Further Battery Fun

So i thought, 'hmm alkaline cells are like 2000mah, maybe i could use them here....'

well the first problem is the voltage. 2xAA cells give you 3v, the pi can run as low as 2.7v, so we could use that!!

So i wired it up, and it worked... for all of 15 minutes... it seems that alkaline batteries drop voltage very fast. we need to boost that voltage!!

So i raided the parts bin, and i found an adjustable buck/boost board, gave that a go

didnt even make the start!! seems the boost board wont work below 3v

so plan C, over to pololu and one of there boost boards boosty which can take an input as low a 0.5v and boost it up so we looked like we were onto a winner

So testing began, and i was expecting big things, 2000mah should technically yield 8 hours, but after just 3 hours, the pi started rebooting. seems that boosing the voltage isnt as simple as it seems!! to get 3.3v @ 200ma on the output, the input needs more current. so as the voltage drops, the current requirements raise.

vin	current (ma)
3	220
2.75	240
2.5	264
2.25	293
2	330
1.75	377
1.5	440
1.25	528
1	660
0.75	880
0.5	1320

so as we can see here, things will quickly snowball. as the voltage drops, the current raises, which will force the voltage lower, and the current higher, and so on, until the batteries just give up.

So using lithium batteries is a much better prospect, as the voltage in general is higher, and there is no need for boosting!

Discussions

K.C. Lee wrote 04/06/2017 at 14:55

The problem with AA cell is that the internal resistance is relatively high, so you'll get a lot of droop and energy lost internally. It simply isn't designed for high current loads. NiMH AA cell on the other hand would do a bit better.

e.g. My old digital camera can take 2-3 pictures with new alkaline AA and die. With AA NiMH, it can take 80 or so and some with flash.

Common LiPo battery is 3.7V to 4.2V. LiFePO4 on the other hand is 3V to 3.6V which can be used without regulator. (see the LiFe Pi project)

Are you sure? yes | no

moosepr wrote 04/06/2017 at 15:22

yeah i am using unboosted generic li-po cells normally, can get 4-5 hours from a 1000mah cell.

i didnt want to try NiMH batteries because i know they dont like being sucked totally dry. I guess 3 cells unboosted would be the ideal, but then there is no space advantage over the li-po really

Are you sure? yes | no

K.C. Lee wrote 04/06/2017 at 15:55

Regular AA has the advantage of easy to find in a store in a pinch when your batteries are running low. NiMH isn't and not usually come fully charge. Hope you LiPo cells are of the protected variety as they don't like over discharged.

Are you sure? yes | no

moosepr wrote 04/06/2017 at 16:21

seems it wont let me reply to your new comment :P

im using the TP4056 protection boards http://www.hotmcu.com/tp4056-micro-usb-5v-1a-lithium-battery-charger-with-protection-p-176.html

they seem to have the protection cut off a little low (2.4v) but the pi will start behaving weird at 2.7v so you will never really get too low

Are you sure? yes | no

K.C. Lee wrote 04/06/2017 at 17:00

There is a limit on the levels of nested replies.
In general Li battery doesn't like to be deeply discharged as the cycling takes away battery lifetime. 3V is commonly used as the lowest point. May be some sort of low battery warning or battery level indicator is needed.

Are you sure? yes | no

moosepr wrote 04/06/2017 at 17:51

ah that makes sense! I did plan on adding an attiny to so some safe shutdown/power metering kinda stuff. I had assumed the protection board stepped in sooner, but hey!

Are you sure? yes | no

Marq Watkin wrote 04/06/2017 at 13:02

Thanks for sharing your experiments with power sources, it's great to read

Are you sure? yes | no

Low power, low sucess!!

further further further battery fun

Discussions

Become a Hackaday.io Member