19

I have a mature product (designed by me) on a PCB. For science, I would like to induce Electrostatic Discharge (ESD) damage to the board so I can see how it behaves afterwards. This test would be purely academic.

One idea that comes to mind is to apply triboelectric effect using a vacuum cleaner. That seems like a plausible idea, since some users really use a vaccum to clean their board. What are other "average Joe"-methods (for example: wrap the board in plastic) to inflict ESD-damage?

What are some exaggerated methods to inflict damage? (Like using a raygun or something other silly method)

jww
  • 101
  • 5
bos
  • 950
  • 1
  • 11
  • 24
  • 4
    Some other high-voltage sources include miniature Tesla coils, Van De Graff generators, Marx generators, neon sign transformer with Cockroft-Walton output multipliers (warning: dangerous), etc. – rdtsc Oct 17 '19 at 12:49
  • 5
    You could also kill yourself and the device with a jacobs ladder .. The possibilities are endless. – Sorenp Oct 17 '19 at 13:03
  • 3
    3 words: electric fly swatter. – dandavis Oct 17 '19 at 18:33
  • I know you said exaggerated, but what type of devices are on the board? IIRC from my training, damage from ESD on some devices can occur can at voltages as low as 20V, which isn't a visible arc. – mpdonadio Oct 18 '19 at 00:50
  • 7
    I am wondering what kind of result you can interpret with a sample number of "one" zapped board. When we test products with a calibrated consistent ESD device, the damages can range from `nothing`, `some register corrupted`, `some components fried`, to `the whole circuit dead` and everything in between. You would need a lot more sample to extract meaningful result. In your case if you use many samples but your ESD discharge is not consistent, it will also make the results ininterpretable. – Hoki Oct 18 '19 at 15:54
  • @Hoki I agree with your sentiment. I don't think the OP will be able to make any meaningful statement about the results of an experiment like this...it means nothing. – Elliot Alderson Oct 18 '19 at 17:29
  • 1
    Pack it in styrofoam peanuts and ship it to a customer? – Suncat2000 Oct 18 '19 at 18:29
  • 1
    If you want to do this on company-time, there are standardized tests that do exactly this. Usually go to 4kV, but nothing stops you from going higher. Naturally, there's equipment pre-built to perform those tests. – Mast Oct 19 '19 at 08:31

8 Answers8

30

My go to method is my ghetto ESD-gun (a cheap electronic lighter) ala - Long neck lighter Split it open, strip the wires and zap away :) It should produce a few kilo volts.

Sorenp
  • 1,287
  • 8
  • 18
6

Scotch tape being peeled was being investigated for defribillators. They also produce xrays when peeled in a vacuum apparently.

DKNguyen
  • 54,733
  • 4
  • 67
  • 153
  • 1
    Interesting. I will take a look into that. – bos Oct 17 '19 at 14:05
  • 2
    @bos I may have been mistaken. It might have been just xrays, not defibs. I can't find anything on the latter. Scotch tape does generate a lot of static nonetheless. – DKNguyen Oct 18 '19 at 15:17
  • 1
    I used to develop my own film. 35mm film was taped to the spool, and pulling the tape off would generate enough static electricity such that it was normal to see an image of the tape on the "tail" of the film after it was developed. – Mark Harrison Oct 19 '19 at 22:16
6

Dry room + carpeting

Find a room where you walk across the carpet and touch a doorknob and get a little zap. Bring your board to that room and repeat using the board instead of the doorknob.

manassehkatz-Moving 2 Codidact
  • 2,522
  • 11
  • 19
2

If you have an old CRT TV or monitor gathering dust, there is a 2nd anode supply of 15-30kVDC typically. Potentially lethal, so make sure you read up on the appropriate precautions.

Old-style (non-electronic) oil ignition transformers produce about 10kVAC at 10-30mA and are similarly potentially lethal. They supply enough power to make a Jacob's ladder (the arc has to be of sufficient intensity to heat the air enough to drive it upwards for the ladder to work).

