Hi everyone,

I've read several threads but couldn't find the answer I'm looking for.
I'm trying to protect my circuit the same way the attached schema works (using PMOS)

I've read this one:
http://www.electro-tech-online.com/...ion-of-input-power-with-low-voltage-drop.664/
and this one

from other threads

However I'm having a specific question regarding the voltage of my own circuit.
I want to support 10-26v as VIN (in fact 12v to 24v but with some margin).

I plan to use either UA78M05CDCYR or LM2937IMP-5.0/NOPB to go from 12v-24v to 5v.
My problem is that Vgs using a P-channel in this case can theoretically be at some point 0 - 26v (-26v) or 5v - 26v (-21v) and most of the Pchannel I find in the package I want are max VGS -20v.

1) Do you believe the case Vgs (-26v is possible?)
2) In the case I have 25v supply, we will be @ Vgs -20v, do you think it's a safe practice to stay close to max limit?

What do you recommend me to protect from inverse polarity?

Thanks in advance for your help!

 

I believe with a Zener diode in parallel of Vgs and a resistor I'll be fine. Just need to check current is alright but by simulating with LTspice it sounds good...

I think I answered the question by myself...

Please let me know if I'm doing something wrong

 

Your P-FET is the wrong way round in the attached png file.

 

If you can stand a diode V drop a single diode would provide reverse V protection, or with 2 D drops, a bridge rectifier does not care what the input polarity is.

 

Your P-FET is the wrong way round in the attached png file.

No, it's correct. The P-MOSFET drain is connected to the input voltage so that it can block the voltage when the drain goes negative from an incorrect connection.
A MOSFET conducts equally well in both directions so will turn on when either the drain or source is positive with respect to the gate.

 

No, it's correct. The P-MOSFET drain is connected to the input voltage so that it can block the voltage when the drain goes negative from an incorrect connection.

Oops. My bad. However, 5V on the gate won't allow on/off control of the MOSFET if the supply is 24V.

 

Oops. My bad. However, 5V on the gate won't allow on/off control of the MOSFET if the supply is 24V.

5V appears to be just a bias voltage to keep Vgs from exceeding its maximum of 20V.
24V will generate a Vgs of -19V which will turn it on, as desired.
0V will generate a Vgs of +5V which will turn it off, again as desired.