TPS22810: Adding reverse input protection

Hi MR Flashlight,

Technically, adding a diode in the ground path could protect against reverse polarity, but it would also be recommended to add a resistor in series to limit the current (~1k should do the trick). Schottkey is preferred since the voltage would be clamped to a lower value. Also, the VIH level of the device would increase due to the shift in device ground.

In general though, the reverse polarity could not be sustained for lifetime conditions, only brief moments of time. It is difficult to classify this since we haven't done any extensive testing with this configuration. 

The only integrated power switch we have with reverse polarity protection is the TPS2660, but this may be overkill for your application.

Also, can you describe your application? We do not get this request often and would like to understand where it is coming from.

Thanks,

Alek Kaknevicius

The TPS2660 would be way too large for my application.

I design flashlights. This particular product has the battery separated from the light source, and it would be possible to plug the two together backwards. It wouldn't work in this configuration, so I don't imagine it would be left like that for longer than a few minutes at a time. Size and efficiency are key, so I've been trying to identify the best way to add the protection I need within my space constraints and without losing too much power.

This solution of putting a Schottky on the ground pin adds 1 or 2 0402 parts without dropping voltage along my power path, which is why I like it. A reverse protection MOSFET would be about twice as big in terms of components added, and also would add a small drop. A reverse protection Schottky I can fit as currently designed, but it would result in a significant loss of power.

Would there be any good way of testing out this solution besides buying the part, breadboarding it out and trying to see if it works? I'm not sure if any Spice models would be complete enough to see how the part behaves outside of spec.

I wanted to provide an update.

I purchased the TPS22810 in the SOT23-6 package. I then breadboarded out the following circuit, putting the TPS22810 on a breakout board.

I applied 5, 10, and 17 V in both forward and reverse configurations. In the forward configuration, I verified that jumping the enable pin to Vin turned the output on and off. In the reverse configuration, I let the circuit sit for over an hour while watching for failures (such as current flow), then switched back to forward to verify operation. All of my tests passed.

I'm comfortable enough with this to move forward with it in my solution. Obviously, to properly verify that this works, the original silicon design would need to be consulted, or this test would need to be executed on hundreds to thousands of units, but for my purposes I am content right now that this will be sufficient.

Hi MR Flashlight,

Thank you for the update, and glad to hear that the circuit is performing to your expectations! I will keep you updated with any further investigation our team does on this configuration.

Thanks,

Alek Kaknevicius