DC Hipot Testing

[am1]

Hi all,

I do not typically have to perform hipot testing in my role, but I recently became involved on an issue where a production device undergoing its routine AC hipot testing had failed. The testing is required to be done on the full electrical system/controls cabinet. In order to successfully pass, it can either withstand a dielectric strength test of 1500VAC for 60s OR 2100VDC for 60s. As there was only access to an AC hipot tester, that was the test that was always done, but I am curious how to perform the DC testing on the system. The AC tester we use has a 20A receptacle on its face, so our device is directly plugged into the tester. (https://www.atecorp.com/products/associated-research/4040at). For a DC hipot tester with the red and black alligator leads, how would this be connected? I assumed the red high voltage lead would clip to the live prong of the power cord, and the black return lead to the ground prong of the power cord, or ground bar in electrical cabinet, or any ground point. Is this the appropriate set up?

[voltsandjolts]

It's Live and Neutral together, versus Earth. For DC HiPot you may wish to test both polarities.
This guide is worth reading:

https://www.seaward.com/gb/support/guides/manufacturing/f4c40-electrical-safety-testing-during-manufacture-a-practical-guide/complete/

[am1]

Thank you for the guide. I have an AC/DC hipot tester. For AC testing on my tester, I jumped live and neutral together and applied the high voltage lead at this point, and the ground is my return. For DC, I tested by clipping the H.V. lead to the Live prong and clipping the return to the ground, and passed (2100VDC for 60s). Based on this excerpt and what you mentioned, do I just retest at the same conditions but with the H.V. lead on the Neutral line now?

[thm_w]

Look at page 22 in the guide, it clearly shows the L and N shorted together.

[Martin72]

Hi,

High voltage test...
We test our products (power supplies) as follows.
Input to ground, output to ground, input to output if necessary, if there is galvanic isolation between the two.
Whereby ground means the housing(earthing point).
As a rule, we test this with DC voltage, which has the advantage that any capacitors against ground (EMC!) are charged once.*
If we were to test with AC, there would be a unwanted constant current flow.
The inputs/outputs are connected together to protect internal components in the event of a breakdown.
The polarity does not play a role in DC tests in the constellation, but generally minus is connected to ground, while plus is connected to the inputs or outputs.
However, we once had a case where a diode was connected in series in the input as reverse polarity protection.
In this case, minus was connected to the input, plus to ground.
Now there was a breakdown against ground and since the reverse voltage of the diode was lower than the test voltage, it broke due to the breakdown, which would not have happened if plus had been connected to the input.
Since then, plus has always been connected to the input and minus to ground.
What to do in the event of a high-voltage breakdown?
Possible causes:
- Test voltage accidentally set too high
- Tripping current set too low for DC, keyword charging of EMC capacitors against ground
- Insulation of installed cables damaged
- Insulation generally too low between the potentials
- Metal chip residue present in the housing
- Wire residues present in the housing/between the potentials due to incorrect insulation of cables.
- Humidity too high during test (creepage distance: 1mm/1000V)
- Rare: Defective components in the system itself

Repetitions after a failed high voltage test are best carried out with reduced test voltage, as long as this does not fall below the minimum requirement.
Reason:
The components experience a stress phase as a result of a high voltage test, which they should not be subjected to too often (shortening of service life).