Tips on Reverse Polarity protection?

[Puffie40]

Are there any suggestions for input power protection for a circuit? My objective here is to make the design robust for mistakes during setup.

Right now on my design, I have the input power running through two diodes before going to the on-board power converter for reverse polarity protection, and a MOV bridging between the two lines to protect from surges.

Would this be overkill?

Somewhat related, I have a signal input to a microcontroller watching for a logic zero for an external control. How might one protect this line from external voltages - an optocoupler?

[NiHaoMike]

Just one properly rated diode should do for reverse polarity protection.

[waymond91]

Without knowing exactly what you're working on, it sounds like an optocoupler would work fine. Optocouplers are definitely the industrial standard for isolation - especially for big industrial controllers with +20V IOs.

I followed this quick guide the other day for reverse current protection: https://www.ti.com/lit/an/slva139/slva139.pdf

[Puffie40]

Without knowing exactly what you're working on, it sounds like an optocoupler would work fine. Optocouplers are definitely the industrial standard for isolation - especially for big industrial controllers with +20V IOs.

I followed this quick guide the other day for reverse current protection: https://www.ti.com/lit/an/slva139/slva139.pdf

It's a work project, so I don't want to explicitly say what it's for, but I can say it is a fairly simple uC-controlled motor driver circuit. Since our product is pretty modular, we custom build the wiring harness depending on the sensors we want, and I just want the inputs to be forgiving if we make mistakes in the wiring.

I guess I'll look into an simple optocoupler for this one input.

[Phil_G]

Just one properly rated diode should do for reverse polarity protection.

A mosfet makes a better diode than a diode    almost no drop & hence little power dissipation
For example https://www.ti.com/lit/an/slvae57b/slvae57b.pdf section 4
Can be much simplified at typical battery voltages

[jonpaul]

Hello: Use Schottky diodes in LV bus to reduce drop.

Extra protection: Add Schottky  or TAZ  across bus

MOVs are useless in this case.

Better info if you provide the schema, and bus specs and load info,

Kind Regards,

Jon

[jpanhalt]

The TI link by Phil_G shows a P-mosfet, which is simple.  But TI includes added voltage protection with the zener diode.  If all you want is reverse polarity, it can be simplified to just the mosfet (Capture 1).  Note the mosfet drain and source are "backward" from when it is used as a switch.    That is necessary because in the normal attachment, the body diode will conduct with reversed polarity.

There is also an N-mosfet version (Capture).  I use the P-mosfet version.

[Mechatrommer]

the "blow" diode and PTC resettable fuse is the nearest i can get to an undestructible and near zero power loss during normal operation. i did try reversed mosfet in one of my project and i blown it on the 1st prototype board for an unknown reason. the problem if current spec start to go above 10A, its hard to find PTC fuse for it... ymmv.

[Phil_G]

i did try reversed mosfet in one of my project and i blown it on the 1st prototype board for an unknown reason.

I dont know what you did wrong, but particularly in heavy-current applications, mosfet reversed-polarity protection is pretty much the accepted way of doing it.

[Zero999]

the "blow" diode and PTC resettable fuse is the nearest i can get to an undestructible and near zero power loss during normal operation. i did try reversed mosfet in one of my project and i blown it on the 1st prototype board for an unknown reason. the problem if current spec start to go above 10A, its hard to find PTC fuse for it... ymmv.

Are you sure the supply voltage didn't exceed the maximum gate voltage rating, which is normally much lower than the drain-source rating?

[Mechatrommer]

Are you sure the supply voltage didn't exceed the maximum gate voltage rating, which is normally much lower than the drain-source rating?

i use AO4423 in the design. Vgs ±25V and i was using 12-15V PSU.. maybe overvoltage due to induction spike with the lack of tank/decoupling capactor, i havent probe deep enough. but my point is on the "foolproofness", you overvoltage the Vgs during prototyping for 1ns and it will be toasted, you overvoltage a PTC or a diode for a lot longer period, nothing happen, at worst the diode will breakdown and overcurrent the PTC and your circuit will be quickly cutoff, few seconds later, it cool down again and you have a second chance.