Reverse battery protection on a switcher charger

Greetings, I'm trying to come up with a way to protect a buck converter Lithium Poly single cell battery charger against reverse battery polarity by the battery being charged if it is connected backwards. I can't depend on polarized connectors as the user may change them. I have two problems. First, the output filter capacitor is a 100 mfd tantalum. Is 4 volts reverse polarity going to damage it? I'd hate to have to use a non-polar cap. Maybe I could do without it altogether? The second is a little tougher. There is two Schotkey diodes that become forward biased as a direct short when the battery is connected backwards. I tried to put a MOSFET switch in series with the clamp diode but the body diode of the FET becomes forward bias the same way. I guess I could use an NPN to disconnect but then that forward biases the base-collector junction when the battery is reversed which may destroy the processor that is driving it. See page

7 of the pdf file at this link:
formatting link
to see the general shematic of what I am using.

Any suggestions would be greatly appreciated.

Regards,

Art

Reply to
Art K6KFH
Loading thread data ...

formatting link

There's a circuit with a backwards P-Channel mosfet described at

formatting link

Perhaps a switch of that kind between your capacitor and the battery would work.

Mark Borgerson

Reply to
Mark Borgerson

formatting link

Hi Mark,

Thanks for the tip but I don't think that will work in this case. That idea works when the battery is supplying power to the circuit but my circuit is supplying power to the battery and I'm only working with

3-4 volts. Also, the real problem is with the Shotkey diodes being forward biased in series directly accross the battery. Nothing to limit the current as I have to be able to supply several amps to the battery when it is connected up correctly.

Any other suggestions?

Art

Reply to
Art K6KFH

formatting link

Hmmm. Seemed like it should have worked if the p-channel MOSFET was connected up so that the charging current went through the internal diode. Whenever the battery was connected in reverse, the gate would be higher than the drain (or source--can't remember which right now), turning off the mosfet and allowing no current to flow in either direction. You would have the problem of having a diode in series with your charger, and might have to have the sense input connected outside the MOSFET. If I get a chance, I'll work up a simple model in spice and see what happens.

Nothing that doesn't involve comparators and relays.

Mark Borgerson

Reply to
Mark Borgerson

formatting link

I did come up with a circuit that involves a large PMOS power FET and an NMOS switching FET. (For the crude ASCII art to look good, use a monospace font like courier)

charger-->----------[ P Chan]---[1 ohm]| | ------- | | |G | |--[1K]------| | | | | | | |----- Battery + | | ---| | | N-Chan | |-[1Meg]--| ---| | | | |----- Battery - | | GND------------------------------------

When the battery is properly connected, the NMOS fet conducts, and pulls down the gate of the PMOS fet. That connects the charger to the battery.

If the battery is reversed, the NMOS fet gate is negative and the FET does not conduct. The PMOS FET gate is pulled high and the PMOS fet does not conduct.

You should probably connect your voltage and current sensing taps after this disconnect, so that the charger can compensate for the drop across the PMOS FET. The 1-ohm resistor may be redundant---as there is a current-sense resistor in the original charging circuit. You might also consider having the microcontroller directly control the NMOS FET. That way, your algorithm might be able handle jump-starting a battery so dead it could not turn on the NMOS FET.

It seemed to work OK in simulation with generic FETs. In the real world, with chargers and batteries in the Ampere range, you may have to choose components carefully--especially to find a PMOS fet that turns on nicely at 4V or less. Then you can start worrying about how it will handle a really dead battery (near 0V), both properly and improperly connected.

Mark Borgerson

Reply to
Mark Borgerson

formatting link

Hi Mark What is the substrate diode doing when the p-channel fet is back biased? Dwight

Reply to
dwight elvey

formatting link

Good question. That diode should never be back biased if the charger is on when the battery is connected, as the charger voltage should always be higher than the battery voltage.

If the charger is off, the diodes in the charger should block any current through the diode in the p-channel FET.

Something worth checking in the design as a whole, I guess.

Mark Borgerson

Reply to
Mark Borgerson

Hmmm. Never thought of it that way. It does have the advantage that, if the p-channel MOSFET is properly chosen, it won't have a 0.7V forward diode drop. I haven't yet decided whether it would be subject to latch-up, or if it would turn off without having to shut down the charger.

Mark Borgerson

Reply to
Mark Borgerson

formatting link

Hi Guys,

Somebody sent me a link to a Siemens site that had the perfect solution to this problem. It's quite similar to what you guys suggested. A P-channel MOSFET in series with the output and a signal diode from gate to ground. A reversed battery can't turn it on. I tried it out this morning and with 1 amp through it it only dropped

15mv across the FET. The body diode does get forward biased when the battery is in backwards but the charging circuit is switched off until it sees a properly connected battery so no current flows. Works great!!!!

Thanks for all your suggestions. Sorry I couldn't thank the specific person who sent me the link to Siemens but I couldn't find the source for that suggestion. This change is actually going to reduce component cost a little bit and greatly increase reliability.

Regards,

Art

Reply to
Art K6KFH

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.