I recently had a situation where a p-mos high side switch has blown. I suspect that its due to over voltage from the supply.
The circuit is used to control a small petrol generator. Power comes from its magneto coil thru series diode into a 12v battery.
I have recorded 17v from the coil but I suspect that it reaches more than that. Together with the fact that this is a 220 vac generator, I
really need to implement some kind of over voltage protection.

Other than a reverse polarity MOS on the input and flyback diode for the starter motor, there is no other protection.

Its time to add some Over Voltage protection to the circuit.

So as the circuit uses several High side switches (gates pulled up to supply) anything over Vgs is bad (20v).

I initially thought to regulate the magneto coil. But this is kinda weird because after a few minutes it regulates itself and normal
lead acid battery charging voltages are constant (max of about 15 volts). Also I need to deal with any transients that a generator
like this will probably attract.

The circuit has worked pretty well, and even the ignition voltages don't seem to bother it, and i get accurate ADC readings every time.
Even the onboard Modem seems to work without problems.

I should mention that there are 4 HS switches, a motor driver and SMPS connected to supply - so its beneficial to get this sorted on the
supply rather than individual ic's/circuits.

max input voltage = 20v due to Vgs of PMOS.
Idle operating current (when the circuit is in standby) is approx 1.6mA @ 12.6v = 0.02W
Peak operating current (modem/solenoid in use) 2.5A @ 4.1v + 5A @12.6v = 65W
normal operating current (modem only) approx 100mA @ 12.6v = 1.26W

I am looking for advise to implement an over voltage solution where its capable of dealing with sustained over voltage (say upto 30v for few minutes) and deal with unknown transients such as static from the generator. There are many solutions out there and TBH a bit of a minefield when trying to decide what is best. Zener regulator, crowbar, tvs, avalanche etc. Can anyone point out a solution?

Thanks in advance.
A

 

The gate voltage being over 20V is not a problem if the source voltage is the same which I would expect for a high side switch when it is off. The problem arises when turning on the MOSFET if the gate is pulled to 0V while the source is still over 20V. You could solve that problem with this kind of circuit where the zener limits the maximum voltage between gate and source.

 

The gate voltage being over 20V is not a problem if the source voltage is the same which I would expect for a high side switch when it is off. The problem arises when turning on the MOSFET if the gate is pulled to 0V while the source is still over 20V. You could solve that problem with this kind of circuit where the zener limits the maximum voltage between gate and source.

That's exactly the problem. wish i had the foresight to implement this beforehand. As a work around could I zener the whole
supply?
To understand your solution the zener drops the excess voltage to the gate? Keeping it at net 15V? and if R2 is the gate pull up, what is R1 for?
is this solution good for HV transients, or will the MOFSET Vds absorb them somehow?

 

what is R1 for?

R1 is there to limit the current through the NPN collector. So supply minus 15V is dropped across R1.

 

I see. Thank you Bert, so could i do the same thing for the supply but with the zener diode on the supply to base and resistor base to ground? diverting the over voltage thru the trans