AC and DC Hi Pot testing

[Faringdon]

Hi,
Suppoisng your  test instuctions for an offline SMPS say flash test at 2100VDC for 10 Seconds. (input to output)
...but you only have an AC flash tester.
-What AC voltage do you use for the 10 seconds?

[Martin72]

Hi,

Appx. 1500Vrms.
Sometimes we have the opposite situation, a DUT should be testing with an AC voltage of xyz Vrms for say 1 minute.
We use a DC highvoltage tester with adjustable testing time.
So we calculate the value Urms*SQR2 = DC value.

Remarks:

-When the hipot test is a repeatable type (was already done, was not the first one), testvoltage should be reduced to 80% to avoid damage - A highvoltage test stresses the components.
-Using AC instead DC voltage could make problems when you test against earth/gnd (possible capacitors against gnd for emc reasons)
-Could also fail when testing input vs output - Some smps got a cap between input and output, when testing with dc, cap will be loaded once, when testing with ac, cap will be loaded permanently, a current will flow permanently.
-Do not exceed the test-current, hopefully this current is mentioned.

[TaylorD93]

BS EN 61010 for example does State certain test conditions for safety testing and does include some options depending on the circuitry and type of equipment.

But obviously if the psu you are testing is compliant with another standard its best to read up on the requirements of why a DC hipot test is used.

I know at my work our products have alot of filter caps (3Ph inverter drives) so we do DC hipot at 2kVDC used to be 2.2kVDC

[Faringdon]

A highvoltage test stresses the components.

Thanks, i always took it that anything over 500VAC will stress the components and cause cumulative damage?

[coromonadalix]

loll 

We  have to test MeanWell medical stuff ast 4000 vac for an smps  ac adaptor, and they must suffer and pass 1 minute of test

At 4010 vac  they blow  loll,  only 6 in 2 years have failed, BUT we've seen some spikes could pass thru them
had to add some surge/spike  suppressor diode in a revised citrcuit design

Other stuff  say with an xformer etc ,,, have to be tested at least at 1600 vac and doesn't go over 5 ma  current test  for 5 second period of time

Iso related stuff    loll