MOSFET logic for Enabling Battery Backup

Thread Starter


Joined Mar 23, 2019

Ive got a bit of a headscratcher here.

In our system, there are 3 primary voltage sources (at any time, 0, 1, 2, or 3 of these sources can be plugged in). Their voltages are 5V, 9V, and 12V. There is also a battery backup that should kick in when no primary sources are plugged in. The battery voltage will be around 9V.

For the primary sources, I will just use ORing diodes, so the highest voltage source provides the current to the system. I should also mention that standby current will be <50uA, and full load can be as high as 2.5A-3A. Ive looked at all kinds of off the shelf components for battery high side switching, but they usually don’t work either due to the high battery voltage (9V), they cant support the max current I need, their disable current is too high (will constantly drain the battery when not in use), or their quiescent current is too high (when enabled, the system current will be too high). When the system is running on battery, I want this to be as low power as possible (again, sub 50uA), and when the system is not running off of the battery, I want as little current to be drawn from the batteries as possible.

Ive looked at making my own driver here, I can add additional ORing diodes from each of the primary source to make a bENABLE line to control the battery switch, but I feel like im spinning my wheels here.

Here is one design:
MOSFET Driver.jpg

The 3 primary sources are on the right side, and the 9V on the left is the battery. This is a high side switch. The NMOS driving the PMOS gate is essentially the switch enable, and the additional NMOS is used to essentially invert the EN signal such that if no primary sources are being used. I don’t like this solution, its too dependant on the resistor values, though it should theoretically draw 1-2uA when its disabled, and ~10uA when its enabled.

Then today, I thought this is much more easily done with a low side switch (one FET to battery negative and ground, and another FET to invert the signal like above), but it would seem the body diode is going to kill that idea, and I cant choke off the current.

LS_MOSFEt _Driver.jpg
Has anyone experienced this issue before? The obvious design challenge here is that Im trying to control a voltage on and off relative to voltages that can be higher or lower than the battery voltage.

Any help that can be offered would be greatly appreciated.

Thanks AAC!