Flybuck Cockroft-Walton

Bill Sloman wrote in news: snipped-for-privacy@googlegroups.com:

Entire production runs get done that way. That is why there is a hole down the middle of them ALL.

In fact, only a few versions sport steel spring clamps. Most do not.

Bill Sloman wrote in news: snipped-for-privacy@googlegroups.com:

Some get pressed pretty accurately, but still get ground.

The problem is that some (most) use a very rough grind and it yields a por surface quality on the mating faces.

The polished core face versions are just that... polished to optical quality face.

You just posted jacking off at the mouth about spring clips when most do NOT use them.

Perhaps your experience is a LOT less than you would like to imagine. I have years of miniature transformer winding experience.

Bill Sloman wrote in news: snipped-for-privacy@googlegroups.com:

You are an idiot. "Page 8".. Sorry, but I do not need to visit your core page.

And just a note... you also need to place a spacer in the center post gap to fill said gap, even when using the clamp spring variety, which they are of course, not all designed that way.

The clamp springs are needed to insure the machined surfaces strongly mate. They not only insure delivering the spec'd result, but maintain result throuhout the part's lifetime.

Winfield Hill wrote in news: snipped-for-privacy@drn.newsguy.com:

Never disputed what the clamp is for, nor its use or operation. I merely mentioned that if a (factory) gapped core is used, the center post id what they mil to produce the gap, and a spacer is needed to keep the core from splitting due to the clamping force being in the center, while the mated faces are on the edge.

You, unlike Bill, are not usually one to express a lean toward guessing about someone. I have already stated that I have considerable experience with pot cores and that includes both spring clamp captivated versions as well as free cores requiring the customer to mate them. It is obvious I spoke about free cores and I never once mentioned clamps for fixation. I responded to Bill jacking off at the mouth about them.

Clearly, the transformer in this case is far too small for spring clip incorporation. Held by transformer tape and transformer varnish. Good for many Gs in vibe. The core would break on impact before the gap would loosen or change. In fact, to add a gap, a new transformer would have to be wound, just to explore and characterize the issue.

Gapping the core has zero effect on interwinding capacitance

Nobody runs their cores anywhere near close enough to saturation for it to make a difference.

Every transformer has a self-resonant frequency, set by the inductance of the windings and the interwinding capacitance. That frequency doesn't change no m matter how you drive it.

Whether you see ring artefacts on the output depends on the way you drive it.

So what. Baxandall class-D oscillators are mostly driven by MOSFET switches these days (though they weren't around when Peter Baxandall invented the circuit) and I've yet to see ringing on the switching transitions.

Each winding on any transformer has an inductance. Transformers are coupled inductors.

Gapping a core does decrease the coupling between winding on that core.

For the ferrite cores I've worked with the coupling coefficient goes down from about 99.9% to about 98% between ungapped and gapped cores.

For the cores I've worked with, your 25 micron tape across the centre and the rim introduces a 50 micron gap which would drop the inductance by a factor of three. That's quite a lot.

You have to know the inductance of each winding, the mutual inductance between each pair of windings, and the parallel capacitance associated with each winding.

They aren't hard to measure, but it does take some time to do it.

Not of a lot the sims that get posted here include all the detail.

Why not?

One of the many ways the less conscientious produce misleading sims.

The certainly isn't a hole down the middle of all of them.

The part numbers that include a "D" do have a centre hole, but "J" means no centre hole.

That data sheet is for an RM core. Older, circular pot cores offer differen t clamping arrangments, but a bolt down the centre of pot core is never a g ood idea, though it can be a handy - quick and dirty - solution if you have n't got the time to do it right.

Obviously not. But we were talking about pot cores, and I produced a documented example.

You haven't described anything in detail, but want to claim to be expert with "other kinds of pot cores" which you haven't bothered to identify.

This probably means that you haven't seen what you are talking about for years, and can't remember much about it.

