Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
zaphod Regular Member Joined: 03/06/2018 Location: United KingdomPosts: 93
Posted: 04:06pm 22 Jul 2018
Copy link to clipboard
Print this post
Something for the toolbox :) This is for testing high voltage VDR's and TVS devices you suspect might be kaput! The voltage available depends on the switching device used and the inductor construction, I just knocked this up from bit's in my junkbox/spares so the circuit and details are just as an idea that you can alter to suit :)
About the only critical factors are to ensure the energy is within the switching devices avalanche rating and the inductor doesnt saturate.
In my case my scope probes are only rated to 600V so a 10Meg resistor is also used to make a 2:1 attinuator with the probes imput impedance (also 10Meg) this resistor has to withstand half the peak voltage e.g. 400V so you may need more than one in series depending on what you have in your junkbox :)
Operation is simply tie the DUT across the switching device and likewise the scope (via the attinuator if reqd), set the latter to single shot and press the button, peak voltage should be the clamping voltage of the DUT, if its 800V (in this case) the DUT is failed open circuit!
Check the DUT isnt failed short circuit before placing in this circuit else you may overheat the inductor depending on your bench psu.
Hope this is a usefull idea to some :)
Note the mosfet model is encrypted on the ON website but of course you want the model for your mosfet :)
OOpps just spotted an error in the energy calculation, it's actually appx 1.9mJ Edited by zaphod 2018-07-24Cheers Roger 1Kwp DIY PV + Woodburner + Rainwater scavanger :)
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 11:24pm 22 Jul 2018
Copy link to clipboard
Print this post
I have never seen a TVS fail open circuit, unless it has literally exploded and is totally gone! Usually they fail shorted due to internal over power dissipation meltdown.
They can easily be tested with current limited dc, say 400v (or whatever) at a milliamp or two for a few seconds, is only a few hundreds of milliwatts dissipation. I have a couple of different current limited high voltage dc supplies that I can use for doing this.
The TVS voltage ratings are usually pretty accurate, and the knee is very sudden so its easy to measure the exact knee point.
It should also be possible to test avalanche rated devices at very low continuous dc current. Its similar in principle to non destructive insulation testing.
Nothing wrong with pulse testing in principle, but measuring the exact value of applied dc is a lot less error prone than measuring the exact clamped amplitude of a very fast narrow pulse.Cheers, Tony.
zaphod Regular Member Joined: 03/06/2018 Location: United KingdomPosts: 93
Posted: 07:48am 23 Jul 2018
Copy link to clipboard
Print this post
Hi Tony thanks for the info, yes I have had some TVS shorts but wondered if any of them had become innefective through being open. Of course had I a high voltage power supply I wouldnt bother with the pulse tester but my valve days have long gone and due to enforced downsizing all the related gear went too (culled).Cheers Roger 1Kwp DIY PV + Woodburner + Rainwater scavanger :)
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 08:28am 23 Jul 2018
Copy link to clipboard
Print this post
I have accumulated some fairly old test equipment here, but its still useful sometimes.
The only thing I still own that has a valve in it is a Keithly Instruments 240A high voltage power supply.
These cannot be worth much these days on the surplus market, but 0 to 1,200v (positive or negative ground referenced) current limited to 10mA, and switchable in one volt steps can be very convenient.
I use this surprisingly often, and if you are messing about with higher voltage circuitry, it can make a very handy voltage reference source for testing or calibration. Cheers, Tony.
zaphod Regular Member Joined: 03/06/2018 Location: United KingdomPosts: 93
Posted: 03:23pm 23 Jul 2018
Copy link to clipboard
Print this post
Looks great and very usefull. I used to have loads of gear, mostly heavy and large but when I got downsized (UK style divorce with kids!)I had to live in a rented bedsit for a while so an awfull lot of stuff had to be scrapped :(
Anyway I built the circuit, works fine except the probe attinuator, I know I should have known better haha only works for DC not pulses and would take forever to calibrate so threw the towel in and ordered a 100:1 2Kv probe, hopefully enough to last me a while.
