Copper flux bands on power transformers

[Rhodesplyr]

It's not unusual to see a copper flux band (known by various other names) around a power transformer, surrounding the laminations and windings and soldered to create a low-resistance loop. I know that their purpose is to reduce leakage flux from the PT to reduce 60Hz hum being induced in any nearby audio transformer or wiring.

What I don't fully understand is how they work. I think I've read that the leakage flux induces a current in the band which then opposes the leakage flux. Can anyone give a more thorough explanation?

Has anyone ever added one to a PT that didn't have one, and, if so, how well did it work?

I've run into a few situations where power supply ripple was actually very low, butt there was still hum from the PT's proximity to other transformers or high-gain circuits.

[J M Fahey]

99% 0f magnetic flux travels through/along laminations path because itīs the one with lowest magnetic resistance, thatīs the purpose of using a core, but still a little travels outside and can induce hum into other cores and/or any piece of wire, which can be considered a coil.
Even a short straight piece of wire *is* a coilbecause curent through it induces a field around and viceversa.
The external copper band shorts that outside flux.

Wonīt it short the transformer itself, overheat it, etc.?

Not really: coupling is very poor and to boot , lossy.

[trobbins]

A diagram or two can help in this situation. The first graphic in the following link illustrates the "99%" of flux lines (generated by current through the winding turns) going completely inside and around the core. Some of those flux lines (actually each one is a 'loop') don't stay in the core all the way around the core's loop, and some of those flux lines don't even make it in to the core (not explicitly shown in that graphic). http://www.encyclopedia-magnetica.co.../flux_fringing

For those flux loops extending out to the side of the transformer, if they try to couple in to a 'flux band' then they cause current to flow in that new 'winding', but that winding is a closed loop (a short circuit), which forces flux line path in the vicinity of that band to be restricted. Figure 28 of the following link indicates the effective change in flux line path with and without a flux band.
https://www.dalmura.com.au/static/Hum%20article.pdf

[Tom Phillips]

I also seem to remember that transformers that are designed such that the core is closer to saturation will exhibit stronger flux radiation and therefore cause more hum coupling to adjacent circuity. Does anyone have references to that situation?

[trobbins]

Tom, I think that has been assessed in detail with large power transformers, especially where they are in oil-filled tanks, and the tank structure and mounting hardware gets heated by fringing flux, and that parasitic heating can get bad when the transformer is pushed more towards saturation (ie. voltage tap settings). Probably need IEEE document access to search around.

[Mike Sulzer]

It's not unusual to see a copper flux band (known by various other names) around a power transformer, surrounding the laminations and windings and soldered to create a low-resistance loop. I know that their purpose is to reduce leakage flux from the PT to reduce 60Hz hum being induced in any nearby audio transformer or wiring.

What I don't fully understand is how they work. I think I've read that the leakage flux induces a current in the band which then opposes the leakage flux. Can anyone give a more thorough explanation?

Has anyone ever added one to a PT that didn't have one, and, if so, how well did it work?

I've run into a few situations where power supply ripple was actually very low, butt there was still hum from the PT's proximity to other transformers or high-gain circuits.

Start with a basic property of E&M: If you have a closed path (like an imaginary loop; no wire need be present), there is a voltage induced around this loop that depends on the rate at which magnetic flux is changing through the loop. For voltage, think of electrical potential, that is, the ability to drive current, if a conductor is present around the loop. In principle, and sometimes in practice, you can check this by putting a small wire around this path, not quite closing it, and connecting an ac voltmeter to the two ends.

Now let's assume that we put a closed loop of wire around this path (and let the transformer be contained within as the source of the field), and, contrary to fact, this wire has zero impedance. Then there is no voltage around this loop, which means that a current must flow that creates a magnetic field than cancels the flux that was present through the loop from the transformer field. The magnetic field produced by this current is not just like the field from the transformer (only the total flux through the loop has been canceled), but it is enough like it to reduce the transformer field everywhere. Also the loop has some resistance and inductance, and this further degrades the performance. But it helps.

[catalin gramada]

Is there a difference between classic transformers EM shields vs toroids EM shields? I never seen a EM shield for a toroid going through the core (electrostatic shields did). It.s a simple silicon steel or mu-metal band arround. Think it never helps the flux to stay stucked by its path. But is damn effective, more effective than any cage or screens I tried to interposed between transformers.

[J M Fahey]

I never seen a EM shield for a toroid going through the core (electrostatic shields did).

You wonīt see them through the core *hole* so fully surrounding the core "single leg" because then they would become a shorted turn for the main flux field, shorting the transformer.

In EI transformers they do not go around any single leg either but around the whole core and its 3 legs (central and 2 side ones)

[trobbins]

Similar to how the belly band on an EI transformer conceptually constrains external field lines from extending out beyond the tangential direction of the belly band, a band around the toroid constrains stray flux lines from looping externally out beyond the toroid in the direction of the belly band (which is the directional plane where equipment circuitry is typically placed). For the toroid, stray flux lines will still loop out externally 'above' and 'below' the toroid, but those directions are usually managed by the chassis underneath the toroid, and any large metal hold-down washer above the toroid. So you end up with stray flux having more constrained loop regions at a +/- 45 deg type angle around the toroid.