Sure. It's bad practice. but it happens. The central hole through a pot core isn't there to accommodate a clamping bolt, but rather a screw-in adjustor.

see page 9 for an example. You can use it fine-trim the inductance of the coil in the pot core.

The fact that you don't understand it makes you stupid - not me.

The gap is part of the transformer design. Fiddling with it successfully after the transformer has been designed demands a rather more complete understanding of what it does than you seem to have.

Or slapping a capacitor into a circuit to stop it oscillating. It sometimes works and lazy people always try it to see if it solves the problem they think they've got.

You are confident that you know it all already. Pity about that.

I've not seen it done, but the gear I worked on wasn't designed to survive high amplitude vibrations.

But imagines that it is an inferior substitute for the bolt through the central hole which isn't always there.

The centre post is ground down - not milled. Milling doesn't work with ferrite cores - the material is brittle.

And nobody who knows what they are doing centre-clamps a pot core, even when the pot core being clamped has a centre hole (which most of them don't).

You've got some experience, but not all that much insight.

Amateur night. If the core and the winding are small enough this could work, but it's not an approach that lets you predict when the structure will fall apart.

Obviously. So why are you telling Phil Hobbs that he should rip his transformer apart and add a gap?

Sintering involves sustained pressure at elevated temperatures. It depends on getting the atoms in the individual particles of the powder being sintered to diffuse into adjacent particles.

The grind is going to be good enough to let the core halves mate well enough to work.

Optical quality is submicron flatness because the wavelength of visible light is about 0.05 micron (0.038 to 0.074). That isn't an issue with pot cores.

They aren't used as much as they should be. At lot of stuff that gets into production is more improvised than designed.

But can't produce any detailed information, because it's all commercial in confidence. The great thing about military and commercial secrecy is that it shields sloppy operators from informed criticism.

Bill Sloman wrote in news:516170be-e9f1-4b10- snipped-for-privacy@googlegroups.com:

The inter-winding cap is what he says is causing the ringing.

If as I first suggested, it is not possible to reduce the cap number, then the gapping fix could work. That is all I said, and then you went off on a stupid tirade about how it will have no effect on this or affects that.

The goal is that it could affect the ringing problem in a positive direction.

Your goal is apparently derision, because that is all you have provided.

Bill Sloman wrote in news: snipped-for-privacy@googlegroups.com:

A ringing oscillator might. But a gapped core might dampen or stop the ringing, which is the goal.

Bill Sloman wrote in news:4a2ecb33-95de-492f- snipped-for-privacy@googlegroups.com:

Show me a sim with a transformer that has a single turn feedback winding where the operation gets modeled correctly.

It was a problem 15 years ago, and I doubt anyone has addressed it yet.

Bill Sloman wrote in news: snipped-for-privacy@googlegroups.com:

You are an idiot. A bolt is required when one does not want to rely on solder joints to hold a device through vibe qual. Also a good choice over solder joints for production too. Yep... not just us... the entire industry does it.

Your brain was quick and dirty. Nice try though.

Bill Sloman wrote in news:6dddc0c1-d4c3-4712- snipped-for-privacy@googlegroups.com:

The fact that you do not get that gapping the core of a transformer is not about a measured gap, but more about a tuned transformer/oscillator pair is a huge tell. I am not stupid and Biden did not sell access to his president. Nice try though.

Bill Sloman wrote in news: snipped-for-privacy@googlegroups.com:

Nice try, Billy.

Bill Sloman wrote in news: snipped-for-privacy@googlegroups.com:

Yes, a Horizontal MILL operation using a grinding head. I have seen the grain direction and know what was used. I am also familiar with the media. You seem to always accuse otherwise.

You are just a putz trying to nit pick stupid shit. And oh look.... ooops, you even failed at that. Entire arrays of cores get ground all at once on a flat bed. The gapped cores, however, have to be hit with a small diameter head that can dress down into the core half. One at a time. Main reason they cost more.