Thane Heins BI-TOROID TRANSFORMER

[Jack Noskills]

Well, my prototype is a piece of crap put together using electrical tape, not sure it would survive the rubmle in the mail lol. I have been working in a post office as a young lad. It is better if someone just builds another one, coils were taken from trafos made by chinese company named Jutec.

Any type of coils will work as long as they are rectangular so you can make alternate path between secondaries. It was just some beginners luck that those bigger E-I plates from Amiga trafo fitted inside the coils from Jutec trafos.

I am considering to buy some cores from magnetec to get more powerful version. If I do that I will then get me a new voltmeter to get more accurate results than just a gut feeling. But maybe someone could beat me to it so here is my idea of design how I would make it:

I am thinking about using two M-283 oval cores, then a toroid that fits on top of that, maybe even on both sides. Primary wind would be of 5*5 rectangular uninsulated Litz wire, maybe 0.25 mm thick thread size for start that connects those oval cores in the middle. This would give about 65 mm of space in the core, so with one winding I would get 4*65*5 turns, 1300 total, then several layers and connect individual strands, experiment with bifilar windings at the same time.
Secondary windings connecting oval from other sides to toroid, also 5*5 rectangular Litz, a bit thicker, lets say 0.5 mm strand size and bifilar winds. Again connecting individual strands and enough layers until output voltage matches input.

Hmm, broli maybe you could make another nice picture using ovals and toroids ? I put the spec of M-283 here, there are smaller and bigger ones available. This one costs 60 Eur from their web shop. The toroid(s) I am considering would be of Magneperm having 450000 permeability, but other values will also work. Trafo build entirely from magnetec cores would be quite expensive, maybe iron as primary would also work giving enough power. A payback time of atmost one year would be reasonable. So 200 Eur for cores, 1 kw free out and this could be sold for 1000 Eur to get a nice margin. Thinking of a final product, hehe.

And a note, those magnetec cores are always cased. It could be essential that those ovals touch each other, or it could not be. So more experimenting is needed here.

[futuristic]

Hi is there any way that you could barrow a camera to take some pictures and post it here?

It would be really beneficial to actually see the coils and whole setup.
I have a hard time imagining how you quickly changed from normal to heins type trafo.

[Jack Noskills]

Heh, simple things are most beautiful, principle works with software development (my profession, for now) and also with hardware.

Look at the 3d picture, just imagine fourth coil in the middle right above/below the primary. Separate output for that so I got two different outputs connected to standard wall outlets. If I connect to one output then I dont use the other so it has no effect on test result. Then I keep swapping cable from electric saw, one test under 20 secs hehe.

Taking picture is tedious process when using mobile, not enough user friendly to transfer it to computer so I rather avoid it.

[broli]

Hmm, broli maybe you could make another nice picture using ovals and toroids ?

I'm pretty sure this is not what you had in mind, but I came up with it trying to visualize yours. In fact this has a special property that I have seen no other BTT variation have. The fields of the two secondaries don't oppose but add up in this setup. Does this change, improve or detriment anything? No clue but I just put the design out there and might even build the setup if the wallet allows it.

red is primary, green are secondaries. Also made the toroid thicker than the oval core.

From Magnetec this could be achieved with:

Oval: M-284  >60000µ @ 50Hz 100 EUR
Toroid: M-113 ~30000µ @ 10kHz 60 EUR

[Jack Noskills]

I am not sure this design would work. Look at the flux from primary, the other secondary sees flux from primary and opposing flux from other secondary, they dont add up. Here is a 2d picture of what I mean.