Today pcb was completed and tested. Unfortunately coil inductance almost doesn't change when inserting magnetic core. Driving coil like in buck converter also didn't work, voltage change is faster than my comparator bandwidth (full swing of voltage is 450ns, comparator output response is 300ns).

Some screenshots from oscilloscope, CH1 is pwm signal, CH2 is connected to pin3 of Q1 (high drive side of coil in this test):

My problem could be caused by measuring voltage on coil connections, not current through coil. But that would be only a problem when measuring inductance. The real problem: why that waveform on second screenshot doesn't change with inserted/removed core? It should change atleast a little, but it doesn't. Only during moving in/out, which is expected (induction).

If anyone sees an error here, comments are very welcome.