Theoretical: distributed gap concept applied to tapewound core

[multi-volti]

Ferrite and powdered iron toroids have an effective distributed gap.

Distributed gaps are used in some industrial transformers.

I unwound a tapewound toroid I found in an old piece of test equipment that was discarded and was surprised to find it had a gapped tapewound core. The whole thing had an epoxy coating, which is typically only done with silicon steel tape and not other alloys.

Another smaller one I unwound had no gap.

I got the harebrained idea to try cutting two gaps partially (about 2/3) the height of the core on opposite sides (Dremel tool and 0.023" diamond wheel), 180 degrees apart on opposite sides (top @ 0 degrees, back at 180 degrees, for visual reference).

I debated whether I could cut deep enough to go all the way thru or if I should...I figured the epoxy would keep the steel tape from unwinding, but two partial cuts ended up being the way I went.

Any magnetophiles have knowledge about the effect of an incomplete gap (not assumptions- I have made too many myself already!).

This is about the physics, not whatcha gonna do next & how.

I do have a toroid winder in the garage I took apart many years ago to 'modernize'. This might be the year...

Thanks

Murray

 

[multi-volti]

OK, a transformer engineer just told me I'm on a valid path...depends on what I'm trying to control.

Now to remember where the parts for the winder go...been 15-20 years. Or hand-wind...

 

[Elvee]

Ferrite and powdered iron toroids have an effective distributed gap.

Powdered iron, yes.
Ferrite, generally no unless it is FPC, which is not very common.
Normal, sintered ferrites are a complete, homogeneous material and can achieve extremely high permeabilities

Distributed gaps are used in some industrial transformers.

I unwound a tapewound toroid I found in an old piece of test equipment that was discarded and was surprised to find it had a gapped tapewound core. The whole thing had an epoxy coating, which is typically only done with silicon steel tape and not other alloys.

That part isn't very clear: when unwinding the tape, did you find regularly spaced "blank" segments?
That sounds extremely strange.
Or maybe you had a cut C or double-C core with a gap (they are very common)?

I got the harebrained idea to try cutting two gaps partially (about 2/3) the height of the core on opposite sides (Dremel tool and 0.023" diamond wheel), 180 degrees apart on opposite sides (top @ 0 degrees, back at 180 degrees, for visual reference).

I debated whether I could cut deep enough to go all the way thru or if I should...I figured the epoxy would keep the steel tape from unwinding, but two partial cuts ended up being the way I went.

Any magnetophiles have knowledge about the effect of an incomplete gap (not assumptions- I have made too many myself already!).














Another smaller one I unwound had no gap.

I got the harebrained idea to try cutting two gaps partially (about 2/3) the height of the core on opposite sides (Dremel tool and 0.023" diamond wheel), 180 degrees apart on opposite sides (top @ 0 degrees, back at 180 degrees, for visual reference).

I debated whether I could cut deep enough to go all the way thru or if I should...I figured the epoxy would keep the steel tape from unwinding, but two partial cuts ended up being the way I went.

Any magnetophiles have knowledge about the effect of an incomplete gap (not assumptions- I have made too many myself already!).

Not perfectly clear either: did you leave a part of the magnetic circuit ungapped?
If it is the case, the low induction permeability will remain relatively high, but the small part of iron remaining will concentrate all the flux and will saturate early and abruptly, leaving only the gapped inductor in service: you have built a peculiar type of swinging choke

 

[Mark Tillotson]

You don't want flux concentrations in a core if you can help it, saturation is very bad news, in fact avoiding saturation is often the reason you want the gap in the first place. With constrictions the inductance will suddenly drop on signal peaks, causing odd-order distortion.

And a partial gap hardly has any effect on the relunctance either, as flux can bypass the high reluctance of the gap

 

[multi-volti]

Thank you for the reply, Elvee. Your answers are always informative & inspirational.

What I meant by unwound was that I took all the magnet wire off some mystery toroids I had no use for as they were.

One was a 25 VA 115 V 400 Hz power transformer that is no longer in the mfr's catalog. So unwinding it gave me a turns count for that voltage & frequency as a reference.

My goal with that one was to attempt cutting a gap with the thinnest tool I could find.

If I could do so successfully, I considered gapping a wound toroidal power transformer for an experiment, but they usually winding turns over the full circumference...hence experiments with an empty core.

Goal is to explore introducing some 'tolerance' for DC imbalance in a PP OPT of toroidal construction.

Never discard anything until confirmed absolutely useless.

The next mystery toroid (a pair, actually...I was destructively ambitious) was already gapped...approx. 0.5 mm.

Some small unused old (70's?) inverter transformers almost got scrapped until I realized they had C-cores...Arnold AH-7 0.004" Silicon steel. Bobbins are eluding me for the time being.

 

[multi-volti]

A 2nd opinion by email compared my partial cuts to the same effect of a composite core toroid made of a gapped & ungapped toroid together...other that that allowing much more quantifiable path & gap lengths.

 

[multi-volti]

Thank you, Mark.

With the end goal being some tolerance for slight DC imbalance, along with effort to achieve balance.

I'm hoping by using a large enough core with partial or complete gap, I can still manage enough primary inductance.

Winding multiples will be incentive to get the winder back together.

 

[multi-volti]

Another response from a core mfr. said valid thinking, but pay attention to remaining cross-sectional area flux density.

So, a large enough core sacrificing a little as a trade-off gives me some room to experiment.

Of course there are easier ways...but I don't seem to stay on paths very consistently.