In Adjustable-ac-current-limiter-for-output-of-variac-for-safe-circuit-testing, I asked about alternatives to using incandescent light bulbs to make an adjustable A.C. current limiter. The challenge with incandescents is that they are relatively large, fragile, and increasingly obsolete. A variable current limiter made with multiple bulbs would be particularly large and fragile.

It doesn't seem like there are any simple solid state circuits.

I came across polymeric positive temperature coefficient (PPTC) fuses which seem to have similar (but better - i.e., steeper) positive coefficient resistance changes with temperature.

Specifically, using a rotary switch and a selection of PPTC fuses ranging from say 30mA to 5A, I could make a more compact, less fragile, and potentially even better behaving A.C. current limiter.

For higher value PPTC fuses, I could even put a couple of same value fuses in parallel to achieve higher current limits (indeed, when the fuses are in parallel, if anything the cutoff should be more responsive since once one device starts shutting off power, the current to the other parallel devices will increase, thereby accelerating them shutting off too.)

I could also add a (red) LED plus resistor across the PPTC fuse bank to signal when the PPTC fuse has cut off the current (as the voltage will rise across the open fuse.) If I end up using variable input AC voltage, I could add a small Zener circuit to feed the resistor with constant current.

For extra protection, I might add a single traditional fuse (rated slightly higher than the highest PPTC fuse) in series. Plus a GFCI for the load socket.

To finish up the device, I would add an ON-OFF-ON 3-way switch that would allow me to switch between

  • PPTC current limiting
  • Off
  • Bypass of PPTC current limiting fuses

Would this approach work?


1 Answer 1


Not really. PPTCs aren't available in more than modest power levels; and they're huge, i.e. expensive and slow. Note that a PTC fuse takes longer to clear than a fast conventional type, and they have much lower voltage and current limits. In particular, you aren't going to find a PPTC for 100s of A of fault current, let alone the 1000s that mains is capable of delivering.

There are high voltage PTCs, made of ceramics, analogous (but complementary) to NTCs; but they don't have as sharp a curve as PPTCs, so may not be as handy. They're mostly in smaller sizes, I think? (I haven't been shopping for them in a long time.) The thermal time constant hits again, of course.

As for conventional means, a current sensor into a comparator (and latch) and relay does well enough for some applications; the relay may open before a conventional fuse opens (~10ms), or a slow blow might be used. A small current-limiting resistance might be used to reduce fault current and thus contact wear, since it'll be opening under load. These are also available as entirely electromechanical units (current limiting relays), and circuit breakers are basically the same thing at a different setpoint (thermal-magnetic types have a bimetallic strip to open at low currents and long time scales, and a small coil to pop it open under fault currents).

Relays aren't very good because, in those 10ms or so, they let through full fault current; anything that's going to blow, has already, and you're just saving fuses or wiring.

Semiconductors aren't very good because it takes a ton of energy handling to switch mains voltage -- so, big devices; even if you don't have to deal with surge ratings. A switching current limiter / e-fuse would be fine, if a bit on the expensive side, but, as you say -- a complex solution as well.

Mind that many solutions have the downside of not handling inrush, if that should be a problem. A starter circuit may be helpful as well.

  • \$\begingroup\$ I'm looking to use this for electronic device repair. Typically several hundred milliamp limit at 110V, max maybe of 4Amp. I could use a 9-position rotary switch to get 30mA-60mA-120mA-250mA-500mA-750mA-1A-2A-4A. PPTC's are available @250V for <$1 for all such current values (except that for 4A I would need to use 2 2A in parallel). They are all pretty small and cheap. I'm not sure they are any slower than what it takes to light an incandescent -- they are really just both heating wire or equivalent. \$\endgroup\$
    – JK1
    Apr 3, 2023 at 7:43
  • \$\begingroup\$ Also, I don't need to worry about 100's of Amps of Mains as I will be on a 15-20A branch circuit. Plus, as I mentioned, I plan to lead with a traditional 5A fuse in series. Unless I am missing something, I really don't see how your objections are relevant to my use case and proposed solution. So, please explain what I am misunderstanding here... \$\endgroup\$
    – JK1
    Apr 3, 2023 at 7:45
  • \$\begingroup\$ I'm not sure you're aware how much current mains can supply under fault conditions--? The source impedance is a LOT less than (120V/15A). Fault current of 1kA is typical. Shockingly, I do see some parts exist (but aren't in stock) for 250VAC, such as Yageo BK250-2000-SZ. But they're 26mm dia., take max. 28s to open, and the Imax of 10A means even a 15A load most likely blows it out (given the size, probably at significant output of magic smoke). You still need current limiting resistance to use it reliably. \$\endgroup\$ Apr 3, 2023 at 8:40
  • \$\begingroup\$ Thanks for the additional clarification. Again, I am not worried by the possibility of high fault currents as I can protect from that by preceding with a standard fuse or circuit breaker. The size of 26mm is still much smaller than an incandescent (even more so given the shape). HOWEVER, the max speed to blow of multiple seconds would seem to rule out such solutions. Thanks. \$\endgroup\$
    – JK1
    Apr 3, 2023 at 13:00

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.