltage at its terminals, draws more current to make the same mechanical work, and runs hotter for it. It will still work. It will just die a couple of years early, and it will tip your battery's low-voltage cutout into tripping on hot afternoons for no visible reason. Water pumps are the loudest complainer: a pump that will not prime, or that runs but never builds enough pressure to satisfy its pressure switch, is very often a wiring problem rather than a plumbing one.

Chargers that undercharge. A solar controller regulates to the voltage it measures at its own terminals. Put drop between the controller and the battery and the battery sits below the target the whole time — the controller believes absorption is complete and tapers off while the bank is still short. You get a system that never quite reaches full, and a battery monitor that slowly drifts out of calibration. This is a measurement problem, not a power problem, which is why sizing controller-to-battery cable at 3 % (or wiring the sense leads your controller offers) matters more than the conductor length suggests. If you are still deciding what to charge and how much of it, battery and solar sizing is the step before this one.

Heat where you cannot see it. The bad case. An undersized conductor that also happens to be overfused, run through a wall cavity, packed in wool insulation and panelled over. No airflow, no inspection, no smell until there is a lot of it. This is why wiring belongs early in the build order — behind insulation and panelling, cable is effectively permanent, and “I'll upgrade that run later” means tearing out a wall.

It is worth saying plainly that this is not an obscure failure mode. If you read the well-known van electrical write-ups — FarOutRide's, Explorist.life's — the same correction keeps coming back: the cable is too thin for the distance. It is the most common electrical mistake in DIY conversions, and it is entirely avoidable with one multiplication.

Measure the run, don't estimate it

Every number above depends on one input that most builders guess: the length. “About three metres” from the fuse block to the roof fan is really 3.4 m up the rib and across, 6.8 m round trip — and that is the difference between comfortably inside budget and quietly outside it. This is where a layout you can measure earns its keep: VanPlot routes real cable runs in 3D through the actual van shell, around the ribs and behind the furniture you have already placed, so the run length is a measurement rather than an estimate — and it changes when you move the battery box, which is exactly when it should.

If you just want the arithmetic done, the wire size calculator is free and needs no account: put in amps (or watts), one-way length, system voltage and your drop budget, and it doubles the length for you, checks both limits, tells you which one governed, and hands you a fuse size that the wire can actually carry.