Metal oxide varistors catch fire in surge protection devices

Posted on Monday, December 24th, 2012 in components by DP


Arhi tipped us off to a potential problem in cheap surge protection devices. The metal oxide varistors (MOVs) found in surge protectors are prone to catching fire in cases of extended current surges.

MOVs are designed to suck up short surges that might occur on power lines. Unfortunately, in case of prolonged surges MOVs break down, usually causing a short circuit. This can lead to heat and possibly fire as shown above.

This entry was posted on Monday, December 24th, 2012 at 11:00 am and is filed under components. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.

6 Responses to “Metal oxide varistors catch fire in surge protection devices”

  1. Anthony says:

    Don’t they have some kind of breaker or fuse to blow when the mov becomes shorted
    I never saw one without it

  2. Sleepwalker3 says:

    Yeah, if you read my comments in the forum it might explain a bit more. I suspect it’s not actually a MOV that’s gone up too. Often the switch is also a thermal circuit breaker, but they aren’t really suitable for surges that come on quickly and in any case, that could be just a plain switch.

  3. Niklas says:

    A fuse in series with the live wire before the MOV is good practise on PCBs. With some electrical systems you might also need a fuse in series with the neutral. Also cover the MOV with a heatshrink tube to keep all the burning parts from spreading in case the MOV explodes.

  4. arhi says:

    The one on the picture that burned down as some others on pictures in the forum (original thread is in Serbo-Croatian language here: ) had a glass fuse and the glass fuse remained un-tripped so the MOV started burning before fuse broke.

    There’s number of videos and txt’s here too:

    Here’s video on how the 110V “protection coord with mov” behave when you put 220V trough it (it also have a fuse that does not blow):

  5. Sleepwalker3 says:

    Arhi, are you sure the two devices on the right were MOV’s and not Y1 capacitors?

    Unfortunately those powerboards are often fitted out with very slow reacting thermal breakers and if you look at the trip characteristics a typical 10A unit, can pass 100A for a couple of seconds, 60A for maybe 8 seconds, etc., so are really too slow for that type of application. The problem is, that this has to be balanced against nuisance tripping (or fuse blowing) with a faster breaker or fuse. Typically fast breakers cost a lot more than cheap thermal ones too.

    Still, if the fuse hasn’t blown when the MOV clamps solidly, then clearly it’s unsuitable and shouldn’t be put on the market, but sadly, many are.

    The video with the guy demonstrating his product against the others, he termed his something like a ‘Line Conditioner’, which is not correct, a proper Line Conditioner such as an Industrial Constant Voltage transformer is a completely different thing. The device he was demonstrating was really an Under Voltage / Over Voltage relay – Effective, but it doesn’t do any ‘Line Conditioning’ as such, simply turns the relay off if the voltage is out of spec.

  6. Niklas says:

    Just as Sleepwalker3 wrote, the selection of fuses can be tricky. Their characteristics are not linear, but instead logaritmic. Usually they need 10 to 100 times more current than the marked one to break very fast. Perhaps a combination of a MOV and a TVS, with a more abrupt current knee, can break the fuse faster if their voltage ratings are paired correctly.
    As an electronics designer you must also take into account the current peaks during the initial connection. X and Y capacitors and also the bulk storage capacitor for switchmode power supplies can draw a lot of current and must not trip the fuse. Dont underestimate what can happen when you have high frequency noise on the power line. Filter capacitors might then be almost zero ohms.
    Connecting a 110 Volt rated power cord to a 230 Volt net is not very clever, even if it is fused. The creepage, clearance and also the overvoltage ratings are totally different. As appliances powered by 110 Volts draw more current, to get the same power, the fuse must have a higher current rating compared to one that is rated 230 Volts. Also the fuse itself might not be ok to use since it might not break properly with the higher voltage across it.

Leave a Reply

Notify me of followup comments via e-mail. You can also subscribe without commenting.

Recent Comments

  • KH: Well, I'm sure the HDD can handle a little vibration *cough*. What I am more interested in is this: the fan appears to be placed...
  • Alan: The Si5351 runs on 3.3V, and several FPGA dev boards now include "Arduino compatible" headers [if you stick to 3.3V logic levels]. Maybe it's time...
  • hli: Sunday++
  • JB: Hi
  • Peter: Sunday