Hi the_big_m,
The transformer patents you mention sound almost like some sort of fractal transformer. The 3 D transformer reminds me of the Meretsky patent, US4210959, "Inductive Device With Orthogonal Windings" but this is certainly not an antenna as such. It does have some unusual properties according to tests. McCreary US2445847 and his other patents are also interesting in this regard. And then there is the famous Wanlass parametric transformer, US3654546 and others.
There are a large group of these cross coil devices and they either were used as a parametric or reactive current control devices, or-- with a movable third coil-- as antenna tickle coils, or variable couplers.
Yes, I used to do a lot of research in the U. of Washington Engineering Library, where I would nail down a computer with my butt and then research until I couldn't stand it any more. If I got up there would inevitably be a student there when I got back. Originally when I started doing patent research in the late 80s the patent database was on nothing but a microfiche reader, so I would go to the library with a list of all the classes I was interested in and scroll through them, stopping to print. I still have a wall of paper from that period.
Now of course most things can be done online. I used Google Advanced Patent, Espacenet Advanced Search, and sometimes the USPTO class search page. Each has advantages and disadvantages. Google Advanced Patent uses an OCR like program to read the patents so a lot of older inventor's names are garbled, and often the program will not give truly all of the patents. Class coverage is also nonexistent before about 1920. USPTO class coverage is perfect but the interface is cludgy. Espacenet covers patents from 80 countries, but only goes back to 1920 in most cases.
Unless I have a certain patent that I'm looking for, I use class searches almost exclusively. For instance, the Daniel McFarland Cook patent was originally found by doing a class search on USPTO of the entire patent class of "inductors" (336), from the start of the patent system in the 1790s-- So I had looked at some half a thousand patents one at a time before running into that one!
It never would have turned up otherwise, since Cook's name is garbled on Google Patent, and the patent doesn't appear with a name or class on Espacenet.
Especially useful are class searches on Espacenet. Espacenet has classes specifically for rotary, solid state, and hydraulic/pneumatic "perpetual motion machines" as well as special classes for all sorts of electrical machines that don't fit into usual categories. The most important of these classes are H02K53/00, for rotary free energy devices, and H02N11/00 for solid state free energy devices. Between these two categories there are about 20,000 patents that claim overunity, with only about 200 of them ever having been discussed in any forum or book. Go to
http://worldwide.espacenet.com/advancedSearch?locale=en_EPand plug in H02N11/00 on the IPC line to have some serious fun :-)
orthofield