Transformer - strange current waveform and "overheating" ??

[max_torque]

I installed an COTS 3kVA 240VAC single phase isolation transformer on a project:


http://www.solahevidutysales.com/pdf/transformers/AutomationXfrmrs.pdf


I wired the Taps to suit (240VAC primary, 240VAC secondary, ie 1:1 setup)

The unit powered up, but sometimes trips the (12A rated) circuit breaker on it's primary feed when you first turn it on.  Without any secondary load, the unit gets extremely hot after around 10 min of operation (i'd guess an external case temperature of around 80 degC).  I measured the primary voltage and current, shown below:


Yellow - Channel A = primary current via current probe at 100mV/amp
Blue - Channel B = primary voltage


it seems the current is extremely non-sinusoidal with 20A "Pulses" lasting 5ms lagging the voltage waveform, almost like something is saturating and the current is suddenly ramping up.

The unit does provide a 240VAC output voltage, as measured on a multimeter - sorry, didn't get time to scope the secondary yet.



So i guess questions to any transformer experts:

1) how hot should this sort of transformer run under no load? - rating is "80 degC rise at peak load
2) It's a 60Hz rated unit, we are at 50Hz, so i guess i'd expect a higher current ripple, but that should still be more sinusoidal surely?
3) Measuring the primary and secondary taps with an LCR meter shows inductances as expected (around 70mH Prim and Sec when wired 1:1). If taps were miss-labelled and hence mis-wired, i'd expect to see odd inductances across the taps?
4) Could a shorted turn cause this sort of behavoir?

[The Soulman]

Most likely it is saturating because it's running on 50Hz instead of the intended 60Hz.

Transformers these days are designed and build to such a precision that they function to spec
but not beyond that, just to save few pennies on resources.

[richard.cs]

Agreed, it looks like it's saturating.

You don't say which part number you have but if it has an option to wire the taps for a higher voltage you could try that. A transformer designed for 240 V 60 Hz is only usable at 5/6 the voltage, or 200 V on 50 Hz unless it was over-specced in the first place. 80 C case temperature suggests a much higher winding temperature, if you leave it on long enough it will probably overheat and fail.

You should probably just replace it with a transformer designed for 50 Hz use, it's not the right component for you.

[max_torque]

I had considered saturation from 50hz operation, but it just doesn't feel quite right to me for several reasons:

1) The transformer is sold in the UK by a major distributor - (RS components) surely they would have noticed if everyone returned the units they bought?
2) There is no "min frequency" spec for the unit in any documentation for it. Surely, if it saturated at 50Hz, it would tell you that?
3) Looking at the wave form, a minor reduction in the time for each cycle (5/6ths) won't be enough to prevent the current rising suddenly, because the rise time only takes 1 ms?



I can certainly try connecting to a high voltage tap on the primary and seeing what happens!

[max_torque]

Unit bought from here:

https://uk.rs-online.com/web/p/wall-mount-transformers/8908801

no mention of not being suitable for 50Hz.

[Zero999]

Unit bought from here:

https://uk.rs-online.com/web/p/wall-mount-transformers/8908801

no mention of not being suitable for 50Hz.

There's no mention of it being suitable for 50Hz either and the data sheet only specifies the HS5F3AS at 60Hz.
https://docs-emea.rs-online.com/webdocs/14a6/0900766b814a6acd.pdf

It appears you have the wrong part for the job. You need a 50Hz transformer.

[max_torque]

I've actually got a frequency controlled AC source, so i'll stick 240VAC at 60Hz into it and see what the current waveform looks like!      

[Benta]

The datasheet is quite clear on this point: it's a 60 Hz transformer.

[capt bullshot]

It's been already said, your current waveform is the typical "saturated core" waveform.
The RS page says max. 60Hz operation frequency, but that's just a sign on non-understanding what they're selling. Pretty normal today.

[ogden]

The datasheet is quite clear on this point: it's a 60 Hz transformer.

