copper "belt" around transformers

[NiHaoMike]

I have noticed that on many power supply transformers, there is a copper "belt" that encircles the entire unit for magnetic shielding. If that's not required, will removing it significantly increase the efficiency?

[Fraser]

Google is your friend......


http://www.diyaudio.com/forums/parts/4841-shielding-transformer.html

The professional audiophile (no not audiofool this time) community know much about transformer eminations and the effect of copper and Mu metal screens. This is one area where appliance poor design can create havoc with mains hum on an amplifier output.

I would always leave the copper screen in place.

[NiHaoMike]

I'm going to be using it (actually an inductor) in a power inverter, so I plan to remove the "belt" if it would significantly increase efficiency, especially at low load. I will leave it if the gain would be insignificant.

[Fraser]

If the copper actually absorbed useful energy from the transformer/inductor it would be a very bad idea because it infers that it is acting as a shorted turn on the core.....not a great idea in any transformer/inductor design.

From what I understand, the copper 'belt' isn't a shorted turn around the cores operational (useful) magnetic flux, and is purely a containment method for spurious E, and to some extent, H field emanations from the core & windings. I don't see how your efficiency will be adversely effected by a correctly installed and implemented copper belt. You could end up with an increased chance of an RFI problem without it though.

[Computeruser]

>>> so I plan to remove the "belt" if it would significantly increase efficiency

The few transformers I have used that have this belt seem to work as designed. The power transformer in my TEK 455A2B2 has just such a belt, and it, too, works just fine.

I cannot see how removing it would increase efficiency in any way. It is probably (in my cases) there for shielding.

... C

[Neilm]

The belt around a transformer can have quite significant effects on the performance of the transformer. In a flyback transformer, adding the belt reduces the leakage inductance. This both improves the efficiency, and reduces the turn off spikes on the switching FET - meaning you can use one with a lower voltage rating which will improve the efficiency even more.

Yours

Neil

[NiHaoMike]

The belt around a transformer can have quite significant effects on the performance of the transformer. In a flyback transformer, adding the belt reduces the leakage inductance. This both improves the efficiency, and reduces the turn off spikes on the switching FET - meaning you can use one with a lower voltage rating which will improve the efficiency even more.

But that's irrelevant for a single winding inductor, right? In that case, wouldn't induced currents in the belt increase losses?

[Frangible]

If you examine the construction of the copper shield, you'll notice that it doesn't go around the transformer in the same manner as the windings.  It's basically at a right angle to them, and thus does not participate much in the inductive loading on the primary.  So no, removing it will add nothing to the efficiency of the transformer, but will definately increase noise and coupling with other circuits.

[ciccio]

But that's irrelevant for a single winding inductor, right? In that case, wouldn't induced currents in the belt increase losses?

In my experience with mains frequency transformers the belt is not a winding, in the sense that it is not a single turn coil. There is a insulation foil that does not allow the shorting of  the "beginning" with the "end" (forgive me  for the imperfect terminology).
If it was shorted, it will carry a very high short circuit current,  with resultant primary winding failure (if no other protection is installed)
No current flows through it, and it only purpose is to reduce the external electric (and sometimes magnetic) filed.
With toroidal transformers I had very good results in terms of external electrical and magnetic field reduction using a steel belt,  made with the same steel tape that was used for the toroid core.
It was possible to increase the magnetic shielding with a two or three turns belt winding, but the added thermal insulation resulted sometimes in excessive temperature increases.

[NiHaoMike]

On the inductor I'm planning to use (PFC inductor from a bad PC power supply), the belt does form a shorted loop around the entire core. Since about half of the belt is right up against the windings, wouldn't there be a significant stray coupling causing an induced current and losses?

[Fraser]

You have to be careful when describing the position of the copper 'belt'. It normally does not pass around the winding bobbin but rather, around the whole core exterior. There is a significant difference as the flux that passes around the core doing all the 'work' mainly stays IN the core and is not effected by the EXTERIOR mounted 'belt'.

If in doubt...do it the practical way. Test efficiency with the 'belt' in place and then without it. I think you will find the view that there will be no appreciable improvement in performance without it, proven.

[Frangible]

Aurora is right.  Most of the flux is confined to the core - the shielding prevents extraneous flux from radiating out and being picked up by sensitive local components.  You seem intent on proving that this shield is an impediment to efficiency - the only way you'll know for sure is to measure the efficiency of the circuit with and without the shield.  I doubt you'll notice much if anything.

[NiHaoMike]

The inverter has not been built yet so I can't do an in-application test. And once the inductor is installed, there's likely to be enough nearby components to make it difficult to remove the belt and even more difficult to put it back on if results are unfavorable. I could do a Q test but it would not be representative of actual operation at high power.

If there has been a simulation or actual experiment done before about this, I would like to see the results.