How to test a varistor?

Thread Starter

xchcui

Joined May 12, 2014
272
Hi.

A quick way to check if the varistor is okay is to test with DMM
if it has infinite resistance.
But this test can't tell me if the varistor meet the requirements,
like if the clamping voltage of the varistor at 1mA DC
meets the value which the varistor is rated.
How can i test that without a special laboratory instruments?
Maybe with a megger in series with a resistor...?
Thanks
 

Thread Starter

xchcui

Joined May 12, 2014
272
You use a DC power supply to apply 1mA through the varistor, and then measure the voltage drop across the varistor.
I would like to take for example the 14D431R varistor.
When you said to use a DC power supply,what value of DC voltage source should i use?
Do i need a special power supply source?and how can i limit the current to 1mA if i don't have a special power source with current limit?
Or if you have a suitable signal source you can try this. A curve tracer would do an even better job.
View attachment 276548
As i mentioned before,unfortunately,i don't have a laboratory
instruments.I just have DMM,capacitor meter and megger.
May you suggest any other simple method?
 

Jerry-Hat-Trick

Joined Aug 31, 2022
544
I would like to take for example the 14D431R varistor.
If I understand correctly, this varistor is nominally 430V +/- 10%. They are a bit like back to back zener diodes with a very high voltage - they will conduct in both directions when that voltage is exceeded.

Suppose you use a 380VAC transformer like https://www.ebay.co.uk/itm/174573531668 where you can select voltage inputs and outputs? I'd suggest a 220 primary with maybe a 12V secondary connected to the 12V secondary of a transformer with 380V primary. 220V direct to 380V in one transformer is a bit scary and it doesn't allow you to limit the current easily. The peak voltage from 380VAC is 537V so the top and bottom of the sine wave will be shaved off. A resistor in the 12V line would drop that voltage when the MOV starts to conduct so current is limited. If I were to try this, I'd make sure I stood well clear! If you use a potential divider (with a string of resistors) to reduce the high voltage, full wave rectify and smoothe with a capacitor you will have a way to measure the MOV strike voltage with a DMM.

But don't try this at home! Seriously, please don't be near anything when you fire it up. I'd welcome suggestions for improvement or even strong criticism - messing with anything more than 50V is scary.
 

DickCappels

Joined Aug 21, 2008
10,153
I would like to take for example the 14D431R varistor.
When you said to use a DC power supply,what value of DC voltage source should i use?
Do i need a special power supply source?and how can i limit the current to 1mA if i don't have a special power source with current limit?

As i mentioned before,unfortunately,i don't have a laboratory
instruments.I just have DMM,capacitor meter and megger.
May you suggest any other simple method?
If you don't have a scope use one of the suggestions in this thread that use a power supply. Personally I recommend that you use a transformer to isolate your experiment from the line. It will still be dangerous (keep one hand in you back pocket) but it is a little less likely to hurt you and prevents current from returning through the safety ground.
1663766991202.png

In this power supply circuit the triangle represents earth ground. Put a MOV in series with R_Load. Do you know how to estimate the value of the resistor to get your test current (1 ma)? You should also be able to get a good idea of the wattage rating for the resistor if you can predict the voltages on the transformer primary and secondary.
 

ronsimpson

Joined Oct 7, 2019
2,987
I like Dick's idea but;
You need a voltage 2X the voltage of the Device Under Test. (DUT)
You need a circuit that will survive if the DUT is shorted.
Use two transformers from post #5. Use two bridges so you can use 1000V diodes.
Make the capacitors very small to limit the energy available.
At about 1KV most hand hold meters are OK.
Choose a "R_load" that can handle 2W, 1000V. You might have to use several resistors in series.
If you short the leads the current is 2mA. With a 500V DUT the current is 1mA.
1663768245295.png
edited
Add blead down resistors across the capacitors.
Because 1kv caps are hard to find, use lower voltage caps in series with the blead down resistors as shown.
1663769666483.png
 
Last edited:

