I am designing a simple circuit that pulses a coil once every 60 seconds to actuate a clock mechanism.
The coil is about 8 ohms, the current is about 375 mA for about 50 mS through the coil.

The circuit runs from two "AA" batteries which will undoubtedly be installed backward at some point, which will destroy the board as designed.
(the body diode of the NFET and freewheel diode provide a dead short under these conditions.)

I want to implement reverse polarity protection while meeting a few criteria:

(a) As cheap as possible

(b) No impact on battery life (very little extra leakage)

(c) Very little added voltage drop - so the system will continue to work with weak batteries.

(d) Must work after polarity reverse event- no fuse to replace, etc.

Here are some of the ideas I have already considered:

Series Schottky diode- eats up maybe 3-400 mV of my battery voltage, this would shorten my system life.
Schottky diode backward across the supply - would need to be really beefy to handle the full short circuit current of the battery. High leakage?
PFET on the input - Expensive, must have low Rds-on value so losses are low.


The PFET option is the most logical, but it's expensive.

Trying to design anything right now feels like shopping at Home Depot after the Zombie Apocalypse, Digikey has been picked clean like a thanksgiving turkey carcass.

 

https://www.ti.com/lit/an/slva139/slva139.pdf

 

I am designing a simple circuit that pulses a coil once every 60 seconds to actuate a clock mechanism.
The coil is about 8 ohms, the current is about 375 mA for about 50 mS through the coil.

The circuit runs from two "AA" batteries which will undoubtedly be installed backward at some point, which will destroy the board as designed.
(the body diode of the NFET and freewheel diode provide a dead short under these conditions.)

I want to implement reverse polarity protection while meeting a few criteria:

(a) As cheap as possible

(b) No impact on battery life (very little extra leakage)

(c) Very little added voltage drop - so the system will continue to work with weak batteries.

(d) Must work after polarity reverse event- no fuse to replace, etc.

Here are some of the ideas I have already considered:

Series Schottky diode- eats up maybe 3-400 mV of my battery voltage, this would shorten my system life.
Schottky diode backward across the supply - would need to be really beefy to handle the full short circuit current of the battery. High leakage?
PFET on the input - Expensive, must have low Rds-on value so losses are low.


The PFET option is the most logical, but it's expensive.

Trying to design anything right now feels like shopping at Home Depot after the Zombie Apocalypse, Digikey has been picked clean like a thanksgiving turkey carcass.

About 300mv drop with a PNP transistor.

 

A battery holder like this one is designed to make reverse insertion harmless. This particular one might not be suitable because of cost or being through hole, or whatever but mechanical protection has no voltage drop at all.

​

 

A 3 A Schottky diode should come in closer to 200 mV.

ak

 

How about the ROHM RQ5E035ATTCL with an RDS OF 50mOhm? It's 18 cents at Tayda Electronics. This would provide a voltage drop of 19mV.

Tayda is an awesome vendor located in Thialand, with a US warehouse. I have been pleased with all the components I have purchased there.

 

The PFET option is the most logical, but it's expensive.

What do you call expensive?

You should be able to get one to meet your requirements (<100mΩ Rds(on)) for well less than a buck US.

Since you circuit is floating, you can use an N-MOSFET to block the negative line side to the battery for reverse protection, as it's generally a little cheaper than a P-MOSFET for the same characteristics.

 

How about the ROHM RQ5E035ATTCL

As you can see from the data sheet excerpt below, that MOSFET has a Vgs(th) that's too high for operation from 3V.

This needs a MOSFET with a max Vgs(th) of <2V.

 

What do you call expensive?

You should be able to get one to meet your requirements (<100mΩ Rds(on)) for well less than a buck US.

Since you circuit is floating, you can use an N-MOSFET to block the negative line side to the battery for reverse protection, as it's generally a little cheaper than a P-MOSFET for the same characteristics.

It had not occurred to me to use an N-FET! that's a nice idea- I could use the same part as the coil driver to minimize the BOM count.
Here is the design with the N-FET this is what you meant? Correct?

The other idea that is really nice is the mechanical solution! I will consider this too.

 

It had not occurred to me to use an N-FET! that's a nice idea- I could use the same part as the coil driver to minimize the BOM count.
Here is the design with the N-FET this is what you meant? Correct?

The other idea that is really nice is the mechanical solution! I will consider this too.

Look at the RDS(ON) at the operating voltage for that device.
https://www.digikey.com/en/products/detail/diodes-incorporated/DMN2005K-7/1778818
https://www.diodes.com/assets/Datasheets/ds30734.pdf


Look at this device.
https://www.digikey.com/en/products/detail/micro-commercial-co/SI2312-TP/6616108
https://www.mccsemi.com/pdf/Products/SI2312(SOT-23).pdf

 

Look at the RDS(ON) at the operating voltage for that device.
https://www.digikey.com/en/products/detail/diodes-incorporated/DMN2005K-7/1778818
https://www.diodes.com/assets/Datasheets/ds30734.pdf


Look at this device.
https://www.digikey.com/en/products/detail/micro-commercial-co/SI2312-TP/6616108
https://www.mccsemi.com/pdf/Products/SI2312(SOT-23).pdf

Wow! thank you!

This part is vastly superior and CHEAPER! The only downside is its MASSIVE input capacitance, but that's irrelevant for this design.

 

Here is the design with the N-FET this is what you meant?

Yes, that's it.

 

The other idea that is really nice is the mechanical solution

The only thing that won't protect against is a wiring error between the battery holder and the circuit.