another strange idea

Hello John,

The UHF section of all those VHF/UHF 4:1 splitters had "phenolic transformers". Not exactly transmission line but close. At least all the ones I mounted in my teenage days were that way. Lower end or higher frequency reflectometer bridges (SWR meters) often contain directional coupler strips made from FR-4.

The E-core thing, well, I am not a fan of that. Seen to many come apart on those desert truck rides. I do it with double-holers.

I see no reason why multi-layer couldn't do it. Except that FR-4's RF properties are, as Archie Bunker would put it, lousay.

Regards, Joerg

I have designed one on an integrated chip, which is more or less the same thing, just in a smaller scale

Regards, Mikko

Hi John

I have made transformers using spiral turns for the winding - 2 times 2 on a 4 layer PCB (FR4). These were however for about 10MHz

I can mention several papers if you are interested. None however for ferrite planar transformers (but wont that be very lossy due to the limitations of the ferrite?)

Regards

Klaus

Philips wrote an app note about using windings in FR4 for a small planar power transformer. It's probably on the Ferroxcube site now ( they've taken on the former Philips ferrite lines ). Includes example pcb foil layout.

Graham

Hello Klaus,

Ferrite is a wonderful material. Remember the old VHF/UHF splitters? The VHF section usually consisted of a double-hole ferrite balun and is useful up to 300MHz.

Personally I am just not a great believer in split cores or anything that relies on adhesives to hold core halves together. At least not for stuff that ships into fields where handling can be a bit rough.

Regards, Joerg

Hi Joerg,

I've built quite a few planar converters, using a combination of clips and glue. My last dc-dc converter had a planar xfmr and a planar coupled inductor, and we've made about 30,000 so far. Some of these get carted about on the back of large trucks )in theory with air riders, in practice without), and while there have been problems, the cores have been OK for 5 years now. Some of the thermosetting epoxies I tested were far stronger than the core, and the operating temperature rise hardens them.

I prefer thermosetting epoxy to glue the cores together, and some form of CTE-, tolerance- and vibration-absorbing structure to hold it to the PCB. its usually a good idea to space the core slightly away from both sides of the PCB, because the permittivity of MnZn ferrite is so high

Of course the production guys then changed to a rather crappy glue that softens when it gets hot, because it was easier to apply. It remains to be seen if this was a wise idea (they no longer do it), but I suspect not :)

Hi John,

I've seen some papers out of South Africe detailing integrated magnetic structures, where the ferrite material is formed around the windings. J D Van Wyk springs to mind, but IMBW.

There have also been a bunch from IIRC Korea about coreless transformers (spiral windings on FR4) and various uses as gatedrives etc. And some spinoff papers looking at the effect of placing Cu and/or ferrite plates on the outside. air core gatedrive transformers sound like a great way to create mysterious blowups, along with maximising EMI emissions and susceptibility.

any transformer you can make with an E-core, you can make with a planar core (E-E or E-I). All they have done is reduce the height and increase the thickness of the cores. Glued cores can have a lot less permeability than a single piece of magnetic material (eg toroid, n-holer), unless they are squeezed very close together. several tens of microns can have a noticeable effect. very high creepage and clearance can be done, but is a pain (buried windings, coatings of gunk, fancy bobbins etc)

Cheers Terry