[catalin gramada]

You wonīt see them through the core *hole* so fully surrounding the core "single leg" because then they would become a shorted turn for the main flux field, shorting the transformer.

That.s happen if put the screen in a closed loop. The same warn apply for toroids screen cage screens which can create a shorted turn through chassis-transformer bolt -metalic screen cage -and presumtive contact between screen and chassis (meant closed loop)

Late: read it- toroids have very low leakage flux arround compared with classics...Could be, but I used lots of toroids and without a proper shielding get more or less noise in speakers even without output transformer be electrical conected somewhere.. and used to keep a lot of room in between. You cannot turn a power toroid, but can turn OT instead if have multiple bracing choices,so there are limited options arround...

[R.G.]

If you have a magic something that is trying to push a magnetic field through open space, how much magnetic flux flows through space from that push is proportional to the push.

Well, duuuuhh, R.G. - so what? The proportionality constant for how much flux flows for how much push is called the permeability of free space. It's value is taken to be unity, and the units are set up so this is true. This is analogous to conductivity - how much electrical current flows in a loop for how much push in volts.

OK so far. All materials have a conductivity of some sort. Copper has a high conductivity, iron about 1/3 of that, other materials usually less. Silver is best for the metallic elements, and you have to do downright strange things with yterbium-niobium-cobalt- unobtainium held at -280C to get less - a superconductor or a partial one. We all kind of know that.

What is a better conductor of magnetic flux than free space? Turns out there are a few. Iron, nickel, and cobalt, and their alloys, and some of their oxides and such are better conductors of a magnetic field. Transformer iron, for instance, may be 10,000 to 18,000 times better, meaning that for the same magnetic push, that many more times magnetic flux flows. There are exotic alloys (e.g. mumetal for one) that have even higher relative permeabilities than the iron alloys.

If you have a copper wire of the same dimensions as a pencil lead, and hook the pencil lead in parallel with the copper wire, then hook a battery across both pencil lead and copper wire, what is the ratio of current flow in the copper versus pencil lead? You're right - you measure the resistances of pencil lead and copper wire, invert to get conductivities, and that's the ratio that current divides into. The copper may be several hundred times higher conductivity than the pencil lead, and that's the ratio that the current divides into.

Now let's think about magnetic fields, iron and free space. for the same magnetic force across a path containing both free space and iron, the magnetic push makes flux flow in both free space and the iron in the ratio of their permitivities - with the iron winning by a ratio of several thousand. One way to look at this is that a magnetic field would rather be inside iron instead of free space by that amount.

However - you knew that was coming, right? - the ratio is not infinity, so there is always some magnetic field flowing in the free space path. If you have a transformer with iron having a relative permittivity of 15,000, then 1/15,000 of the magnetic flux escapes into the free space around the iron.

It gets worse. Iron and all other ferromagnetic materials have both non-constant permittivities and saturation. Iron with near zero magnetic field has a high permitivity, and this permitivity falls off as flux increases, and at some point the permitivity decreases toward the unity of free space. This decrease is saturation. For low magnetic fields, iron and other magnetic alloys retain their high permitivity. So a conservatively designed transformer with low flux levels in the iron leaks some small amount of flux by the ratios of the supposed permitivity. Push the flux densities higher and the iron gets less "magnetically conductive" and leaks more flux out into free space as its permitivity goes down. So more flux leaks out the harder you drive the iron.

Then there are the air gaps. We don't make transformers out of continuous loops of iron. Sure, that's the idea of a toroid, but as a practical matter iron core toroids are wound from strips of iron in a spiral, the idea being that the air gap has a very broad area from layer of iron to layer of iron and it doesn't hurt too much. Toroid leak a little all the way around. E-I cores leak significantly from each but-jointed E and I. Interleaved stacking helps with this, but there is a leakage zone at the end of each window where an E and an I are butted.

Belly bands are helpful with the E and I laminations in forcing the leaked flux from having an easier time in free space. the high conductivity copper bands helps make the amount of free space available to the escaped flux small, and so it's a higher "resistance" to escaped flux. Hence, less flux escapes.

[GainFreak]

Toroid leak a little all the way around.

I read that if a toroid is not wound properly (whatever that means) it can emit stronger than an EI iron. Somebody wrote in forum that his Tele's (not shielded) singles are picking up noise (kind of combination of buzz and hum) couple of meters from an amp that has a toroid inside (in a steel case).
Also I had this rack amp with a toroidal power transformer that was noisy and no matter what I tried the hum/buzz was still there - not much but annoying. It looked like it was inducing hum in the wires coming from the volume pot. As a last resort I started rotating the PT in place CW and CCW and found a sweet spot where the hum was almost completely gone.

[catalin gramada]

I'm out of subject but think is not a matter of winding but more how the core was formed. First, steel band should be serious pretensioned during rolling process.This mechanical stress destroy grain orientation structure of the steel band. It need a controled thermal process to be done to recover the steel proprieties. That means precise manufacture and time.