See attached schematic.

The end result is a latching relay schematic. When start is pressed it latches, when stop is pressed it unlatches.

I’m pretty sure this will work as intended but I figured I’d get a second opinion.

When you press start, it will activate the n-ch mosfet which in turn activates the coil which triggers the contacts and its a standard latch.

When you press stop, you turn the mosfet off, the 100K in series with the coil will de-activate the coil which cuts the 5V source from the circuit.

It would be simpler is I just used a NC push button to break the latching but it turns out that there are not many NC tactile buttons that are cheap, share the same footprint/look as its NO counterpart and are PCB mountable. With this design I can use the same button type for both start and stop.

100k resistor has to go to +relay, not to the drain. Leo..

Or to +5V - you want the MOSFET to be fully on, not acting as a source-follower.

MarkT: Or to +5V - you want the MOSFET to be fully on, not acting as a source-follower.

Look again. When the start button is released, the relay's NO contact takes over, so the 100k resistor stays connected to 5volt. After the stop button is pressed, the 100k resistor has to be disconnected from 5volt.

Current diagram uses the mosfet in it's lineair region. Leo..

No, the relay contacts open once the stop button is pressed so it doesn't matter if the MOSFET is on or off, the relay coil is isolated at the high end.

Yes, its confusing, I had to redraw the circuit the right way up to convince myself.

Maybe a misunderstanding. I think you said "connect the 100k resistor to +5". Then the mosfet will turn "on" again the moment you release the stop button. If you connect the 100k resistor to +relay/cathode diode, the gate gets 5volt when the relay is on, but nothing after the stop button is pressed. Leo..