Or use some 74HCxx schmitt triggers to at least
The previous schematic I posted already had a scmitt-trigger (though it was interverted 
) so that point was clear 
Latest plans incorporate a 74LVC1G17 in sot23-5. This time I am sure it is non-inverted .
When I laid my eyes on the 3D view the thought to employ an attiny1616 with some shiftregisters (74HC595) has crossed my mind. When I use an uController to control mosfets I use a 22R resistor to limit the inrush currents and a pull-down resistor. So that is 2 resistors and a freewheel diode per mosfet.
This analog variant now has one sot23-5, one resistor (0603) and 1 capacitor more than that. The costs are about the same. So the extra amount of components isn’t that much of a deal. And considering I dont have to pull my programmer out of a drawr… analog has my preferences.
I am however slightly worried about the contact bouncing. I will propably need an extra RC filter (atleast) for that and at that point I may be inclined to take the uController approach instead.
Therefor I am going to breadboard this circuit and perform some tests. The partial simulation I managed to do in Kicad were looking good. I have opamps laying around to make a schmitt-trigger with. With an arduino I can monitor the pulse length on the output side. I only can’t monitor how steep the flanks are. I am looking forward to it.
Bas