I have a device which runs on two AA Duracell batteries. The batteries (brand new) together produce 3.2 volts (1.62 volts individually), which is estimated to run for at least 2 years. The device is always running and consumes 300 micro amperes, which isn't a lot.

Could someone tell me how long can the batteries will last on 3.2 volts (1.62 volts individually) before dropping to 3.0 volts (1.5 volts) in years/months/weeks/days/hours/minutes? So basically, how long until the batteries voltage begins to drop?

  • 2
    \$\begingroup\$ You really need to find the datasheet for your battery. professional.duracell.com/en/product-datasheets . AA batteries come in different chemistries and rating all of which will give you different lifetimes. You need to see the curve for your battery. \$\endgroup\$ – Will Oct 13 '14 at 22:49
  • 2
    \$\begingroup\$ eevblog.com/2011/01/23/eevblog-140-battery-capacity-tutorial this has a good graph and explanation of battery capacity \$\endgroup\$ – Will Oct 13 '14 at 22:53
  • \$\begingroup\$ @Will, I think he wont be able to find curves for such low currents. \$\endgroup\$ – Kamil Oct 14 '14 at 0:07
  • \$\begingroup\$ @Kamil it's true that there is no curve at 300uA, but it does give you an idea of 'what is the voltage when half of the energy is used?' or 'what proportion of the energy is left when the voltage is down to 1.5v?'. The curve drops quickly to start with for alkaline batteries, so the answer is most of the energy is left... \$\endgroup\$ – Will Oct 14 '14 at 0:33
  • \$\begingroup\$ @Will I know. Maybe my comment was unnecessary. I added my answer and... I have same problem - estimation from higher current curves is just necessary. \$\endgroup\$ – Kamil Oct 14 '14 at 1:07

I don't have good news for you.

Below is the discharge rate for Durcell batteries done by some company. Look at the DC label; this is a Duracell coppertop battery. The complete test can be found here: link.

Enter image description here

There was no µA test (it would take too long), but I guess the voltage will drop below 1.5 V after about 1-2% capacity discharge. There will also be battery self-discharge (very slow). My guess estimated time for 0.3 mA (300 µA) before dropping to 1.5 V will be somewhere between 50-150 hours. You can test this; it's not that long.

I would suggest to use different kind of batteries. Alkaline and a 3.0 V requirement is just the wrong battery for this application.

You can also consider different battery chemistry. If you use some lithium AA battery, for example Energizer L91 - they have much more energy "available" before voltage drop below 1.5 V, but be careful - they have also higher initial voltage (about 1.7 V) and they are expensive (in my country they cost 2-4x more than alkaline Energizer or Duracell).

Enter image description here

Image from Energizer Ultimate Lithium L91 datasheet: link

However if you want to pull 0.3 mA (300 µA) from battery for two years (over 17000 hours) - you need more than 5 Ah (5000 mAh) before the voltage drops below 1.5 V. This is probably too much for any AA battery available on the market.

There are also nickel-zinc AA batteries, they have nominal voltage 1.65 V, but they have less capacity (50% less than Energizer L91).

Another idea might be three batteries with a low-dropout linear regulator. Three AA batteries in series will provide more than 3.0 V until they are completely empty, but a linear regulator may be necessary for some devices - three new alkalines may have 4.95 V initially.


The batteries (brand new) together is producing 3.2 volts (1.62 volts individually), which is estimated to run for at least 2 years. The device is always running and consumes 300 micro amp which isn't a lot.

Your math is off: A high quality AA cell has about 3000mAh, which gives you about 10,000 hours run time, or a bit more than one year.

But the discharge voltage will drop quickly lower than 3,0V - end voltage is about 0.9V/cell. I would expect the drop from 3.2 to 3.0V to occur within one or at most two weeks.

  • \$\begingroup\$ "A high quality AA cell has about 3000mAh" - Do you have a reference for that? What chemistry? I don't think I've ever seen an alkaline AA cell spec above 2200mAh. \$\endgroup\$ – marcelm Jan 11 '19 at 12:54
  • 1
    \$\begingroup\$ Number taken from Energizer E91 at 25mA discharge current. One of the first that pop up for me when searching for "AA datasheet". \$\endgroup\$ – Turbo J Jan 12 '19 at 13:59
  • \$\begingroup\$ Fair enough! I'm a bit surprised; perhaps the figures I'm familiar with were at a larger drain current... \$\endgroup\$ – marcelm Jan 15 '19 at 16:00

Using a reulator allow power or current to be stabilised as the battery voltage drops.
Using any regulator you need to define the load that the battery sees unless your load is a resistor. To be able to calculate battery life you must follow the manufacturer's datasheet

Example-only graphs are from here - Duracell AA Alkaline-Manganese Dioxide Battery.

enter image description here
enter image description here

If we use a linear regulator we are following the constant current.
If we use switching regulator we are following constant power.

To be possible to follow these load-lines we must use a test load in the output of the regulator and read the voltage during test .
A single measurement give you different results each time so needed put at least 20 measurements in a buffer and create a moving average (shift so make last out new measurement in and divide by 20) the time between measurement (if your design is for years) can be between 30 mints to one hour. In this way measurement get good stable results

If we are using only DC current in the circuit Average and RMS are the same.
If you are measuring voltage with Microcontroller you must use moving average to get good reading otherwise you have noise.

For the chart here is the polynomial that I have developed of Voltage wrt current and time.

Voltage(t) =- (60.10778E-2 • CRNT• t)⁵ + (84.63365E-2 • CRNT • t)⁴ - (10.91466E-1 • CRNT• t)³ + (12.58045E-1 • CRNT• t)² - (10.32172E-1 • CRNT• t) + 1.600119 • n

CRNT-current in mA
t-time in hours
n-number of batteries in series.

Generally an electrical circuit contains different currents, if the design relies on AVG So can be a big mistake between reality and measurement.

  • 2
    \$\begingroup\$ Hi, As required by this site rule, when you include something in an answer (e.g. photo, image or text) which isn't your own original work, you need to properly reference (cite) it. Those images have come from somewhere else, so in order to comply with that rule, can you edit your answer and add a link back to the original source web page / document, please? Thanks :-) \$\endgroup\$ – SamGibson Jun 8 '20 at 10:35
  • \$\begingroup\$ @Koren Make sure you use the edit button instead of creating new posts. Thanks \$\endgroup\$ – Voltage Spike Jun 8 '20 at 15:27
  • 1
    \$\begingroup\$ @Koren Reuben - Your polynomial looks interesting and useful. An explanation of its basis and how you derived it would be valuable. || I have provided links to the source of the diagrams. Please do so with any future answers wit links to material from elsewhere. \$\endgroup\$ – Russell McMahon Oct 4 '20 at 18:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.