An old disposable camera (if you can still find one) typically has an electronic flash circuit that generates a few hundred volts and a storage capacitor that can deliver perhaps 5J for a small one (120uF charged to 300V), from a 1.5V battery.

The above goes well beyond what you would expect from static discharge, in terms of energy, also perhaps peak current and voltage. If you want to do scientific studies of ESD, you should use a standard ESD test circuit (some resistors and a high voltage capacitor of something like 100pF). You'll also need a DC high voltage supply to charge the capacitor, usually to some kV.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
  • 2
    @DmitryGrigoryev Yes, I was just adding some text on the differences. I suppose the more extreme methods might be suited for lightning protection testing- the kind with gas-discharge tubes and spark gaps as with telecom devices. – Spehro Pefhany Oct 17 '19 at 22:59
1

A targeted zapping with a "gun" as Sorenp mentioned is useful if you want to test something in particular, like a button that will be touched by a possibly charged-up user.

But if you want something more random/general, your typical office chair is a great generator of ESD and EMI, because of the fabrics and foam interaction, plus good isolation from ground because of plastic wheels, plus typically dry environment. In fact, a good percentage of chairs keeps snapping and cracking for a while after the sitter leaves - just listen!

So: if you are "lucky" to have such a chair, just sit, shuffle a bit, stand up and touch your circuit; or put it on the chair while connected.

More detail at http://www.emcesd.com/pdf/uesd99-w.pdf

hmijail
  • 119
  • 3
1

You've never taken a sweater off in a dry room, then? I'd think this is a pretty easy way to generate static charges without really moving much. Also you can just rub wool on cotton, but the sweater method seems pretty tried and true.

hexagod
  • 121
  • 3
1

Having had a lot of hassle with ESD inflicted damage of circuitry I'd like to add my 2 cents.

There are a lot of methods described in other answers which will work for sure to generate discharges which can damage electronics. But what exactly is it you want to know?

If you cannot measure or reproduce the damaging event with some exactness, the gain of knowledge is going to be limited.

The damage depends on a lot of variables:

  • Voltage
  • Charge
  • Series resistance in discharge circuit
  • Behaviour of surrounding circuit
  • Possible pre-damage of compontents due to other ESD events or different stresses

Furthermore it is very difficult to find out, if you have successfully inflicted any damage. Single MOSFETs may show significant change of gate resistance upon a single ESD event. But little more complex device are very difficult to analyse and may hide an initial damage until they are completely broken by other events or other consequent deterioration to happen.

What I want to express is:

  • Before fiddling and trying to damage your board cleary define, what exactly you want to find out
  • Design your experiment in a way you can repeat. E.g. by measuring your ESD source, or even probing the damage capability with a set of identical MOSFETs (not that expensive)
  • Make sure your device hasn't been damaged in advance (can be very very difficult)
  • Make sure you have reliable means to detect a damage

If you do not take those measures, the outcome of your experiment will be mostly useless in my eyes or the rationale might be reduced to "hey, they always were talking about this so called ESD damage. I tortured my board and I think perhaps there is such thing as ESD damage. Perhaps, probably, eeeh, perhaps."

Ariser
  • 3,846
  • 3
  • 23
  • 43
-1

There is no point in doing what you propose. If you want to perform ESD testing on your devices, what you really need to be using is a calibrated ESD simulator. With a calibrated source you can test your device at all relevant test points up to the ESD tolerance limits you have designed for. If you have designed well, you should be able to complete your tests without damaging your board. This step verifies that your design is correct and that you have achieved the performance you intended.

With a calibrated device you can then also test to destruction if you're curious about what the absolute limits of your device are. In this case, as with the verification tests, you complete the test with a quantified answer to your question.

If you just zap your board with whatever you have lying around it may or may not fail in a way that you may or may not notice. If it does or doesn't, in any case, you will have no idea what magnitude of discharge caused the damage and you will learn nothing. There's no point to blindly damaging equipment unless your intention is simply to satisfy a sophomoric desire to wreck stuff, in which case this question is not about electrical engineering at all and does not belong on this site.

J...
  • 1,261
  • 8
  • 14