High Voltage Diode Testing Question

[Dennis P Brown]

How can a high voltage diode (5 KV and up) be tested to determine if it is working? I have tried an ohm meter but this reads null (no reading) for either direction.

[David D Speck MD]

Dennis,

HV Diodes are made up of many series connected diodes of lower voltage ratings. They will look like an open circuit in either direction to an ordinary DVM.

Whereas a single silicon diode has a typical forward voltage drop of 0.4 to 0.6 volts, a HV diode may have a forward drop of 20 - 30 volts or more, depending on how many series connected individual diodes there are inside.

To test that the diode works in the forward direction, connect it in series with a variable low voltage DC power supply, a 1K resistor and a milliammeter (or the current function of a DVM). Increase the applied forward voltage gradually, and see if the diode starts to conduct in the forward direction somewhere between 15 and 30 volts. If you get much above 50 or 60 volts without any current flow then you should suspect a bad (open circuit) diode. Be sure not to exceed the forward current rating of the diode during the test.

In the reverse direction, the diode should look like an open circuit to pretty much any applied voltage below ~50 KV or more.

To check the reverse leakage of a HV diode, you would need a variable HVDC supply, a hefty 100K or 1Meg resistor rated for high voltage, like a Tiger Stripe resistor, and repeat the process with a microammeter while reverse biasing the diode. The reverse leakage should be only a few microamps as you approach the rated operating voltage of the diode. For X-ray diodes, it is unlikely that a typical experimenter would have any HVDC supply that could even begin to approach their rated reverse voltages of 50 - 150 KV or more.

Generally, if you find a reasonable forward voltage drop across the diode, then it will probably work just fine. Trying to test the reverse leakage current is inherently dangerous if you are not absolutely certain of what you are doing.

Be extremely careful not to touch any part of the test setup while conducting such a reverse biased HV test. You could easily be shocked by voltage arcing right through the case of a meter to you. That's why many lab surplus HV supplies have long insulating shaft extensions and Plexiglas front panel extensions built onto their fronts.

The diode, resistor, meter and all leads would have to be completely isolated from ground. Also beware of any stored charge that your HV supply might have inside, waiting to bite you.

Dave

Addendum:
I initially read the request as applying to an X-ray diode, which would be rated at 50 - 150 KV or more for typical X-Ray machine use. I think everything I said above still applies, though the forward voltage drop of a 5 KV diode will probably be more like 3 - 5 volts, rather than 20 - 30.

D

[Chris Bradley]

The Vf of multi kV diodes will be of the order of 20V or so, so a regular multimeter's diode setting won't work. Essentially, high voltage diodes are actually just a string of diodes in one package, so you get a higher forward voltage than for just a 'one single diode' package.

Connect it in series with a 10k resistor and a multimeter set on a mA range of current, then put that across the terminals of a 0 to 30V power supply. Turn it up slowly and at 20V or so you'll see a current beginning to flow, assuming it is the right way around for the current to flow. It'll probably read one mA or so by the time you get to 30V.

Turn it around and you shouldn't see any current.

In the 'flow' direction, you could also put a second high impedance volt meter to measure the voltage across the diode. It will get up to some limiting voltage (the Vf) then not go higher as the diode 'opens'.

If it is some crazy high voltage diode, 50kV or something, then obviously the Vf could be a lot higher and you'd need a higher voltage supply to do this test. But expect a Vf of around 15 to 30V for a 5 to 10 kV diode.

You can do likewise with an HV power supply and a higher resistor value (of course), if you like, if you feel a need to test its Vrr (stand-off voltage) is above some voltage you put across it.

[Chris Bradley]

[snap! - you beat me to the post button...]

Just one caution on the choice of resistor to use - bear in mind many HV diodes are just a few mA. Don't run more current through it than it is rated for, hence I suggested keeping it to just the one mA with a 10k.

[David D Speck MD]

That's OK, Chris,

Great minds do think alike!

Dave

[Dennis P Brown]

Thank you both - this makes more sense than what I was finding on the internet. The issue with diodes above the standard microwave voltage is an issue and this clarifies a procedure that makes more sense.

Also, the thanks for the notes on cautions on testing (I overlook that issue due to my lack of first-hand experience with testing of digital electronics.) That would have been a great way to fry a good diode!

