Toroid winding question.

Weak but some will stick. (hold against gravity) Brass screws, probably from China, I assume with a fair amount of iron. I use to have production sort out the magnetic ones, but now we get screws from a better supplier. I had some issues in the past with the field homogeneity of our optical pumping apparatus... I found the magnetic screws, but that was not the problem.

Brass is not a very well defined alloy. (I assume you know that.)

George H.

Well, I wasn't aware of Iron in brass, but Wiki says there can be.

Cheers

Putting an odd number of half-turns on a pot core can get pretty exciting if the gap is in the centre post. The odd half turn sees the full permeability of the outer ring of the pot core, which can easily saturate.

Cheers

Phil Hobbs

There's no such thing as a half turn (and nothing else) on a pot core: the flux goes up one side, down the other side, using the core as a ferrite bead. Spooky behavior ensues! (Core saturation, high radiation, really strange inductance values.)

It is nonetheless possible to implement such a thing: you need a balancing winding that enforces equal flux through both sides of the core. Ye Olde Unitrode appnotes show the way:

Tim

With that kind of pedantry and hair splitting, anyone would think you'd been spending time in Oz. ;)

You put an odd number of half-turns on the bobbin, and the behaviour is very different. I think that's a perfectly reasonable way of putting it. How would you describe the wiring configuration if not that way?

Cheers

Phil Hobbs

One of my ex-guys, a self-declared magnetics pro, did that once, a winding with an extra half turn on a pot core, in a power inverter application. It got really hot.

I need a replacement for this 15-year old custom dc/dc converter transformer:

and the header is no longer made. I want something that will be available for 10 years at least. Pot cores are multi-sourced, so I need the windings. The old design had a lot of windings, a DC-DC converter with multiple outputs. Nowadays, I could generate a single isolated output and use cheap load-side switchers.

I was thinking about making a little PC board as the windings... just drop it into the pot core. If I go 1:1, I'd just need 4 pins, maybe 5 or 6 turns per winding, and the windings could be done on a 2-side board.

Coilcraft does similar stuff.

Huge power ratings!

For some reason, the good meters of yesteryear didn't trust brass much, the nonmagnetic screws were usually copper-nickel (nickel silver, Monel, German silver...). I've got a surplus stash of those somewhere...

I've also seen molybdenum used as steel-substitute, when the magnetism issue was REALLY important. That got complicated.

Printed windings do make sense, but you want to get them from a printed circuit shop that specialises in making them.

The specialists put relatively thick copper onto very thin substrates, and can get copper fill factors that come close to conventional wound coils. Coilcraft would be using something like that (as can be seen from the pictures).

It would be a nice way to do banked windings (for lower parallel capacitance across the whole coil).

Two-sided boards would be a bit too fragile to be attractive.

We used nylon screws and nuts to hold the pot core halves together.

That works for a simple functional explanation. Sometimes it helps others to see things explained several ways. :)

Tim

They tend to get weak and brittle after awhile at higher temperatures, though. Aluminum is reliably nonmagnetic for this sort of use, but expensive.

Cheers

Phil Hobbs

I've found that nylon fasteners "creep"... stretch and lose their tightness.

Been like 30 years since I've done any transformers, but I pretty much settled on pot cores back then, with the spring clip kind of PCB mount. ...Jim Thompson

True, but ours were also potted in Scotchcast. No problems.

of

sorry I wasn't more specific....

A metal screw through a torrid can cause a problem if the screw forms a com plete circuit shorted turn with the top and bottom of the metal chassis, yo u have a shorted turn through the chassis. Obviously an issue once you thi nk about it. The screw itself isn't an issue.

I understood you perfectly. I guess I drifted off topic by mentioning pot cores. We found it beneficial to use nylon screws to help prevent over tightening the gapped pot cores. They held quite well and the transformer was potted in semi-rigid material.

My apologies for misleading you.

Yes, the toroid gets definitively torrid when that happens.

Jeroen (Sorry, couldn't resist that one) Belleman

magnetics I can't see being a problem, but thermally it's also more insulated, and less robust, could you go with a 7 strand rope of thinner enamelled wire?

This is what I did in case anyone else needs something similar...

I laid out some Kapton tape on the bench, sticky side up and taped down at the ends. I laid some flat tinned copper braid down the middle, taped the ends down and ran a screwdriver shaft along it to stick it to the Kapton. I then cut into the Kapton at each end and folded it over the braid, pressing down to stick it, then cut the ends and trimmed the excess Kapton.

So I have nice flat flexible insulated braid which works well. Fortunately, I'm not making many of these.

Can you buy Kapton insulated braid? I couldn't find any.

Cheers

Try the generic term, polyimide. 200C rated enamel, commonly used for demanding applications: motors, inverters, etc.

Downside: I forget if there's a "sold-eaze" version of the stuff. The chemical strippers I understand are kind of nasty...

Tim