interesting transformers

They aren't - for an ungapped cored. The coupling factor would be 0.997 which is pretty ordinary.

At least one of them has a printed winding, but you don't seem to have noticed.

Too bad they didn't think to interleave the output winding terminals.

RL

There are likely geometric tangles on the PC that restrict crossovers. It must be interesting to design multilayer PCB windings like this. There are thermal considerations too. I could take one apart and x-ray it.

I've considered making transmission-line transformers integral to PCBs, in which case they have no terminals as such. The geometry issues are tricky. I've always used coax so far.

That's what blind via's are for.

People have been doing it for some twenty years now.

It's a specialised area of printed circuit manufacturing - you need thin substrates and thick copper to get a decent fill factor and tolerably low winding resistances

Of course you have.

You don't have to think nearly as hard about what you you are doing. Twisted pair is another sort of transmission line, but harder to terminate neatly.

Guys, in all fairness, I think that Jan was kidding.

boB

Tolerance issues are likely as well, especially if you use more than one board house. The epsilon of 'FR-4' can range between 3.8 and 4.5 depending on the vendor.

Interesting transformer, if a bit on the pricey side--with all those windings you could do some useful things, such as isolated supplies, supplies riding other rails, and so forth.

Cheers

Phil Hobbs

Yes, I was thinking about playing with windings. I'd have to violate their rule about "windings to be connected in parallel".

I've ordered some to play with.