From my point of view, the 10000uF of the capacitors are not sufficient 
to supply the current the pulse draws from the powersupply. Take in 
mind, that for the time the MOSFETs are powered the supply sees some 
what like a short circuit. So the internal resistance of the pwr supply 
is very high compared to the inrush current of the coil.
And the capacitors are not able to support the pwr supply as needed.
You have to improve:
First the pwr supply, so that it can source more current in a short 
period of time to withstand the short of the pulse from the coil and to 
recharge the capacitors.
And second, add more capacity to the capacitors.
Search this forum for "powersupplies" to find the correct formula to 
calculate the capacity. The user MaWin has plenty of knowledge to this.

Hope, that will help you further.

sirlakidis giannis schrieb:

> the problem I presented is that the voltage +12 volt creates ladders ..
> Look at the picture 2 ..

What did you expect?

Simply simulate this piece of shit and you'll (hopefully) understand 
what happens.

Hint: It's not required to simulate the whole driver stage to understand 
what is wrong in this design. Replace it by simple pair of complementary 
driven switches for simulation. This will speed up the simulation and 
allows you to focus the core of the problem. Keep in mind: the impedance 
of real world power supplies is never zero.

Did anyone ever calculate this shit before it was build? I'm afraid 
none.

c-hater schrieb:
> this shit before it was build?

Ack!