I am making a circuit which may draw up to 3-4 Amps, so I am avoiding using a Schottky diode as fool-proofing it from incorrect polarity on the battery (12V, automotive style) input.
I have used P channel MOSFETs, one as an input polarity protection, and another after a fuse for ON/OFF by way of using a PCB mounted slide switch, which is a Single Pole Double Throw switch. This allows me to either pull the gate high of the second MOSFET shown below as Q11, into an "OFF" mode, or pull the gate to ground into an "ON" mode.
I have used a PFET instead of the physical switch in series due to the size of switch needed to have the proper current ratings. This way seems like a better way to handle current for a PCB mounted ON/OFF switch with more than a few hundred milliamps going through it.
Should I bother with the 10K resistors, or will it be okay without them?
I am thinking of putting an ~18V zener clamp/TVS diode to protect it from inductive spikes from connecting long cables to the input, and perhaps some 10uF ceramic capacitors very close to the drain of Q11 to also help reduce spikes on connection.
Here is my possible improvements. Have you guys got any better ways to do this, which is cheaper/will get the job done and be robust enough for 12-14V input, 5A fuse in-line as shown, and loads being in the range of 2A-3A maximum (usually will be more like 1A).
In the second version I also moved the fuse up to be the first thing from the input connector, because it's possible for the first PMOS to die and short in strange ways, so the fuse will blow if ANYTHING weird happens. I think this is better.