Is completely discharging AA Batterys worth the trouble?

[PeterFW]

Hello!
In the past i have allways worked of Li* cells with portable things, this time i want to use AA cells for various reasons. After a great deal of deliberating.

No Lithium, i want to use two AA cells in series.

Now i have to consider something, i need a voltage regulator, not a huge deal but i have trouble selecting one for a perticular reason.

My go to for this is the MCP1640, and this one works very well, i need 3.3V out with a decent efficiency at a light load and it still should be able to deliver short peaks of 300mA.
It starts up at 0.65V and that means it will run non rechargeables completely dry.

But, i am considering to add the possibility to run the thing of USB as well and that would lend itself to change the voltage regulator to a Buck-Boost.

Now... i have set my eye on one but, it has a minimum input voltage of 1.8V.
That is not a problem when i would use NiMh, those should not be discharged below that, so that is fine.

But... when i would use non-rechargeables, they have spend most of their power allready when they reach 0.9V per cell and i think it should be pferctly fine to use another voltage regulator.

But the back of my brain does not like it, even after looking at the discharge curves of AA cells, it still thinks it would be a good idea to stick with the regulator with the lower start voltage.

That is completely non reasonable, i should just use the buck-boost and stop discharging the cells at 1.8V (0.9V per cell) right?
It is not worth the extra trouble and circuitry.

Greetings,
Peter

[Ian.M]

I suspect you'd be FUBARed by it dropping out of regulation or even shutting down due to your load spikes + rising internal resistance with discharge before you could get to 0.8V per cell anyway.  Take 0.9V/cell as an end of discharge voltage as, assuming 85% regulator efficiency, to deliver 300mA peak will require something like 650mA from the battery. At 1.2V/cell it would need 490mA.  From fig 10 of http://data.energizer.com/PDFs/alkaline_appman.pdf, it appears unlikely that you can sustain your peak current once discharged past 1.2V/cell

[wraper]

At 0.9 there is almost no energy left, maybe 5%. The other problem is if you leave deeply discharged batteries, often start to leak and damage the device.

[PeterFW]

Hello and thank you for your replys!

I suspect you'd be FUBARed by it dropping out of regulation or even shutting down due to your load spikes + rising internal resistance with discharge before you could get to 0.8V per cell anyway.

Oh... i had not thought of that.
Thanks for pointing that out!

At 0.9 there is almost no energy left, maybe 5%. The other problem is if you leave deeply discharged batteries, often start to leak and damage the device.

Yea, you are both right!
Bad idea, out goes theMCP1640, it will be replaced by a TPS63001.

I have not crunched all the numbers but mostly the circuit will be drawing around 1mA and 10 to 20mA when running.
The spikes are either running the buzzer, writing to an SD card or sending data over wireless for <1s.
All that goes out the window when the backlight comes on though, but that is just for emergencys.

Greetings,
Peter