Testing an Isolation Transformer

There is almost always some leakage in anytithng, it's a matter of how much . I would test it with an incandescent lamp. Each side of the seconadry to each side of the primary. An incandescent should not light, even dimly when connected to either side of the seconday and anythong else except the othe r side of the secondary. If it does, the transformer is shorted.

You're up in the UK huh. There is a specified hot line leakage test in the US wwhich is supposed to be ran on every piece of cunsumer electronics serv iced if it runs off house power. Most people just use an ohmmeter and do th e cold check but the hot check is alot more revealing. It actually appplies the voltage across.

I am sure they have someting similar there. Just run it on that just like a TV set or something, but also the output. Here they want you to probe each exposed metal part as well, which isn't a bad idea really. You don't want any shorts or heavy leakage to the core either.

I looked it up real quick and didn't even hit the US method, which you prob ably could use. However, if you can get a service manual for a TV set there , there is probably a section on that. If not, really, you could just do th e US one but consider that it is double the line voltage.

Another thing you might do is plug the thing into a GFCI and then ground ea ch side of the output. If the GFCI doesn't trip it is probably safe.

Think about it for a while. The transformer seems to be acting normally.. YOu are probably getting some capacitive coupling form the primary to secondary windings.

Is this in the US or somewhere else ? In the US there is no neutral on a

This is in the UK where it's getting rather late so I'll take it up again in the morning and check for any further replies. I have precisely zero experience in this area, I'm afraid.

There may be enough capacitive coupling between primary and secondary windings to deliver enough current to light up a neon bulb. They don't take much.

Is this a *true* isolation transformer, sold as such?

Medical-grade isolation transformers would (I believe) include a conductive shield between the primary and secondary windings. If the shield is grounded (as it is supposed to be, in this sort of service) it will prevent capacitive coupling to the secondary.

You chose your moniker well. You have analysis paralysis.

You have the ability to use absolutely the worst, least precise, tool to make a measurement under conditions outside the normal specification range of the devices, then fret over the result.

Measure the AC current between the output and ground with a real AC current meter. Put a resistor in series with it to limit the current justincase.

It would not be unusual to measure lethal currents for the worst-case connections to your body and ground. More to fret over.

Working on line-attached stuff is very risky. There is no such thing as "floating". There are always AC paths. When I do it, I drag out the TEKTRONIX A6902a. Depending on the voltages involved and what I'm doing, I might also use an isolation transformer. A GFI in the mix is a good idea.

I like caution, but you could have fixed the scope by inspecting the schematic in less time than this thread has consumed.

You're right there! I've slept on it and have come up with a better idea. No one here will like it and I'll probably get abuse for it, but I'm comfortable with it and it will get the job done with reasonable safety. I'm going to slowly step up the volts via the variac on the primary side whilst sticking my finger into each of the secondary terminations in turn. If I don't feel a tingle by 100V it's probably working as it should and I'll keep winding up to 230V. I shall of course, keep one hand behind my back. I'll let you know how it goes later. Hopefully.

If there were no lethal voltages present to freak out about, yes. What's a "GFI"?

Ground Fault Interrupter senses imbalance in the hot/neutral and switches off if there is imbalance, which likely goes from hot to ground by external means...possibly through you.

There are no lethal voltages on a schematic diagram you're reverse engineering...no voltages on the paper at all.

Diagnosis is topology dependent, but, in general, I wouldn't expect a switcher to behave predictably outside its design input voltage.

Varying the input voltage while testing the output with your finger is so ludicrous that you're beginning to sound like a troll.

If your representations are truthful, you probably shouldn't be using a scope, much less trying to fix one.

I guessed it must be what we call a "residual current device" = RCD.

Mike.

Yes, well I expected such a response. Anyway, my method didn't actually prove anything; I got no tingle at all even on around 160V on the primary side so I guess I'm not adequately earthed. And I'm not about to stand in a bowl of water to improve things, or whack the volts up any further so I'll defer to someone's earlier suggestion and measure it properly with a mains meter. I admit it was a bit daft but it seemed a useful expedient at the time. BTW, we call those things RCDs here and I do already use one for extra safety.

If it were that simple I'm sure someone would have already suggested it. AFAIK - and I could be wrong I admit; it's not my field - most transformers, isolation type or not, will pass this test so it doesn't really tell us anything.

The other kind ( aka auto-transformers) will not.

.... Phil