Hi-Pot tested PSU's never sent to customer?

[cupoftea]

Hi,
The Cosel TUNS500F48 ACDC power module datasheet says it can survive 2000V Hi-pot test from input to output. However, would you agree, that every production unit should never be exposed to such high voltage?...since it causes some damage which is cumulative. Rather, this test should only be done on one prototype part. Then that unit should never be shipped to customer...but just scrapped? Would you agree, for production units, any Hi-Pot test voltage should never be more than 500VAC?

TUNS500F48 datasheet (Page 8)

https://en.cosel.co.jp/tool/tag/pdf/SFE_TU.pdf

 

[FvM]

Please read the safety standards and respective test specification to understand the difference between type and routine tests. 3000 VAC 1 min listed in the datasheet is a standard requirement for type tests of 150 - 300 VAC CAT II equipment with reinforced insulation. Routine tests are carried out for 5 secs only.

 

[Easy peasy]

We routinely test psu's to 2kV ac and higher ( 3750 Vac rms ) there is no damage to the transformers or opto's ( change in leakage currents after the test )

in any event most off mains power supplies should tolerate this test for 1min without issue - esp only 2kV ac rms, mains to earth - or - mains to output ...

if it don't its not safe when there is a lightning strike on the mains,

there needs to be more C to earth on the output to limit the volts generated across these caps during the mains to output isolation test.

2750 Vac for 1 min is a minimum test for safety isolation ( mains to o/p , i.e. anything you can touch )

 

[cupoftea]

Thanks, the datasheet is quoting 3000vac for 1 min with 10mA cut-off current.

10mA sounds very high...after all, this is current thats getting "rammed" through the insulation. Can we choose 1mA cut-off instead?

And what of these Hi-Pot testers that only allow you to set isolation voltage and time......what current level do you think they would have as cut-off?

Page 8 for ACDC module spec

https://en.cosel.co.jp/tool/tag/pdf/SFE_TU.pdf

 

[FvM]

What do you exactly want to achieve? If you you are planning routine tests for instrument production, consult the applicable safety standards. For measurement & control instruments, this is IEC 61010, requesting 5 sec AC or DC tests.

It makes no sense to test DCDC modules that already performed production tests at the manufacturer, but tests of the complete instrument are usually required. The applied test voltage must be tolerated by all components connected to the power supply input, in some cases it's only possible with DC test voltage.

10 mA limit value accounts for leakage currents of Y capacitors.

 

[cupoftea]

there needs to be more C to earth on the output to limit the volts generated across these caps during the mains to output isolation test.

Thanks, are you referring to fig 1.1 (Page "TUNS23") of this?....

https://en.cosel.co.jp/tool/tag/pdf/CME_TU.pdf

...where they show no Y caps from output ground to earth?

Though i wasnt sure where the high voltage was getting caused, by the carrying out of the "mains to output" Hi Pot test?