Bad transformers

Are you sure that it wasn't a regulating (ie Sola type) transformer?

awkward that it is encapsulated. Sounds like the laminations are not properly assembled. I've had transformers with a gap of 2mm between the E and I 's. Result - much higher Imag and overheating. Solution was to return to supplier.

This seems quite high for no-load copper losses. Have you measured the magnetizing current for the similar Dagnall transformer you have?

I have a 12 VA transformer that draws 32.2 mA unloaded from 120 VAC, and the primary resistance is 50 ohms, for a no-load copper loss of .0518 watts. The total no-load loss, copper+core, measured with an accurate low power factor wattmeter is .48 watts.

I checked a couple more small transformers and the no-load copper losses are more like 1/2% to 1% of the full load rating, rather than 10% as the OP is getting.

"The Phantom" "Phil Allison"

** Nope.

** Yep - it's 31mA, with a 6VA tranny.

Makes the copper loss 0.76 watts, off load.

The case rises only a few degrees C.

The primary wire rises by 19 degrees C.

..... Phil

"The Phantom"

** Transformers of the same nominal VA vary widely in their I mag figures - over a range of 20 to 1 or more.

Toroidal types in particular have very low I mags - circa 3 mA in 10 to 20 VA sizes.

Though a low Imag might seem desirable, it is not by any means *essential* to good operation.

The OP's 10 VA tranny is speced as being 70% efficient - so at full load it dissipates 4.3 watts.

From the OP's figures, the off load dissipation is about 2.3 watts.

Nothing strange there.

.... Phil

It actually this one:

The verbal repsonse from the manufacturer is that this is normal performance.

DC primary resistance after 90 minutes for a transformer on load is

478 ohms so temperature rise is 50C. For an off load transformer it is 466 ohms, temperature rise is 42C.

With a 23C ambient that gives primary temp on load of 83C, the hot spot is about 93C. Off load it is 75C with a hot spot about 86C.

I guess the hot spot readings means the infrared thermometer is reading high.

"Raveninghorde" "Phil Allison"

** Nothing to worry about there. 65C rise is a commonly accepted value with transformers.

** Try 73C.

** Bollocks.

The primary copper IS the hottest part of the transformer.

** By about 25 degrees C.

Ever heard of the term " CALIBRATION " ????

Yaawnnnnnnnnn..........

....... Phil

Sounds like they used the wrong core material. Might have been an attempt at cutting costs by using Asian crap.

Does it audibly buzz too ?

Graham

The buyer will only buy what the engineers specify.

You do have a documented approved supplier list ( ideally vetted ) and component specs I assume ? If not, blame yourself.

Graham

Very true. It's a standard method.

Graham

Bear in mind that that Imag is largely reactive. It does NOT represent actual dissipation as VI or are you not familiar with the difference between real and apparent power ?

I make 57mA into 400 ohms, 1.3W of copper losses. No big deal.

Graham

In this particular case we needed 50 transformers.

Clairtronics were probably the biggest supplier of this type of transformer in the UK as they had distribution tied up with most of the catalogue companies. The demise of Clairtronics disrputed the supply chain as Farnell etc were all looking for an alternative source.

In this situation it was a matter of getting transformers from where we could at a sensible price.

Given that we had on going field failures with the Clairtronics transformer from a batch of 900 purchased 3 years ago I am being cautious in making sure I understand the difference in performance.

Phil poked me to cross check the temperature rise which has shown my infrared thermometer is reading 20C high. That is my fault I should have realised if I didn't burn myself when I touched the transformer it could not have been 95C.

And my fault for not having the thermometer regularly calibrated. It is now on the list with the rest of the kit.

The German manufacturer has confirmed the Imag currents etc are normal and not a problem.

I spoke to several UK transformer manufacturers yesterday and they all now import this type of transformer despite web sites saying made in the UK. It is a low cost commodity product and everyone is cutting corners to keep the price competitive. I just have to accept they will run hotter than I am used to.

But the fact that it's getting hot means there's quite a bit of resistive power dissipated as heat.

I did some work for a transformer co many years ago (Belclere) and their stuff was conservatively designed and got gently warm with no load. I guess they are driving the cores much harder these days, but would still reject any transformers that get significantly above ambient with no load. The quoted 75 C is way over the top. Just cheap 'n tacky design imo.

That's quite apart from any green issues w/regard to power efficiency...

Regards,

Chris

Raveninghorde mentioned more than 400 ohms hot, IIRC 466 ohms warmed up with no load. That's 1.5 watts copper loss in the primary, plus core losses. Total heat dissipation with no load is probably almost 2 watts, which is enough to warm up a 10 VA transformer quite a bit.

A 10 VA transformer dissipating almost 2 watts with no load, maybe

2.5-plus watts with full load... Drawing 13 VA with no load... It sounds to me that the manufacturer really cut corners. I am guessing that the core material is only a step or two better than recycled truck bodies, the laminations are a bit on the thick side and the primary turns count is on the low side.

I am also guessing that the transformers were designed for 230V, but I have heard that many parts of England have 240 (True?). If this is true, then cores that already saturate somewhat will do so even worse.

- Don Klipstein ( snipped-for-privacy@misty.com)

Round here, 240 volts on the dot most of the time, perhaps 238/239 at cooking time. Transformers in the old days would take a few percent either way, though I do remember old valve equipment that had tap settings in 5 or 10 volt steps. Probably more to do with internal calibration than anything else though.

Everything gets cheaper and tackier and quality kit can't compete where everything is rated on 'how much does it cost', rather than 'what does the product really need to be a sound engineering design'...

Regards,

Chris

"ChrisQ"

** The OP's temp figure were *wrong* !!!

I told him so and now he agrees his IR meter was faulty.

The vast majority of transformers on sale have 60 degree C or more temp rises, on load.

In the case of PCB mount units, most never run off load.

Capice ??

..... Phil

"Eeysore" is always 10 paces behind any thread.

** Bollocks.

The temp figure is completely WRONG !!

The 57mA off-load Imag figure is of academic interest only.

Cos when on load, it drops by a factor of nearly 4 times, reducing the related copper loss by a factor of 14 and so becomes insignificant.

If the tranny is operated with 60 Hz power, Imag drops by a factor of 3 time off-load and 6 times when loaded - making the related primary copper loss trivial in both cases.

** Its a fully potted tranny - d*****ad.

..... Phil

Well, let's exam "The total no-load loss, copper+core, measured with an accurate low power factor wattmeter is .48 watts."

Does that sound like I'm not "...familiar with the difference between real and apparent power"?

over

=20

=20

=20

Good luck with that. The UK has gotten out of manufacturing anything but "Pounds".