Thread Starter

xchcui

Joined May 12, 2014
272
Thanks for your guidance included the attached circuits,
but i understand(as you warned me not once)that there is
a risk in that testing.This was the reason why i mentioned
before the megger,which may be safer,i think.
I have a megger(not expensive one),which has:250v,500v
and 1000v options.(see specification at the attached photo).
The spec. says that in short circuit the current will be NOT-LESS
than 1.5mA(i got 2.45mA,for example, while i short it with 250V).
Can't i use the megger for that testing,by adding ,maybe more components(in series or parallel to the probes)to set the right
voltage/current to the varistor on testing?specifications megger.png
 
But this test can't tell me if the varistor meet the requirements,
like if the clamping voltage of the varistor at 1mA DC
meets the value which the varistor is rated.
Going back to your original post, I think you are looking for a quick result rather than an exact result. By using a cheap transformer which changes your grid voltage to 380VAC you will have a peak (positive and negative) of about 537VDC. If you connect the MOV in series with a resistor across the output the waveform will be clipped top and bottom when the the voltage exceeds around 430V +/- 10% if the varistor is working correctly. For practical purposes the MOV has a very high resistance below 430VDC - both negative and positive - and a very low resistance above 430VDC. So the voltage across the resistor will be a series of positive and negative humps with peak voltage of 107VDC. I'm guessing this might look something like 20 to 30VAC when measured with a cheap DMM, different with a full RMS DMM as cheap DMMs average assuming AC signals are sinusoidal. To limit the maximum current to maybe around 5mA the resistor should be around 20K. It is possible to do the maths on this if you are looking for a more accurate result.

Nice if you could look at the voltages with an oscilloscope? You are not measuring resistance but voltage across a known resistance. If you followed my previous suggestion going from 220VAC (or 110VAC if this is your grid voltage) to 12VAC with one transformer and back up to 380V with another, then put a variable resistor in the 12V line would draw current thereby reducing voltage into the transformer as soon as the MOV starts to conduct. Then, measuring the AC voltage across the resistor in series with the MOV, increase the variable resistance in the 12V line until the measurement approaches zero. The AC voltage across the MOV will then be peaking at the MOV knee point. Multiply this voltage by 1.414 - the square root of two - for a result which will be within the +/- 10% tolerance band of the component.
 

ronsimpson

Joined Oct 7, 2019
2,987
I have a megger
If your megger is used to measure resistance only, it will not work.
Some meggers can also measure voltage. That will work.

You can use the megger as a HV power supply and use a 1kV meter to measure the voltage.
Put a meter across the megger's output. Now attached the Device to Test.
 

Thread Starter

xchcui

Joined May 12, 2014
272
If your megger is used to measure resistance only, it will not work.
Some meggers can also measure voltage. That will work.

You can use the megger as a HV power supply and use a 1kV meter to measure the voltage.
Put a meter across the megger's output. Now attached the Device to Test.
Yes,that what i was meant to,but since the megger will give
500VDC,shouldn't i put a resistor in series with the mov in
order to limit the current more?as the short current of the megger at 500VDC position is several milliampere.

Going back to your original post, I think you are looking for a quick result rather than an exact result. By using a cheap transformer which changes your grid voltage to 380VAC you will have a peak (positive and negative) of about 537VDC. If you connect the MOV in series with a resistor across the output the waveform will be clipped top and bottom when the the voltage exceeds around 430V +/- 10% if the varistor is working correctly. For practical purposes the MOV has a very high resistance below 430VDC - both negative and positive - and a very low resistance above 430VDC. So the voltage across the resistor will be a series of positive and negative humps with peak voltage of 107VDC. I'm guessing this might look something like 20 to 30VAC when measured with a cheap DMM, different with a full RMS DMM as cheap DMMs average assuming AC signals are sinusoidal. To limit the maximum current to maybe around 5mA the resistor should be around 20K. It is possible to do the maths on this if you are looking for a more accurate result.

