I have recently been designing an arduino controlled power board to run all my non engine management electics in a kit car i am building see this post :
I am now planning the main protection board this needs to provide protection from various car elctric nasties :
voltage spikes up to 120v positive or negtive 10-100 ms
Reverse polarity
24v boost during starting (although could isolate the board)
The arduino has it's own off the shelf USB power suppley unit working form 12v which i thisnk will add some of it's own protection i have alos designed an optocoupler sheild for all the arduino i/o pins
So the main power supply i plan to protect using this circuit:
The main features being
double pole isolation switch
20 Amp fuse/mcb
TVS bi directional diode (smcj40ca) that will short at voltages over 30v but allow 24v jump starting and manange the spikes
p channel mosfet acting as a reverse polarity valve with v low resistance 0.026 ohm to allow 20 amps although will still need large heatsink
zenner and resistor to clamp Vgs to -10v
i took the reverse polarity idea from this video which explained it well to me.:
Please let me know if anyone thinks my power boards or arduino are still gonna go bang !
I am a long way from getting an engine running so have plenty of time to modify bits
thanks for all your help as usual
in particular i wasn't sure how to arrive at a value of R2 i assume i need to draw enough current through the zenner to reverse polarise it but i wasn't sure what current to look at for this on the data sheet there seemed to be 2 currents quoted IZT which was 25mA and IZK which was 0.5 mA ???
You're missing the diode used to get turn-on volts to the circuit side of the FET
Your arduino on-board regulator circuit won't like you trying to run it from 24 volts (even for a short duration) so you might want to look at providing a 12 volt regulated supply dedicated to arduino power feed only (to minimise heat loading)
Thanks for your help, the implication on the video was that the MOSFET had a built in parasitic body diode which does appear on the schematic on the datasheet for the MOSFET quoted in the circuit although wasn't on the schematic in circuitlab although they didn't have my exact MOSFET, is this correct or do i need to add an additional diode in parralel and if so will this carry significant current that i have to worry about it's power dissipation or will the MOSFET once on take the lion share of the current ?
Also i should have explained better that i have a dedicated 12v to USB power unit i think similar to what you get in the cigar lighter chargers for cars but on a small PCB.
can confirm circuit works well, the mosfet switches on without an additional parrallel diode so must exist within mosfet.
Also played with value of R2 and seems happy to drop 10v accross the zenner with 10K or 1mA.
However I did notice that if you disconnect the Gate completely the current still flows through the MOSFET !!! and it gets v v v hot presumably this is because it is then all flowing through this body parasitic diode with a more sig voltage drop and so lots more wattage i was driving a headlamp of 4 amps and it was far too hot to touch after a couple of seconds so switched off. With the gate negatively biased the mosfet passes the 4 amps hardly getting warm as the RDS on is only 0.026 ohm for this MOSFET so 0.5 W or so. The problem with this is that if the zenner short circuits then the MOSFET will switch off but current will still throw and i presume the MOSFET will fail, not sure if there is anything i acn do about this. Can someone please confirm if my analysis is correct about why current is still flowing when the gate is disconnected ?
