SMPS transformer hipot test without any damage whatsoever

[Faringdon]

Hi, we want to test the isolation of a offline 240vac one tran forward transformer.
But as you know, these cause small cumulative damage...what is the isolation test voltage which does no damage whatsoever?
This txformer looks like just ECW against ECW, with no margins...which, as you know, wouldnt pass normal hipot...but i dont want to blitz it with 3500VAC for 1 second.

[jonpaul]

FTTS just more click bait.
nonsense Q as always everything in your question is false
j

[Faringdon]

Thanks, so what voltage would you put on the HiPot tester for this case.....we are not allowed to take apart the transformer. We have no  manufacture spec for it.
So we need to find out if it will pass regs for isolation...but we dont want to damage it at all.

What kit would you use for this?, ie prooving it is enam copper wire on enam copper wire with no margins.

[jonpaul]

FTTS, review basic magnetics designed for safety and EMI.

1. voltage rating, leakage  and insulation required
see regulations for your location, type of equipment

eg different for industrial, consumer, medical, avionics

2. wire insulation....your usual vague terms..

single, heavy triple, quad,
Formvar, soldereaze nylease polynylon, HPTZ depends on temp grade eg classm155, class200.

3, creep/ strike distance...use rated bobbins split if possible

4. Hypot..V, I ..see appropriate regulations and compliance. May be AC, DC, 10 uA/100uA...5 mA depending on the application

Contact regulatory authority and safety test equipments manufacturer like Rod-L, Hypot, Associated Research in USA.

bon chance

j

[Faringdon]

Thanks, nobody has kit for what we want....because people always use the 3500vac test.
We cant do that.
What we must do , is make a 500VDC power supply (1W)
We must then connect it to pri of transformer.....then sec will go to a 1MEG resistor, and then to ground.
Every microamp of current that leaks through the insulation will cause 1 volt across the 1MEG resistor.
But how many microamps would we expect with a transformer simply (and wrongly) would with enamelled copper wire for primary and secondary...and no margin tape?

I actually suspect we will get much less than 1 microamp. and in fact, we must put the transformer in series with the 240VAC mains...and measure the AC leakage current through a series 1 MEG resistor.
But again, how many microamps AC would  we get?

There is no advice on the web that is good for us concerning this.....people are expected to know how their transformer is wound...but we dont know, and we are not allowed to take it apart....the supplier says so. If we take it apart, we must pay them £100k

[jonpaul]

measure C p-s
reactance Xc is 1/2 pi F *C
I = v/x

j

[Faringdon]

Thanks,  as you kindly say, we can calculate the leakage current, but wouldnt know what leakage current corresponds to a wrongful , totally enamel wire wound transformer, with no margins.

[jonpaul]

identify the wire insulation build, type.

Find spec the of manufacturers

Enamel is too generic a term

j

[Faringdon]

Thanks, they wont give us details of the txformer wire  unless we pay big bucks.
We are evaluating this PSU, and they dont tell us anything about it.
I am thinking of  just making two tranxformers...one with all enamelled wire, and one with one coil triple ins wire
Then i will shove both across the mains, and measure current in 1 MEGOHM sense resistor...then apply the learnings to the actual transformer....and measure its current. then i reckon i will know.

[Gyro]

<sigh>

That's no way of validating a vendor's product, and the...

There is no advice on the web that is good for us concerning this.....people are expected to know how their transformer is wound...but we dont know, and we are not allowed to take it apart....the supplier says so. If we take it apart, we must pay them £100k

...is complete rubbish.

If you are importing the PSU (presumably from China) then you are responsible for it's safety rating, the vendor must provide all relevant information, including transformer winding spec to submit to the test house (you are going to do that, aren't you).

If the vendor wants $100k to provide details then they either really don't want your business or are more cowboy than you are. Drop it and go elsewhere!


P.S. The ignore user function of this forum is utterly useless!

[tggzzz]

P.S. The ignore user function of this forum is utterly useless!

Graybeard wizards encountered this back in days of old, and developed ways to deal with it.

The most effective filter was to ignore any subthread started by a person. That meant you ignored their nonsense and ignored any nonsense caused by their nonsense.

That is very effective at continuing to participate in a thread while not being derailed by idiots.

Not possible with "modern technology"; not everything new is forward progress