I thought this was very interesting when I first read it.
Given that Bill spent 22 years being unimpressed by other peoples wound components, he should perhaps direct us to an article of his own on this subject so we can all learn how to do it better.
It sure impressed me. I think it's filled with some ferrite loaded epoxy, like the wideband conicals.
As they say, great for wideband bias tees.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
He's just here to make friends.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Am 07.11.25 um 17:06 schrieb john larkin:
We used conical coils made by Piconics in our 10 GB/s fiber optic modules in 2004. Performance-wise, they were simply the best. It took some work and several SMD parts to get 10 GB/s eye diagrams that were even nearly as good. But the cost of $/€ 35 per inductor was simply too much.
<MCL had nothing that could compete. I told the MCL FAE that they probably could sell one inductor or even bias tee for each ERA-X if they performed well & the price was acceptable. That triggered big $$$ signs in his eyes, on the spot. :-) But it took some years until the Piconics patents ran out.
And yes, you get usable simulations from the s-parameters that are measured by a VNA, including all these series and parallel resonances. It takes HFSS or Keysight ADS or ... to simulate that.
3D electromagnetics included. LTspice won't do.But if all you've got is a hammer, the world looks like a box of nails.
Gerhard
Prices are better now that the patents have expired.
But those tiny leads are awful. Hard to solder, super fragile, and they can even dissolve in the solder.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Am 08.11.25 um 01:06 schrieb john larkin:
I've seen someone trying to repair a golden bond wire by soldering under the microscope. The bond wire dissolved in real time. :-)
Gerhard
Edward Rawde has posted quit enough nonsense here, to let me be well aware of his intellectual defects. I'd probably be happier if I was ignorant of them.
It's called putting forward a plausible hypothesis. Edward Rawde isn't into that kind of subtlety.
If you are as dim as Edward Rawde, you can think that that is what I'm pretending to do.
Does Edward Rawde have a brain that it would be safe to link to? Words are what we use to swap ideas between brains, and if you don't understand precisely what particular words mean you can have very strange ideas about what might be being said.
I spent three years at EMI Central Research. Some of their emplpoyees did get stuff wrong, but they were very high level misunderstandings. I got the staff briefing on their nuclear magnetic resonance imaging system, and asked them why they weren't using super-conducting magnets, and got told that that was a naive question.
John Larkin doesn't understand what he is doing all that well,but if you keep experimenting and testing for long enough you can come up with pretty impressive products. Not as good as they could be - but quite good enough.
So what. People as good as Brian Kibble are thin on the ground, and British society wasn't great at delivering good education to kids who didn't have at least middle class parents.
That was a me-too article from the period. No mention of parallel capacitance at all. If you thought that it was very interesting when you first read it, you hadn't read any of the ferrite core data books around back then.
Coil-winding machines make that relatively easy. When I worked in a lab that had one, I did too.
The Cambridge Instruments coil-winding shop on the production line had some very experienced coil winders, and they could wind finer wire than I could without breaking it. Sometimes that mattered.
I'm sure you told them what fools they were.
A motor like that can sometimes be salvaged by winding it with thicker wire and running it from a spare transformer*. It's crude, but worthwhile if it saves an expensive piece of equipment.
What I actually posted was.
"I spent three years at EMI Central Research. Some of their employees did get stuff wrong, but they were very high level misunderstandings. I got the staff briefing on their nuclear magnetic resonance imaging system, and asked them why they weren't using super-conducting magnets, and got told that that was a naive question."
You've chopped my text - which made it perfectly clear that I wasn't going around telling people that they were fools - so that you could post something gratuitously offensive.
You seem to be almost as old as I am (and I'm 82) so I can't tell you to grow up. The joke in that story is that modern nuclear magnetic resonance imaging systems all seem to use super-conducting magnets. I thought that everybody knew that.
I thought you considered gratuitously offensive posts were the normal way of replying.
I designed and wound this myself!
That thing takes 48V from a wall wart and makes 1400 volt pulses at MHz rates into a Pockels Cell.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
It is true that when I first read that article I would have read almost nothing else about ferrites. But let's assume I had read the ferrite core data books you mention. What differences would I have encountered between the information in that article and the information in the ferrite core data books?
There's a Siemens Ferrite data book here:
Our newest microwave doesn't have a turntable and seems to not need one.
It's a Panasonic with great thermals and a repulsive user interface.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
X112 was a multi-tile multilayer proto board. Several people contributed circuits that they wanted to try, so I threw in that HV thing. I wanted a monitor output and a pot that was linear on voltage.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Artistic license.
If we had to make a lot more, we could have one of our guys 3D print a winding mandrel or something. Or more sensibly, send it out to a coil winding shop.
The big deal on that inductor is cooling, which the gap-pad is part of. Peak currents are ballpark 50 amps and skin losses are huge.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Nothing dramatic. Data books are usually more comprehensive, and less worried about keeping the reader's attention.
The Siemens ferrite data book I did find useful around 1978 included a section on the transformer equation
V1 = L1.dI1/dt + M.dI2/dt V2 = M.dI1/dt + L2.dI2/dt
where M = k. square root 0f L1.L2 and k is the coupling coefficient between first and second winding, typically about 0.999 for an ungapped pot/RMcore, dropping to about 0.98 for a heavily gapped core pair.
If you have more than one winding you can have different coupling coefficients, though they are rarely dramatically different. LTSpice lets you specify them coil pair by coil pair, if you want to. Section K in the help files - Mutual Inductance.
The transformer equation doesn't seem to have made it into the 1983 data book. I hadn't run into it before then, and I wasn't able to persuade Win Hill to put it into the third edition of "The Art of Electronics". It works for me, and it lead to a useful discussion on coupled scanning coils at Cambridge Instruments a few years later - my boss hadn't run into it before then either.
I've no idea when it made it into Spice - I didn't run into it there until LTSpice became available, which was a lot later.
Have something to add? Share your thoughts — no account required.
Ask the community — no account required