On last week I played with coils, and played a lot
Now I have 18 coils, all what's needed to complete one modular electromagnet:

As I'm too lazy to wind them by hands, I had some extreme prototyping experience ->

Yeah, it's winding machine made from a building kit! Works surprisingly well!
I would design something more reliable later, of course, )

Since I have some troubles with accessing 3d printer and drilling machine for now, I haven't completed modular electromagnet yet, however I fixed problem with transistor's opening! 

This is how schematics looks now. Nothing changed in current monitoring circuit, but now we have IR2117 (or IR2118) MOSFET driver here. What it does - it provides extra voltage over voltage on capacitor C3, so voltage difference between Gate and Source stays positive then needed even then transistor is in high side switching mode. That means, that transistor is fully opened, outputs that needed and dissipates less heat as well, so it wouldn't require hardcore cooling : )

Also, there is additional transistor Q2 - while opened, R8 and R7 work like voltage divider and voltage on IR2117 "IN" pin is close to +0.4V (R7/R8 * +15V), then Q2 closed, we have +15V there.
IR2117 has different logic levels:
- low 0-6V
- high 9.5V-20V
So it's necessary to convert 0-5V output from TTL's to an appropriate level, that's why Q2 is needed.

---

I think, tracing PCB's with MOSFET controllers only is OK for test purposes, so new revision of control circuit's PCB would be later, I don't want to waste time waiting for a delivery. And I want to go camping next week, so modular electromagnets experiment results release is going to be delayed )