What is the typical practice for protecting VBAT from a reversed coin-cell insertion?

[DHase.1]

It is possible for a CR2032 coin-cell to be inserted backwards in the holder. Therefore the VBAT (on a STM32F446) input needs some protection for reversed polarity. What is the usual method of protection?

The datasheet shows a -5 ma as the current limit. It isn't clear if there is a difference between when the other power pins are powered or not powered. The simplest way to limit the current would be a series resistor. A value greater than 680 ohms would limit the current and the voltage drop would not be great when the RTC and backup registers are powered. Above about 100K the voltage drop comes into consideration.

Other approaches would be a Schottky diode and suffer the voltage drop. A PFET could be used and avoid the diode drop, but the gate leakage of the PFET might be a factor in the life of the coin-cell and there isn't a need for high current.

A series resistor looks the simplest. I was wondering what approach others take.

[waclawek.jan]

One - cheapest and simplest - of the options is just ignore the problem, educate the users and suffer the occasional mistake.

The drop on a Schottky at 1uA is less than the datasheet value which is usually given at much higher currents.

Some small coin cells have enough internal resistance themselves.

JW

[gbm]

This is a non-existent problem. If you insert the cell incorrectly into a metal holder, it will be shorted.

[AScha.3]

just use a cheap standard batt. holder, inserting wrong direction - nothing happen, no contact on "-" .

problem away...

[waclawek.jan]

In one design we use a vertical holder because of spatial constraints, and the cell can very well be inserted reversed.

JW