Nice if you could look at the voltages with an oscilloscope? You are not measuring resistance but voltage across a known resistance. If you followed my previous suggestion going from 220VAC (or 110VAC if this is your grid voltage) to 12VAC with one transformer and back up to 380V with another, then put a variable resistor in the 12V line would draw current thereby reducing voltage into the transformer as soon as the MOV starts to conduct. Then, measuring the AC voltage across the resistor in series with the MOV, increase the variable resistance in the 12V line until the measurement approaches zero. The AC voltage across the MOV will then be peaking at the MOV knee point. Multiply this voltage by 1.414 - the square root of two - for a result which will be within the +/- 10% tolerance band of the component.
Although your method and your description(which clarified
also the calculation i need to do) seems
to me a simple and good solution to my problem,i don't have
a 380V transformer right now.
And since i have the megger at hand,which can provide also 500VDC,i wonder if i can also connect a resistor in series
with the varistor(or perhaps there is no need,since the megger
has already current limiter)and check the voltage variation on the varistor with a DMM?
Btw,what are the current/time value that i can pass through the varistor(during testing)without degrade it or damage it?
 
since i have the megger at hand,which can provide also 500VDC,i wonder if i can also connect a resistor in series
with the varistor
I think I'd try it first with a resistor in series although the megger current limiting may suffice - presumably the 500V drops to reduce the current when the internal current limit is reached? A 68K resistor in series would limit the current to around 1mA if the MOV is working correctly.

MOVs are designed to absorb short voltage spikes, typically amps rather than milli-amps for a few micro seconds. They degrade a little every time they see high energy spikes. If you stick to 1mA for testing I expect that'll be okay.

Since the megger is designed to measure high value ohms I'm thinking it does this by measuring the current through a smaller resistor - analogue ones maybe by just passing the current through a moving coil meter. Since it actually doesn't know you have a MOV in the circuit it may present you with a resistance value based on the 68K resistor (which is dropping 70V) in series with an equivalent resistor which is dropping 430V - i.e. a total resistance of 68 + 68 x 430/70 = 485 ohms. But the specification for the voltage output implies that the 500V range can be as low as 450V which won't be enough to push current through the MOV at all. If this is the case you may need to use the 1000V range and change the series resistor to 560k and the megger might read something around 560 + 560 x 430/560 = 1,0M ohm. All this depends on the actual megger output voltages and the MOV voltage which both have quite wide tolerances. Be sure that you use a resistor that can handle a high voltage, if not sure make up the 560K with ten 56K resistors in series.

Please take care - high voltage DC is dangerous. Only turn the megger on when the chosen resistance and MOV, and any attached DMM are firmly connected. It may be that the megger has it's own current limit set below the accepted level of current that is fatal, but this is something you really don't want to test!
 

Thread Starter

xchcui

Joined May 12, 2014
272
Thanks,Jerry-Hat-Trick for your explanation and for your examples,it makes the thing clearer.
In regards to that:
I think I'd try it first with a resistor in series although the megger current limiting may suffice - presumably the 500V drops to reduce the current when the internal current limit is reached? A 68K resistor in series would limit the current to around 1mA if the MOV is working correctly.
So,when the MOV is not working correctly,should i expect to a lower current that will indicate that the MOV is not clamping the voltage(MOV resistance still high)?or should i expect to an higher current that will indicate that the MOV is clamping the voltage too soon(MOV resistance decrease too earlier)? and how much far from 1mA should i expect to be that will indicate that the MOV is not work propely?
 
Typically a MOV will fail open circuit so you may just see a very high resistance with the megger. If it has failed as a short circuit you will be putting the high voltage across your resistor. If you don't have a DMM you will see this when you switch first to the 250V range which, if it shows a low resistance your MOV has failed as a short circuit, a very high resistance is what you are hoping for. If you get a very high resitance at 500V it may mean that the MOV has gone open circuit but it could also mean that it's avalanche voltage is above the actual voltage the megger is supplying.

You keep mentioning current but I'm thinking your megger just measures resistance? If you switch the the 1000V range it's important to have increased the resistor value as already explained (with the megger switched off!!) otherwise the current will be clamped by the megger itself.

Maybe try it and see what you get?
 