As for high voltage diodes construction, I remember from my solid state physics course being confused on how the theory would apply to devices that operated with just a few tens of volts must less the many thousands that some diodes could handle; even after extensive literature searching I couldn’t find a word on that issue. Of course, it was rather easy using the basic principle of diodes to finally figure out how to create a HV diode; just as you mention one just uses stacking – afterwards, I found it highly odd that no standard books on solid state physics (or any that I have ever found) ever mentioned that method even once.

[Richard Hull]

Solid state physics is one thing, solid state application with extant, real world devices, another.

HV engineering books related to the application of solid state devices would be a better choice. The engineering atmosphere in the solid state world over 30 kilovolts is rather rarified.

Most of the world's electronics communications and computational work today is handled at 5, 3.6 and 3.2 volts.

Richard Hull

[Chris Bradley]

When I started my practical work on my project 5 years ago, I had no real idea on electronics, other than basic circuits from 'high' school. So I asked a couple of companies to make the kit I wanted. They simply said it couldn't be done, didn't even quote me for the work.

So I got out components and figured out how they worked, or could be made to work, for myself. Books and references played an almost insignificant part in my learning. The component data sheets, however, were valuable.

I ended up building stuff that wasn't quite what I had originally asked for, but that's partly because I learned what was possible and what was not. As none of it came from 'books', I now have one patent with novel circuits in some of the claims, and one patent application for a novel high voltage stack in progress.

Generating new ideas/creations from scratch in a field of electronics where there is little readily accessible reading material is not easy, but you end up at a better and more 'informed' place when you have worked your way there by trying things out in practice.

[Dennis P Brown]

Well Chris, your experience and methods for diode testing appear very accurate; tested the diodes in the voltage multiplier (Villard type) and they all turned on nicely and allowed current in only one direction - thanks again for the warning on limiting the current! I used a varac and kept the milliamps under 1.0.

The caps are 20 kV; I assume about 4700 uF (they have 4700 Z listed along with the 20 kV voltage; also, they are 1.5 inches by 0.5 inch disks so not exactly small. So, I realize that they could be highly dangerous when charged.) I assume the diodes are also 20 to 25 kV to match the caps.

With sixteen stages, I hope this device (when under oil) reaches a voltage usable for my acclerator project - it might be needed, now.

With that in mind, I will place the assembly under oil in a week or two (after pulling it down to under 50 microns to remove all air.) If I am satisfied with my safety designs, I'll do a high voltage test (air gap using a 10 meg, 50 watt resistor.) I wil luse a 30 mA, 10 kV neon sign transformer to charge the assembly (yes, the ripple will be bad but this is not for power.) I do not need any significant current (just what would be on a level similar to a VdG) since I just need the electric field. The thing just needs to charge my globe with the accelerator tube.

I hope to do this test in a few weeks if I can seal the assembly in tube with oil and add all the required leads (sealed properly.)

The reason I am doing this now, is that I may be forced to use this voltage multiplier to create the high voltage field for my now completed electro-static accelerator.

My problem last weekend was that after testing my VdG by itself, it appeared to get a full 200 kV on it (7 - 8 cm spark in dry air (33 kV per cm, approx.) Then I installed it on my accelerator tube for a voltage test and, of course, total failure.

Always one step forward, two back ... getting very old.

So I removed the globe assemblies from the accelerator and tested the VdG system with all added features (teflon tube for deuterium gas), metal end seal (to hold the accelerator tube in the globe), and internal vari-resistor shaft - the only item not added was the accelerator tube; the voltage was its full 200 kV. Yet when I include the glass accelerator tube again - absolutely zero voltage - I can touch the globe and don't even feel anything ... madding!

I'm lost ... unless the vacuum tube (glass) is conducting ... the internal pressure was of the accelerator tube was under 2*10^-6 torr; so, I can't believe the internal metal lens were to blame. Could be the wood platform nearby but that seems unlikely but with my luck ... .

[Tyler Christensen]

Those caps are definitely not 4700uF. Based on the size and marking, quite likely 4700pF (a million times less than 4700uF).

This is a good thing, because at 4700uF it would hold 1.8MJ (which is a way you can know it's not 4700uF). Instead, it's about 1.8J. Unlikely to do anything more than surprise you and maybe kick you back. Of course you shouldn't find out! Sometimes the kick back is the most dangerous part as you could get thrown into something sharp, hit something else that is even more dangerous, etc.

[Dennis P Brown]

Thanks - I have little knowledge of real world electronics. I agree (lol), that that must be pF ! I do have a single oil cap (50 KV) that is in the uF range. That monster is heavy.