Enclosed pic shows 180Z VDR (nominal 180V) with normal scope probe at 50V/div of course the other aspect of this tester is the current starts high (~8A) then ramps to zero so the clamping voltage changes through the pulse although perhaps a pulse test will give you more of an idea how the device will perform with a real spike! Anyway the purpose is served of a basic go/nogo tester :)
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 09:28pm 23 Jul 2018
Copy link to clipboard
Print this post
That is a very interesting way to test, I have never seen it done that way before. Simple and effective. It should in theory be possible to build a properly compensated high voltage probe, but its not that easy to do in practice. I have a home brew 100:1 oscilloscope probe and a Fluke 20Kv EHT probe, both of which are only really useful for dc or mains frequencies.
MOVs are particularly difficult to test because of the very soft saturation characteristic. Old ones that have been in service for a very long time seem to become increasingly leaky. Very difficult sometimes to decide if one is o/k enough to put back in, or to toss it.
Much prefer TVS, they are either perfectly fine or completely FUBAR leaving no room for doubt.
Its very difficult to protect anything that has to be directly mains connected, and I gave up on several projects in sheer frustration as being impractical.
Biggest commercial project I tackled that required very effective transient suppression I used large powdered iron toroidal chokes and MOVs plus a second stage with TVS all inside a well grounded metal box. It was a case of the equipment having to work with very high reliably over a long period where size and cost were only a very minor consideration.
We had access to the Government Telecom high voltage testing laboratory, and they had some fearsome equipment for doing lightning suppression tests to international standards. Pretty interesting actually, although I am glad I was not paying for the tests...Edited by Warpspeed 2018-07-25Cheers, Tony.
hotwater Senior Member Joined: 29/08/2017 Location: United StatesPosts: 120
Posted: 07:23pm 26 Jul 2018
Copy link to clipboard
Print this post
I built one for work which had a small transformer. The output was provided by a diode multiplier. Caps were small enough that it was inherently current limited.
zaphod Regular Member Joined: 03/06/2018 Location: United KingdomPosts: 93
Posted: 09:42am 27 Jul 2018
Copy link to clipboard
Print this post
Just as good if you happen to have a small high voltage trafo in the junkbox :) Here is a picture of mine
zaphod Regular Member Joined: 03/06/2018 Location: United KingdomPosts: 93
Posted: 09:50am 27 Jul 2018
Copy link to clipboard
Print this post
Your right direct on grid is hard although I dont usually have to much difficulty with grid input smps! I must admit one reason I built this gadget was to test that TVS's do not indeed fail open as it would be disasterious if they did :( I have also heard but not experienced the business of VDR's wearing and again maybe the lack of any means to test them is part of the problem. I guess normally it's not something I have to worry to much about but I am having trouble with a semi directly grid connected H-bridge popping 1200V IGBT's hence my sudden interest in surge protection devices and testing them :)Cheers Roger 1Kwp DIY PV + Woodburner + Rainwater scavanger :)
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 12:06am 28 Jul 2018
Copy link to clipboard
Print this post
[quote]I am having trouble with a semi directly grid connected H-bridge popping 1200V IGBT's hence my sudden interest in surge protection devices. [/quote] You are not the only one, its a very worrying problem.
TVSs should be fast enough, but you need a series impedance as the TVS itself has a limited energy absorption capacity. Shunt capacitors are useless, they just turn a short high voltage spike into a much longer lower voltage bump, without absorbing any of the energy.
Series inductors are a much better bet, because the magnetic core can be made deliberately lossy so that eddy current and hysteresis loss from a very short high energy pulse can swallow a fair bit of the energy, while adding negligible series impedance at 50Hz. The crappiest grade of powdered iron with a lot of turns can be quite effective if its designed to work at the required power level. Not too big (lossy at 50Hz) and not too small (ineffective).
Ferrite is nowhere near as good because it saturates at a much lower flux level than powdered iron, and its much less lossy at the frequencies we want to soak up.
Its very unlikely a single stage of suppression can be made effective enough. A layered approach is the way to go.Cheers, Tony.
zaphod Regular Member Joined: 03/06/2018 Location: United KingdomPosts: 93
Posted: 07:43am 09 Aug 2018
Copy link to clipboard
Print this post
100:1 scope probe arrived good for 2Kv so here is the tester open circuit showing the mosfet avalanche at 1kV. This will be useful not only for testing protection devices but also for checking avalanche charecteristics of other devices at energies low enough to avoid destruction or ageing (hopefully).
For comparison here is a 1.5KE170A note the attinuator calibration is not good on these cheap USB scopes!
Edited by zaphod 2018-08-10Cheers Roger 1Kwp DIY PV + Woodburner + Rainwater scavanger :)
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
