# Does anyone understand the voltage ratings on metal-oxide varistors (MOV)?

#### Robin66

Joined Jan 5, 2016
275
I'm hugely confused. I'm running a brushed universal motor with a PWM rectified AC supply, so 330V pk (from 240ACrms). I'm looking for a device to protect the output drivers from the voltage spikes caused by the brushes. I think I want a device with a clamping voltage ~350-400V. However the below device has a maximum DC/AC rating way below the quoted clamping voltage. Does anyone understand what these mean? I thought any voltage below 395V would see the MOV as a small capacitance. so why is the max DC voltage rating a mere 200V?

https://docs-emea.rs-online.com/webdocs/126d/0900766b8126db7d.pdf

#### dl324

Joined Mar 30, 2015
14,346
The AC voltage rating is RMS, not peak.

#### Robin66

Joined Jan 5, 2016
275
But 150Vrms = 212Vpk. This is far less than the 395V clamping voltage.

From reading the definition of clamping voltage you've supplied I think 395V is for a well-defined pulse , whereas the 150Vrms limit is a consequence of the power dissipation due to the non-ideal characteristics ie. leakage current, capacitance. So I shouldn't consider anything with an AC rms voltage tolerance <240V, which unfortunately will probably put the clamping voltage above the max of my IGBTs

Joined Jul 18, 2013
25,019
MOV's are rated for operating voltage and energy (Joules) rating.
I use an old General Electric TVS Manual for reference from 1976, you might find it on line, if lucky.
Max.

#### crutschow

Joined Mar 14, 2008
29,528
If you need clamping to a more precise value then you should consider a Zener based surge protector (TVS).

#### Robin66

Joined Jan 5, 2016
275
Ok, thanks. I'll look into TVSs as an alternative

#### cmartinez

Joined Jan 17, 2007
7,633
Yeap, a TVS of about 375V ought to do the trick. How many amps does your motor draw?

#### Robin66

Joined Jan 5, 2016
275
It's 1800W at 240V so approx. 10A. I've ordered some of these, which have a min. breakdown of 380V. I've got a free-wheeling diode which should take the brunt of any off-time surges.

https://uk.rs-online.com/web/p/products/5438832

To give a little background about why I'm now investigating MOVs and TVSs, below is my circuit. I programmed the PIC to give a 10% duty cycle. After initial testing with a 500W lamp, I hooked it up to my universal motor. The motor smoothly run at low speed for about 10 secs, so I thought this a success. However, when I tried to run it a second time I found that the 3A fuse had blown. Rather stupidly without checking any of the circuitry I replaced the fuse with a 13A fuse and bang! Q1 was blown to pieces, the AC712 was fried, D1 was shorted, and diodes 3-4 and 1-2 of the bridge were shorted. The rest of the circuit seems to be unaffected and working normally, including, to my surprise, Q2.

I don't know which component failed first. There was no transient suppression at all on the circuit and none near or inside the motor housing. I was testing in a domestic setting ie. no heavy machinery near-by.

#### Attachments

• 47 KB Views: 1

Joined Feb 20, 2016
4,116
What rating is the diode across the motor?
I would look for more than the stall current of the motor, at least. It may have been too small to handle the current.

#### Robin66

Joined Jan 5, 2016
275
It's rated at 5A average current. I thought this would be sufficient because it's only a free-wheeling diode. In my test I expect the current was well below 5A because the duty cycle (at 5kHz) was so low.

Joined Feb 20, 2016
4,116
Obviously, the diode current rating was too small, so it popped.
This is a case where bigger IS better

#### cmartinez

Joined Jan 17, 2007
7,633
It's possible that the TVS you chose has a breakdown voltage that is too low. Those diodes' ratings list at what voltage they will be fully conducive, but they start conducting partially at voltages significantly below that. If that's what happened, then the diode was conductive for too long a time per cycle during the first run, and that might have resulted in its warming up until if finally failed on the second test.

And it must've failed acting as a closed switch, rather than an open one... that's possibly why everything went *poof*!

Check the datasheet.

#### Robin66

Joined Jan 5, 2016
275
s a breakdown voltage that is too low. Those diodes' ratings list at what voltage they will be fully conducive, but they start conducting partially at voltages significantly below that. . If that's what happened,
I'm not sure what you mean. I haven't used a TVS yet. My schematic only shows an ordinary diode and a bridge rectifier.

Obviously, the diode current rating was too small, so it popped.
I'm not convinced because the quoted peak current for 1/2 cycle at ~50Hz is 100A per datasheet. And I was applying a 10% duty cycle pulse where the max on time was <100us. However I doubt the alternatives too so I'll beef up the diode

#### cmartinez

Joined Jan 17, 2007
7,633
I'm not sure what you mean. I haven't used a TVS yet. My schematic only shows an ordinary diode and a bridge rectifier.

I'm not convinced because the quoted peak current for 1/2 cycle at ~50Hz is 100A per datasheet. And I was applying a 10% duty cycle pulse where the max on time was <100us. However I doubt the alternatives too so I'll beef up the diode
Argh! I thought that D1 was a TVS! ... my bad ... either way, it's obvious that the diode failed.

Anyway, I've just looked at its datasheet. And the only type available for the PDR5G diode is SMT. Is that the case? That alone could cause significant heat dissipation problems.

Also, its datasheet says that its RMS reverse voltage is only 283V! ... so that by itself might be the culprit.

#### Robin66

Joined Jan 5, 2016
275
Argh! I thought that D1 was a TVS! ... my bad ... either way, it's obvious that the diode failed.

Anyway, I've just looked at its datasheet. And the only type available for the PDR5G diode is SMT. Is that the case? That alone could cause significant heat dissipation problems.

Also, its datasheet says that its RMS reverse voltage is only 283V! ... so that by itself might be the culprit.
Yes it is a SMD. However I did check after the first 10sec run and nothing was particularly warm. Is 283 not enough headroom from 240Vrms?

#### cmartinez

Joined Jan 17, 2007
7,633
Yes it is a SMD. However I did check after the first 10sec run and nothing was particularly warm. Is 283 not enough headroom from 240Vrms?
I think the 43V difference is too low in this case... what are your transistors voltage ratings?

#### Robin66

Joined Jan 5, 2016
275
Around 370-430V depending on the temperature and the leakage. I do think an over-voltage failure is more likely than a over-current failure due to the low duty cycle I was using.

#### cmartinez

Joined Jan 17, 2007
7,633
Around 370-430V depending on the temperature and the leakage. I do think an over-voltage failure is more likely than a over-current failure due to the low duty cycle I was using.

View attachment 183819

View attachment 183820
It's possible. But from what you've described, the diode failed first, and then things cascaded from there. Otherwise why did it fry at all?

#### Robin66

Joined Jan 5, 2016
275
if the bridge rectifier went first then this could short Q1 which would put D1 straight across the mains, albeit with the motor in parallel. I don't know to what extent the motor would protect D1. However I didn't see the motor turn at all when disaster struck, but it was obviously over in a fraction of a second

#### cmartinez

Joined Jan 17, 2007
7,633
if the bridge rectifier went first then this could short Q1 which would put D1 straight across the mains, albeit with the motor in parallel. I don't know to what extent the motor would protect D1. However I didn't see the motor turn at all when disaster struck, but it was obviously over in a fraction of a second
That bridge is rated at 600V (420VRMS), 10A ... I very much doubt that was the case.