A ductless mini split unit I am interested in specifies a maximum breaker size of 15 amps. Why would this be the case?

The unit is going to be installed on the opposite side of the house from where the breaker is, so for a 220v system that runs at about 5 amps, a 70 foot run of electrical wire calls for 8 gauge wire.

I was going to throw a 20 amp circuit breaker for this run until I came across the spec sheet for the unit. I found a similar post where the top answer states that you can always use bigger wire which only reassures me that 8 gauge wire is fine in a 15 amp breaker, but I can't seem to find why the AC unit would call for no bigger than a 15 amp breaker.

In other words, why would a device on a circuit be sensitive to the potential max amperage of its circuit

  • 1
    What's the exact wording (or a screenshot) from the manual, regarding the max breaker amps? Feb 18, 2021 at 1:21
  • I haven't gotten through the manual yet but the overview on the webpage says MAX BREAKER SIZE 15 amps MIN. BREAKER SIZE 9.5 amps
    – deefo
    Feb 18, 2021 at 1:34
  • 1
    That chart keeps popping up. It's for 12/24V DC systems, not 120/240V AC. The original source is bluesea.com/resources/1437
    – nobody
    Feb 18, 2021 at 2:39
  • 5
    NO NO NO DON'T USE THAT CHART! That chart is for 12 volt DC systems. Voltage drop @ amperage is 20 times worse than 240V). I don't even bother looking up the voltage drop on 240V systems until it's 180' or farther, and even that assumes max current. You could probably go 400' with #12 wire. Google was not kind to you that day, I hope you didn't buy the wire yet! (Harper's Law: Buy the wire LAST.) Feb 18, 2021 at 18:31
  • No purchase yet, but it sounds like im going the way of 12 awg on a 15 amp circuit. I appreciate the insight
    – deefo
    Feb 18, 2021 at 19:06

3 Answers 3


The unit won't be protected properly by a larger breaker

While hermetic compressors and fan motors are protected against overload by internal means, or by their drive electronics if you have an inverter-based system, the wiring of the unit still needs to be protected in case it shorts out internally. As a result of this and the fact that air conditioners universally are complex, multimotor loads, UL has taken pity on installers and inspectors and requires the manufacturer to do the breaker sizing math for us. (Otherwise, you'd have to dig into Articles 430 and 440 of the NEC and size the breaker yourself, which is a bit fiddly for a multimotor load.)

So, simply use the 15A breaker the manufacturer calls for and call it a day, although I doubt you need 8AWG wire for a unit that draws less than 10A at 240VAC. In fact, for that run, 14AWG is perfectly fine. (The chart you found is probably for low-voltage DC circuits, where voltage drop is a far greater limiting factor than at 240V.) However, there is some merit to future-proofing an air conditioner circuit, and to that end, I'd run ½" ENT as that'll get you up to 50A with THHN wire and the assumption that your panel supports 75°C terminations, then pull the 3 14AWG wires needed through it.

  • 2
    Though running 12 AWG instead of 14 AWG would provide some future-proofing at relatively low cost in case this is ever upgraded to a 20A circuit for a bigger unit. Feb 18, 2021 at 3:03
  • 3
    @manassehkatz-Moving2Codidact -- yeah, 1/2" ENT is a much more cost-effective futureproofing measure though, I reckon -- with a reasonably modern panel, that gets you all the way up to 50A Feb 18, 2021 at 3:10
  • I agree with manassehkatz, but would add I have been asked to twice to add a second minisplit, and in that case since mains power is usually fed to the blower, then regulated power from blower to outdoor unit larger wire is mostly unhelpful. Feb 18, 2021 at 17:58
  • I see. I realized the charts I was referencing are for low voltage circuits. I found this website(paigewire.com/pumpWireCalc.aspx) which suggests an 80 foot run at 240 volts/15 amps can be sized at 14 awg copper. I'm not certain I understand the purpose of running multiple wires through the conduit or why to use the conduit at all, but thanks for reassuring me by answering my initial inquiry.
    – deefo
    Feb 18, 2021 at 18:25
  • Yep. To extend the discussion, most consumer equipment protects itself internally over the circuit's capacity. A consumer router that needs 500mA of current will probably have a 1A fuse inside because the circuit will easily supply more than can damage the device. A 15A breaker would feed 10A into a faulting router indefinitely (starting a fire, etc), so it must fuse to stop the overload because the breaker only protects the wire. Motors (like an AC compressor) generally rely on the breaker or a thermal overload device to protect not just the wire, but also the equipment.
    – J...
    Feb 18, 2021 at 19:44

It's to limit the time it takes for the breaker to trip in the case a failure less than a dead short. The time curve for all breakers aren't identical, but suppose the unit fails and it draws 30A, which is 5x rated current. A 15A thermal-magnetic breaker that trips on an inverse time curve (see below) should trip in less than a minute, a 20A breaker could hold for 5 minutes. That heat could be the difference between replacing components and replacing the whole thing.

The unit itself may also have some internal protection less than 15A, but 15A is the smallest size the NEC recognizes in electrical panels.

I've seen the chart you referenced above, I think it still had the source, but it is completely inaccurate for AC circuits and modern wire insulation. I suggest you consult some online voltage drop calculators (1 2 3) to see the actual loss, less than 5% is good, 3% is better.

enter image description here

  • 1
    "That heat could be the difference between replacing components and replacing the whole thing." or replacing the whole house!
    – FreeMan
    Feb 19, 2021 at 19:26
  • If I read that chart correctly, the breaker will allow up to 7 times the rated current for up to 1.5 seconds in the "not tripped" state, then enter the "unknown status" (during which it might trip, or might not) for about another 5.5 seconds (to 7 seconds total time), then will be guaranteed to be "tripped". Additionally, it could take a spike of up to 700 times its rated capacity for up to 15 thousandths of a second before being guaranteed to trip. Is that correct?
    – FreeMan
    Feb 19, 2021 at 19:30
  • That's how I read it. The 700x reflects tripping up to the AIC rating within one full cycle. The 7x seems to explain why motors with overloads are allowed 250% rated breakers to handle inrush that reportedly can exceed 20x rated motor nameplate. Feb 20, 2021 at 2:44
  • Wow. That's quite a range and is totally not what I would expect. Granted, 0.015 seconds really isn't very long in the grand scheme of things, but 140 amps for 7 seconds seems like it could allow a fair bit of heat damage to the wiring. I would guess that it's within spec for what the wire is designed to handle, though. </off topic discussion>. Thanks for the info!
    – FreeMan
    Feb 20, 2021 at 14:32

Probably you got 1 hp aircon,at Nec table it is 8 amps. Multiplied by 1.75 equals 14 amps so ,15 At CB make sense. For wire 8 x 1.25 equals 10 amps only so #12 wire can be used.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.