Right. 50Hz generates 20% more flux lines than at 60Hz. That thing is expensive, do not have any resale value in the UK even new. I would try to negotiate about resolution of this mistake.

[T3sl4co1l]

I had considered saturation from 50hz operation, but it just doesn't feel quite right to me for several reasons:

1) The transformer is sold in the UK by a major distributor - (RS components) surely they would have noticed if everyone returned the units they bought?

You're putting an awful lot more confidence in RS than they deserve, or so it sounds based on others' impressions of them!

Tim

[max_torque]

Swapping it for a different unit shouldn't be an issue, i just want to confirm the root cause.   In my pics, if you reverse the current or voltage wave forms (ie the same as swapping L & N around, or changing the direction of current through the current probe) then the current "spike" is leading the voltage by 90deg, which is what you would expect for the position of peak magnetic field strength, so that does tally up.

[BMK]

It's been already said, your current waveform is the typical "saturated core" waveform.
The RS page says max. 60Hz operation frequency, but that's just a sign on non-understanding what they're selling. Pretty normal today.

I've used RS for 20 years. Even back in the paper catalogue days you could bet your life on the accuracy of specifications. In the past 3-4 years I have seen a lot of nonsense in specifications. It seems that they don't use technically literate people anymore to parse datasheets into their systems. You cannot trust those search filters.  In my opinion they have become the worst offenders of all the vendors I use.

You should be able to return the traffo, I've done this many times even  after their supposed 14 day cut-off. One thing for sure is they will answer the phone and solve problems like this with little argument.

[ogden]

You're putting an awful lot more confidence in RS than they deserve, or so it sounds based on others' impressions of them!

They do not know and do not care - you will use component locally or you will export it as part of product you manufacture. Nobody complains that UK RS components sell various kinds of power plugs that would not fit local mains sockets. The same about transformers. RS components is not your local grocery store and even there buyer is responsible to know what he is buying.

You should be able to return the traffo

Sure. By customer protection rules/laws they are obligated to handle returns even if mistake is yours.

[richard.cs]

I would expect it to be marginally better on the 252 V tapping and to not saturate at all on the next one up which is 456 V, though it probably won't be useful for your application either as the secondary voltage will be about 126 V on the 240 V output.

I agree the datasheet is poor not putting upper and lower frequencies on it, but as they describe it as a 60 Hz device without giving limits it could legitimately fail at 59 Hz and still meet its specification. It's a transformer, the physics dictates that it has limits to its operating frequency and them not specifying those limits should be a warning sign. I have no idea why RS stocks it in the UK, it's clearly made for American applications with its 60 Hz rating and the options for voltages like 208,  277 and 480 which aren't really used in the rest of the world. I suspect 60 Hz systems probably exist in a few UK factories for imported specialist machinery but it's quite possibly less common even than 400 Hz.

If RS are willing to exchange or refund that would probably be your best approach. Otherwise stick it on ebay or something (to paraphrase CQ magazine it might make a fine anchor for a small boat). Take a look at Airlink Transformers for a UK manufacturer of similar transformers in a wall-mount case.

I generally find that the data RS pull out of datasheets and onto the product pages is very poor, worse than Farnell and much much worse than Digikey.

[max_torque]

The company that bought the unit spend some thing like £200k (  ) with RS every year, so i don't see any issue getting it swapped (or a refund, as i can't find any 3kVA 1:1 enclosed iso transformers listed that specifically say "50Hz" on RS .........  )

[ogden]

The company that bought the unit spend some thing like £200k (  ) with RS every year, so i don't see any issue getting it swapped (or a refund, as i can't find any 3kVA 1:1 enclosed iso transformers listed that specifically say "50Hz" on RS .........  )

With such purchase amount you can just call them and let them hop while they offer what you need Hard to believe that you can't find 50Hz isolation trafo in the UK. It would be truly laughable.