coil impedance

It's not talked about much, but a real-world inductor is not a lumped component and can have multiple resonances.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Yes indeed. There is a frequency where the impedance reaches a peak and then it drops. In fact, it may do that several times over.

The first resonance frequency of many inductors is depressingly low.

Jeroen Belleman

A real inductor is a nightmare. Especially a long solenoid. Every turn inductively couples to every other turn with all possible coupling coefficients. Distributed capacitances will be similarly complex.

And then there's the speed of light.

In a bit of googling (I do have a day job) I haven't seen a good sim or measurement of a real inductor. Everybody keeps citing the simple LC formula as if it's true.

We make some laser drivers that include a multi-section home-made bias tee, trying to synthesize a super-wideband high current inductor. Yuk.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

This is ignorant nonsense. The current through each turn does generate field of flux, and that individual flux field threads all the adjacent turns to an extend depends on the distance between the turns and the extend to which the flux is concentrated by the high-permeability core of the inductor.

You can almost always lump them together.

The capacitance between each turn works the same way.

When the propagation delays are long enough to be worth worrying about, you model the inductor as a transmission line. .

It mostly is. I once went to the trouble of modelling an inductor as two closely coupled inductors in series, and got exactly the same result.

I'd been worried about parallel capacitance, but splitting it into two capacitors of twice the size in series didn't make any perceptible difference at all.

John Larkin is very reluctant to design his own transformers and get them wound, and when he posts a LTSpice simulation, his inductors rarely have a parallel capacitance.

LTSpice has built-in models for a huge range of Wurth inductors and those do have parallel capacitance.

John doesn't seem to willing to exploit this resource.

Wide-band air-cored chokes for eadio sets were always a problem because of this. The choke needed a large number of turns to cover the lower frequenies but then the extra capacitance caused internal resonances at the higher frequencies. The stray capacitance and inductance formed a parallel-tuned circuit which meant the choke was virtually a short circuit at the resonant frequency and looked like a coupling capacitor above that frequency.

Some wide-band designs used segmented windings or two very different chokes in series but the problem was eventually overcome by 'progressive-wave-wound' chokes, where the winding head oscillated axially to give a low-capacitance basket-weave effect but simultaneoulsly crept along the former so that the winding was spread out and the stray capacitance between the ends of the choke was minimised.

Edge-wound coils, such as the overwinding of television line output transformers, gave low capacitance and enough separation between the ends of the winding to prevent arcing at high voltage, but they were often deliberately tuned to resonate at the third harmonic of the switching frequency to give a higher peak voltage. They don't generally make good chokes.

I suspect the ignorance is yours.

I worked for a company that built its reputation on the R.F. inductors it designed; these factors were among the many problems they tackled.

Probably not very well. The English language text books I could find were largely useless, and it took the Siemens ferrite data book to get me on the right track. I'm sure that there were people around who knew what they were doing, but in the 1970's and 1980's rubbish like E.C.Snelling's text was all I could find.

What makes you say that about a leading radio company that I haven't even identified.

What is the relevance of that?

A real world inductor is a lumped component - some of the flux lines thread all the turns.

Are we talking about different resonant frequencies or harmonics of a single resonant frequency? In fact, since the permeability of a real core varies with frequency, life gets very complicated.

The resistance of the windings, and the resistance around the current loop in the core means that there's never a short circuit (unless you manage to create a shorted turn).

I think that it has been demonstrated that highest resonant frequency comes from a close packed single layer winding on a toroidal core. Deforming the wire (from round to slightly elliptical) so that you can get a few more turns into that single layer seems to help.

Transformers used in high voltage generators tended to use banked winding.

If you opened up a photomultipier power supply you'd find a perfectly ordinary ferrite core with a custom made coil former that let them spread the secondary along at the core as series of narrow banked windings which they connected in series.

When I found out that you could buy self-bonding transformer wire I had fantasies about winding each segment in succession, putting enough current through each segment to set the bonding resin, then stacking the bonded segments. In principle you could make a series of pancake coils that way, but I'm not sure that it would have won you much.

As you makes the banks narrower, you have less and less capacitance between each layer of the bank, but you have more capacitance from bank to bank as you move along the core.

Don't make trouble. Sloman knows everything, and you don't.

If you ever need it, Coilcraft has a nice super-wideband power inductor.

We tried the dreadful conicals (comicals?) and they were frankly dreadful.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

The fact that he does not know what he's talking about might have something to do with it.

I spent 22 years in England in the high tech end of the UK electronic business, and the understanding of the wound components they used was never impressive.

It makes a point about the local culture. Kibble and Rayner's "Coaxial AC Bridges" had some great stuff about wound components, but interwinding capacitance was treated pretty superficially.

Not exactly impressive. A single layer cylindrical coil is not going to surprise anybody,

Edward Rawde thinks that everybody else shares his problem.

This is ignorant nonsense.

And it underlines the fact that Bill Sloman is able to tell anyone else what they think. One has to wonder how he can do that. Does he have a direct link to everyone else's brain?

Just because the company you worked for didn't understand inductors, it doesn't exclude the possibility that other companies did understand them.

How does that support your contention that an unnamed company, that built its reputation on the excellence of its R.F. inductors, didn't design them very well?

Sorry, of course they do.

Multi-band radios have shorting segements on their wavechange switches to prevent the *parallel* resonance of adjacent coils from 'sucking out' the signal from the coil of the waveband in use. (That was what I was mistakenly thinking of.)

With coils of many turns, all sorts of things can happen and the impedance may go through several cycles of maxima and minima over a wide bandwidth.

Parallel-tuned circuits have an impedance *maximum* at the resonant frequency.

Jeroen Belleman

Yes, a tiny surface-mount transformer and a C-W multiplier is a lot nicer than making and rectifying HV AC.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics