NNNnnnnnooooooo! There's so many wrongs about your transformer
1. Core appears to be #26 powdered iron. mu = 75, and it's about as lossy as a steel brick. While it's still possible to get good transformer action out of it, your gate driver will have a hard time dealing with the large magnetizing current. Expect to need hundreds of turns!
2. The windings are opposed, giving maximal leakage inductance. Expect a coupling factor of... maybe 0.95? Which is pretty atrocious for GDT duty.
As for phasing, this is easy to remember:
Observe that there is one opening through the core. You can enter it from one side or the other.
Each time a turn passes through the opening, it makes one "turn". (So a straight wire going on for infinity, but threaded through the core, consists of one turn. The trick is, it makes a "wrap" around the core because, whether or not you've kept the wire close to the core as a single wrap, or it goes on straight for all of infinite space -- it's still making a complete loop around the core, and that's all that matters.)
If you denote one side of that opening as "positive", the other is "negative". Any wires entering through the "positive" side will all have a positive sense, together.
As for making a good GDT, first, hunt for a good hi-mu ferrite core, ungapped. You might have something salvageable from a common mode filter, or a car amp's DC-DC converter, or EMI beads/rings, or something like that. These will all be moderate to high permeability examples, and will make good transformers.
Second, minimize leakage. Leakage inductance arises directly because there is space separating the wires of the windings. The inductance due to this space has almost zero to do with the core -- it's almost independent of the core being there or not. The best way to minimize this is to wind the coils on top of each other, closely (alternating layers), or by using twisted pair.
Tim
Home
Help
Search
Login
Register
