How to determine remaining hours/days for a AA battery source?

Hi guys,

I am building a low battery indicator which lights up a red led when the voltage remaining is 2 days before the battery dies.

The device draws about 1ma all along and 50ma for about 10 seconds every 6 hours. So how do I check when the battery would have enough voltage to last for 2 more days?

i have attached the battery current performance 

Picture of How to determine remaining hours/days for a AA battery source?
sort by: active | newest | oldest
Orngrimm4 years ago
I was in charge for such a thing years back...

we had a steady power-demand (Wattage) from the battery and needed to calculate the remaining time left.
Our setup was:
The converter was designed to output a constant wattage.
That for defining the setup...

Now how we did that (it is no secret).
we used always the same type of battery (Varta PowerOne AA). Then we took 10 batteries from different batches and ran the load.
we took a voltage-measurement every few seconds.
That gave us a curve liek you posted.
We then added all the curves together and generated a channel of where the voltage can be for a certain % of the battery (simply put. we also used statistical methods on the data to get a high propability-level of a single % off every voltage-level with the given load).
Now it was simple:
In the product, we simply ran the load and measured the voltage. Then we ran it thru the calculations and we had the % of the battery. Now since we knew (with a certain certainity) the runtime of a full battery, we were able to determine the leftover runtime in the battery.

Now some hints on pits and downfalls you may encounter:
- be aware that a hald-empty battery which is left unconnected, will "gain" voltage. If you only measure the voltage now, it appears to be full. But thats not the case since as soon as you start discharging the battery, the voltage will drop massively.
- Internal resistance (DeltaVoltage/DeltaCurrent=InternalResistance) is a good indicator to see if the battery is only "refreshed" (see above) or really full. the higher Ri the less full the battery is in reality.
- If you want to go the whole 9 yards, you can get very precise measurements with methods like described here: http://powerelectronics.com/site-files/powerelectronics.com/files/archive/powerelectronics.com/mag/50122.pdf
- also for very precise measurements, you can use the AC conductance method but this will require a lot of external circuitry and sophisticated measurements... I dont recomment it here for simplisticy's sake. But if you want to dig in and for information, check @ http://batteryuniversity.com/learn/article/how_to_measure_internal_resistance The page itself (http://batteryuniversity.com/) has some very good ideas and explanations about batteries...
kurtselva (author)  Orngrimm4 years ago
Nice explanation.
Instead of generating the curve myself, Can I use the curve in the photo I posted to gauge the % of battery ?
If you happen to have either
- a constant power-consumption of either 100mW or 250mW
- a constant resistance as load of either 24, 43, 3.3 or 3.9 Ohms
- a constant current-consumption of either 500mA, 250mA, 100mA or 1A
and if you use them at exactly 21°C

THEN and only then can you take one of the graphs and translate the graph to a table and look the voltage up against the total duration.
Every other situation is not covered by the graphs.
Orngrimm4 years ago
Oh... Also a good comparison of some methods:
The guys who like answering questions tend to hangout in the Q+A section. You'd get more answers faster there.
caitlinsdad4 years ago
I would think you need to determine the values for a fully charged battery and the minimal acceptable voltage/power needed for your device to operate. From there apply the rate of discharge to see what to measure for 2 days worth of energy.