Surge Suppressor MOV Failures

Status
Not open for further replies.
L

Longroof

Member
I'll start off by saying that I work in a lab that has the sole purpose of identifying root causes of failed electrical components/systems. I am not an electrician by trade, so forgive my lack of knowledge of that side of things. I do have a EE degree so I'm not without some book-learnin' but my experience is not in field work.

A while back we had a facility that experienced the failure of several surge suppressors (power strips). These are the standard, 6ft cord, multiple plug outlets. (WABER/TRIPP-LITE) At the time of failure, three power strips were daisy chained together and a single laptop was plugged into the final strip. All three strips were burned identically. In each strip is a small circuit board with three MOVs and a couple thermal fuses. In all the strips, the same MOV (#2) was burned (melted and in some cases blown open). Not sure how to easily put up a schematic I have (PDF or Power Point) but it's the MOV between H and G.

Since the three power strips were all connected, it's not surprising they all had the same fault. And with no excessive load on them (1 laptop, still working), we figured it had to be a spike on that circuit somehow.

We checked all the other plugs on that circuit and found no damaged equipment.

On all the plugs we were measuring (AC)
H-G ~ 119
H-N ~ 116
N-G ~ 2.7

At one plug in the building we got almost 8V N-G.

Then we took measurements at the feeder panel... Very similar with the same ~2.5 VAC N-G But switching to DC it then reads 13.1V.

Obviously this didn't sound right to me, but it didn't explain my surge suppressor issue and the electrical crew was going to look into that N-G voltage. (I have yet to hear from them.)

Fast forward a few months and I get another call. Same building, same issue. Three more strips, each plugged in separate plugs, with no load plugged into them, all fried the same way.

No other facility has had issues. I'll try to track down the details on the supply transformers but wanted to see if I could get some thoughts.
 
Could be an intermittent neutral?
If neutral opens it is possible for the H-G voltage to rise above the level of the MOV clamp point.
 
I am thinking the 13VDC is adding to either the top or the bottom of the sine wave.

119VAC + 13VDC = 132 Volts across the MOV every half cycle.

A device that pre-dates switching power supplies, like an old overhead projector could be causing the DC bias. If someone saw the strips burn, then you may be able to determine what everyone was using at that time. It may take a specific combination of devices on together to cause the excess voltage across the MOV.

8volts N-G seems excessive to me. I was taught 3volts was the limit.
 
8 volts:

High levels of 3rd harmonic flowing back on the neutral,will cause a high IR drop betwwen neutral and ground.We have measured over 12 volts at some sites.

MOV Failures:

Sporadic electrical arcing between hot and neutral/ground somewhere in this electrical system will cause these "cheap" MOV's to fail everytime.Would recommend installing a high speed power monitor(1 microsecond response time) to track down this electrical arcing.
 
ron said:
What was MOV#2 conencted across? L-N, N-G or something else? or were all three MOV's connected in parallel across L-N or L-G?
Click to expand...
That would be my question. Which Mode was fried. My guess is the N-G Mode.

In a good design a term strip will have 3 modes of protection L-N, L-G, and N-G. Going one step further in a good design the L-G and L-N modes will use around a 300 volt MOV of equal joule capacity. But the N-G mode will be a much lower voltage value as 300 would be worthless. Typically down arounf 20 to 40 volts with the same joule capacity as the L-G and L-N modes.

My guess is you have something wrong in the service entrance or the neutral circuit in the facility because you should not see any difference of voltage between N-G at or near the service entrance. Downstream you will see some difference that varies with the load current on the neutral circuit, but 8 volts is just way to high even with a heavy load current. If designed properly the highest you should see is 2 to 4 volts way down stream from the main on a heavily loaded end branch circuit.
 
Last edited:
Can I send someone a PDF or Power point file to post to show the schematic? It might clear things up as far as which MOVs are damaged.

All three MOVs are the same, (P/N CNR-20D201K) and are listed with max allowable voltage of 130 (AC rms); 170 (DC); and a clamping voltage at 340.
Also on the data sheet it says the varistor voltage at 0.1ma is 185-225 (Min-max). Joule rating is 95 and a 1 W power rating.

Why would a 300 voltage rating for the N-G MOV be worthless?

I agree that the 8V N-G is excessive, but it seems that this is constant, but I'm not sure I understand how this would burn the MOVs. I know that they do have a life/cycle limit, but it seems unlikely that they would all go at exactly the same time.
 
Are these plug strips on a multiwire balanced circuit?
If so then the upstream neutral could be intermittently open circuiting causing a voltage divider amoung the balanced circuits.

This type of failure is discussed here often. Depending upon the respective circuits loads the MOV voltage could be exceeded. In this case they would all be destroyed because the available energy is very great compared to their normal intended purpose of suppressing a transient.
 
Longroof said:
Why would a 300 voltage rating for the N-G MOV be worthless?
Click to expand...
Because much of the sensitive electronic equipment, especially those with switch-mode power supplies have passive circuits installed N-G like RFI filters and MOV's which are rated at low voltages of 30 volts or less. Does not do the equipment any good if you let a 300 volt transient through. It lets the magic smoke out. :grin:

I would start and look for very simple wiring faults first like:

Reversed neutral and ground wires. To check for reversed neutral and ground wires, measure the hot-to-neutral and hot-to-ground voltages under load. The hot-to-ground reading should be higher than the hot-to-neutral reading. The greater the load, the more difference you'll see.

8 volts between N-G is out of any reasonable expectation. If that is the case caused from excessive load or circuit distance would mean the L-N voltage would have to be 104 volts on a 120 volt circuit. I bet you your L-N voltage is not that low.
 
Last edited:
I have never seen a (with a 120VAC or greater input rating) piece of equipment with a component from neutral to ground rated at 30 V?

All the equipments I have worked on used the same voltage MOV that it is connected between Hot to Ground for the connection from Neutral to Ground.

Not saying that you could not theoretically use a lower voltage MOV from Neutral to Ground in a bonded system, it just has not been my experience.

Any capacitor that is part of an EMI filter is rated at higher than line voltage and must be specifically rated for that duty. This is whether it is located from Hot to Ground or Neutral to Ground. They are usually the very same component value ratings.

Another good reason that the Neutral to Ground MOV is rated the same as the Hot to ground is that so many SMPS now a days are Universal power.
 
Status
Not open for further replies.
Top