Hi - I've had to deal with a similar requirement.  I used the linear=20
LTC1440 micropower comparator + reference to switch a PFET; you can=20
choose the voltage to switch at as well as how much hysteresis you want,=20
which is helpful b/c the battery voltage will rise of course when the=20
power is cut.

That design used two li-ion cells, so I set it to be about 6V, not sure=20
the LTC1440 works down to 3V, probably but do check that.

The part is pretty cheap but not dirt cheap.

Another thought - this wasn't possible due the other design=20
requirements, but a trick I was going to try:

If you have a microcontroller and want a pushbutton to turn on the=20
board, you could have the LDO shutdown pin (or pfet) be wired to the=20
pushbutton and a resistor that pulls down/up as required to have the=20
LDO/pfet be in the OFF state as well as an output pin on the=20
microcontroller (thru a low-value resistor).  So when the user presses=20
the button the LDO/pfet is briefly on.  This powers up the=20
microcontroller, and the first thing it does is read the battery=20
voltage.  If OK, it sets the output pin so that the LDO/pfet is held ON.=20
  If not OK, it doesn't and blinks a "battery low" indicator and goes to=20
sleep.  Then when the user releases the power-on button the LDO/pfet=20
gets pulled back to the off state.

If you use a double-pole pushbutton, you can have another=20
microcontroller input pin monitor the other pole, and detect when the=20
user has pressed the button again, and then shutdown.  So now your=20
pushbutton works as press to turn on, press to turn off, and you've got=20
low-battery detection "for free" with a LDO that has shutdown.  Of=20
course if your firmware crashes, the board is "stuck on" and eventually=20
kills the batteries.  have to use the watchdog timer for sure.

All this said, I didn't build said circuit.  If you do let me know=20
how/if it works.  :)  A bit of an aside all this but I thought I'd=20
procrastinate a moment on what I should be doing (writing a mini=20
application in Windows, yawn) and write it up.

Cheers,
J




Mark Rages wrote:
> I'm building a circuit powered by a small Li-Ion battery. The battery
> will be connected to a 3.3V LDO, which then powers the entire circuit.
>
> Li-Ion batteries are damaged if they are allowed to drain below about
> 3V. There are chips to do this LVC cutoff.
>
> But I figure the LDO will already have the pass transistor and voltage
> reference, so it seems wasteful to add another part. This seems to be
> the kind of thing that could be easily integrated.
>
> Does anyone make an LDO with low-voltage cutoff? Parametric searches
> and Google seem to be failing me.
>
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.