Thread Starter

xchcui

Joined May 12, 2014
272
Typically a MOV will fail open circuit so you may just see a very high resistance with the megger. If it has failed as a short circuit you will be putting the high voltage across your resistor. If you don't have a DMM you will see this when you switch first to the 250V range which, if it shows a low resistance your MOV has failed as a short circuit, a very high resistance is what you are hoping for. If you get a very high resitance at 500V it may mean that the MOV has gone open circuit but it could also mean that it's avalanche voltage is above the actual voltage the megger is supplying.

You keep mentioning current but I'm thinking your megger just measures resistance? If you switch the the 1000V range it's important to have increased the resistor value as already explained (with the megger switched off!!) otherwise the current will be clamped by the megger itself.

Maybe try it and see what you get?
Yes,i understand that my megger just measures resistance,but
i thought to connect also a DMM(current position)in series with the MOV and when i press the megger ON button,i will be able to see the flowing current and find-out if the MOV is working correctly.
I would like to mention that my intention is to check if a new MOV'S are working properly(clamping on the right voltage)
rather than testing old MOV'S from a circuit,whether they shorted or have an open circuit,since the latter(as far as i understand)can be done with just a DMM(resistance position).
This is the reason why i asked the last question about the current.
 

ronsimpson

Joined Oct 7, 2019
2,987
I think the megger has an internal resistor. The current is limited.
You need the megger as a high voltage supply that is current limited to 2mA, or about.

There are many different types of meggers. Some are DC and some are AC. The varistor does not care but if you add a volt meter it will care. I think you want DC to get right meter readings.

The meter in the megger is not used.
 

Thread Starter

xchcui

Joined May 12, 2014
272
I think the megger has an internal resistor. The current is limited.
You need the megger as a high voltage supply that is current limited to 2mA, or about.

There are many different types of meggers. Some are DC and some are AC. The varistor does not care but if you add a volt meter it will care. I think you want DC to get right meter readings.

The meter in the megger is not used.
My megger is DC and i think that we have already mentioned that the megger has an internal resistor and that the current is limited(it is also mentioned at my attached specification,post #9).
I didn't quite understand your answer as i was talking(post #16)
about connecting DMM(ammeter)in series with the resistor and MOV and you mentioned the volt meter.
When you said:"The meter in the megger is not used",
What did you mean?
 
I think he means that you can simply connect the MOV across the probes of the megger. With the 250V range you should measure around 250V across the MOV and the current will be very low. In the 500V range it will probably current limit and the voltage across the MOV should drop to the actual avalanche voltage at that limit current. The knee is fairly sharp so you won’t see much difference in the avalanche voltage between 1mA or even 5mA. You could probably put the DMM in the line to measure current but this will either be very small if the voltage is below 430V or the current limit if the voltage is high enough. I was trying to explain a way of testing without using a DMM but as you have one then it’s easier and more accurate
 

ronsimpson

Joined Oct 7, 2019
2,987
When I want to test a Varistor I want to know if it still works like when it was new.
In many data sheets, they are commonly tested for 20uS at high current and at 1mS at a lower current. I have machines that do that, and you don't.
Some tests are done at 1mA. I think your part will test, 387V, 430V, 473V with the current held at 1mA. (min/typ/max)
At 430V and 1mA the part will get hot but will survive for a short time. Don't leave the device on for a minute.
Connect a volt meter across the Varistor. Set to 500V or 1kv.
Your megger makes about 2mA. So the voltages will be slightly higher. And the heat will be 2x more. Do not send power for more than one second.
Apply the megger across the varistor and apply power for just long enough to read the volta meter. (not the megger meter!)

Now you will know the 2mA voltage clamp level. Look at the graph. The 1mA voltage is almost the same as the 2mA voltage. So if your megger makes 1ma or 3mA the voltage reading will be with in 1%.

I know this megger + voltmeter test does not test all numbers but it is a good test. A number in the 380 to 475 range indicates the part is good. A very low number show a leaky and damaged part. A high number suggests an open part.

1664305448570.png
Your part is good to 0.6Watts and at 2mA you are past that. Test fast.
ronsimpson
 
Top