Battery life calculator

[DenisJ]

Hi there all,
I just find this site that offer a battery life calculator.

My node sensor need 19mAh when work and about 9uAh in sleep mode.
I'll use 2 X AA (1,5V + 1,5V) battery and will send data every minute.
After I insert all the numbers it give me only 60 days.

So I want to ask, it's normal only 60 days or I'm wrong something on the inserted numbers please ?

Thanks all.
Denis.

Anyone can tell me if

[mfalkvidd]

@DenisJ being awake 2 seconds every minute will drain the battery pretty quickly (as shown by the calculator).

If you want longer battery life, consider sending much less frequently, and see if you can get the awake time down to much less time. With wakeups every few hours, and awake times in less than 1/10 second you can get multiple years battery life.

[DenisJ]

Thanks a lot for the answer,
I think is not just 2 secs... I must do some variable and misure the time.

Also the temp/hum will be send only if delta difference is more than 0.5 Celsius.

So all the number I inserted is right...
I was hopping in some mistake

Thanks a lot for the answer.
Denis

[Gilles BILLARD]

May I suggest you think about self discharging parameter in case of a reloadable battery and a long life time ?

[DenisJ]

No... I must use 2 X AAA battery because of all project.
Thanks for the suggestion

Denis

[BearWithBeard]

@DenisJ Well, the result of the calculator is right, but are your inputs, too? They seem rather pessimistic for a temp/hum sensor node. Did you measure the wake time of 2s?

In my experience, sending a message takes 80ms on average. Updating and reading a sensor may take 20 - 250ms depeding on the sensor and bus speed. A full wake cycle will likely be shorter than 500ms in this scenario. You could take two timestamps with millis() - right after waking up and before entering sleep - to get an estimate of the actual wake time.

If you read the sensor every minute, but send new values only every 5 to 10 minutes, two AAA alkaline batteries should rather last two years than two month (provided that all components can work in the full voltage range of the batteries).

Note that writing lots of debug messages to the serial port or using blinking LEDs as RX/TX indicators may add a considerable amout of time spent in an active state.

[DenisJ]

@BearWithBeard said in Battery life calculator:

Did you measure the wake time of 2s?

No... I have inserted a wrong wake time,
cause now I have misure the wake time using millis() from initial LOOP
until before sleep command, and and I have only about 600 millis (sometime 300 sometime 900).

Then I send the data only when the temp have a difference of 0.3°C.
And the humidity only at 5% difference.
Also I put 2 minutes (really 128 sec because of 8s X 16 in order to use only thw WDT of 8Sec)

And now the tabel is like this:

But it give me only:

Yes... I have blinking LEDs as RX/TX but now I'll take off it and I'll blink only one time at every 2 minutes.
I'll see the measurements.

I will not discharge the batts too much because I dont buster for the voltage.
The 2 batts (3V) it's go directly to the Atmel and all the rest.
But I use an internal clock of 4mHz and If I understand correctly this will discharge the batts. until 1.8V of total.

Here I paste is the prototipe (cause I know you all like the pictures:-)

Thanks a lot
Denis

[BearWithBeard]

@DenisJ With 120 wakeups per hour as per your calculator screenshot, you'd wake up every 30 seconds instead of once every 2 minutes.

You should easily be able to get more than a year of runtime with that hardware setup and the mentioned thresholds for temp/hum.

[DenisJ]

@BearWithBeard said in Battery life calculator:

you'd wake up every 30 seconds instead of once every 2 minutes.

Whoawwww... I had entered wrong times
Now I take 409 days... great !!!

Thanks a lot again
Denis