Battery circuit why these two mosfets

Reverse blocking ? This example is NMOS, but principles apply -


But your circuit not back to back, error ?

This is a TI reference design.

newbe_always said:

This is a TI reference design.

Click to expand...

And datasheets and reference designs are commonly full of mistakes

completely agree

Not sure what this circuit is doing....note spice model does call out the body diode in it

So the orientation of MOSFET is wrong in the TI reference design

Unfortunately, this wouldn’t be the first time this has happened.

newbe_always said:

View attachment 149761
So the orientation of MOSFET is wrong in the TI reference design

Click to expand...

I dont think we know enough about the design intent, but do a sim, see what you
get for currents vs V applied to connector....

May I know any relation with this.

newbe_always said:

May I know any relation with this.

Click to expand...

Back in the day, before inexpensive EEPROM memory was widely available, the configuration settings were stored in static RAM. And they were kept powered during the main power-off periods via non-rechargeable lithium cells.

As the article you linked mentioned, UL were very paranoid about a schottky diode failure causing current flowing back into the cell, which could potentially cause a fire. Thus a redundant diode was employed. This caused a double diode drop on an already low (compared to 5V) battery voltage. We would have loved to use a P-channel Mosfet instead, but in those days there were no suitable devices for this purpose.

Fortunately static RAM consumption at idle was below minuscule, and the actual schottky forward drop each was a tad above 100 mV, and the circuit would be backed up by a significant amount of time.

However, for the circuit that you show, the coin cell appears to be the primary supply, and therefore two Mosfets in series doesn’t make sense.

newbe_always said:

So the orientation of MOSFET is wrong in the TI reference design

Click to expand...

That's only if the MOSFETs are connected to some type of controller.
It makes no sense when their gates are both connected to ground since, if they are back-to-back, they won't conduct in either direction.