dht11 CR2032 longevity

[fhenryco]

Hello,

I'm wondering if CR2032 is an interesting option for powering a pro mini 3.3V (needs the step up regulator) running the temperature and humitidty sketch on this site for a DHT11.
How long can i expect the CR2032 to last if :

• the arduino sleeps 1 minute between each wake up and sendind of temperature&humidity
• the arduino sleeps 15 minute ...........

if it's no more than a few weeks i would extrapolate that there is actually not any sensor project where the CR2032 could be actually an interesting option in practice right?

so what is the best powering option (and arduino) to have such kind of sensor (sleeping almost all the time) running at least 1 year on a battery ? a standard 12V battery on a nano Vcc ?

thank you in advance for helping advices and comments

[hek]

The CR2025 is typically rated for 163 mAh. The thicker CR2032 is typically rated for 240 mAh.

You need to do some math based on the consumption of your node (and when the onboard sensors/mcu fails).

[fhenryco]

thank you,

No link ?
the problem is not the lack of information but the excess of informations on the web, often inaccurate, often contradicting each other, almost all the time incomplete ...
but is there somewhere a complete and rigorous reference guide/method on how to reliably compute those things not forgetting any major contribution? :
consumption of arduino, RF and sensor during sleeping time
consumption of arduino, RF and sensor during active time,
actual relative duration of active time ...

here i'm talking about one of the most basic and most used Mysensor sketch : the humidity sketch on the Build page, and i did not change any parameter (default RF power, default sleeping duration)
so i guess that people who have worked on the protocol to optimize energy consumption have already figured out this in details ... so i prefer to get this information from Mysensors admins and developpers than from anywhere else (often people not exactly using the same sketches and protocols ...)

Also i don't need an accurate estimation but an order of magnitude ...

regards

[fhenryco]

Found this interesting page
http://www.home-automation-community.com/arduino-low-power-how-to-run-atmega328p-for-a-year-on-coin-cell-battery/

but the gnd pin and quiescent currents of the minipros regulator LQ33 the guy is talking about in this page (80 microA at 0.1mA load current), may be much bigger than for my minipro which regulator has LG33 written on it , so if the following is the correct datasheet:
http://pdf1.alldatasheet.com/datasheet-pdf/view/308176/ELM-TECH/ELM1117LG-33-S.html
it has a typical 5mA quiescent current. This is much much higher than the previous (LQ33) one but the load current is also not specified so i'm not sure...
All in all do the analysis of the above page apply to the 3.3 minipro which i bought through the store on your site ?
It says that "Even though the standard voltage regulator is not efficient at low current, it is still good enough to run the 3.3 V Pro Mini on 3 AA Alkaline batteries for more than one year, based on the measured power consumption of 54 μA in power-down sleep and 3.90 mA in active mode."
Thank you in advance, it's the first time in my life i try to understand a datasheet...

[Nca78]

@fhenryco said:
CR2032 .... step up regulator

Save your time, the answer is no
You cannot use a step-up converter with a button cell, it will die in days at best.
You need to :

• use a sensor that will accept low voltage, for example si7021/sht21 or BME280
• remove voltage regulator and led on the pro mini (and use 8Mhz / 3.3v one)
• update bootloader/fuses so the BOD (which will reset arduino under a minimum voltage) is set lower, typically to 1.8V instead of 2.7V by default. For that you can use another arduino as a programmer, there are many tutorials around
• add a capacitor (100-200µF typically) parallel to the battery to help it during radio transfers
• in your script, go to sleep between consecutive data sending to let the capacitor recharge

[fhenryco]

@Nca78
thanks but ...
!! that's much less trivial than i anticipated... OK to use small voltage sensors and to remove Leds but :
Is it really necessary to remove the regulator ? my previous message linked to an analysis according to which even with a regulator like LQ33 (what about LG33 ?) the arduino could last one year on CR2032 if sleeping most of the time ... i really don't mind replacing CR2032 once a year ...

[fhenryco]

and if i can keep the arduino regulator (with no step up regulator) i could still use two CR2032 in series , true ?

[fhenryco]

@fhenryco said:
" Is it really necessary to remove the regulator ? my previous message linked to an analysis according to which even with a regulator like LQ33 (what about LG33 ?) the arduino could last one year on CR2032 ..."
And apparantly according this more complete datasheet for LG33 , which has the same very low GND current as LQ33 for vanishingly small load currents :
http://ww1.microchip.com/downloads/en/devicedoc/mic5219.pdf
it's not really mandatory to remove such regulator to get long live CR2032... no objection around ?

[Nca78]

From the link you gave, 54µA and 200mAh available that's 200 000 / 54 = 3704 hours = 154 days = less than 6months. (the one year claim is with AA batteries...)
And that's not including the wake up time and transmits.

Anyway if you keep the regulator, you have to connect the power on Vcc and not RAW, case not measured in the page you link. It should be better but you would still have an unnecessary leak of current just to save the 2 mn it takes to remove the regulator.

[fhenryco]

MMM OK, hesitating just because i'm not sure removing this regulator is going to be as easy and safe for board as i would like ...

just to be sure , if i keep the regulator what you say about connecting power directly to Vcc is in case the power is provided by a single CR2032, right ?

About your third other points (BOD modification) : this is necessary in case the voltager that the atmega eventually receives (either through RAW without regulator , or through Vcc) is lower than a regulated 3.3 V, do i understand well ?

about your last point (capacitor and sleep between consecutive transmissions): is this again related to the above harware choices for powering the board ? i mean is it made necessary again by the fact that the atmega is running on a less than 3.3V CR2032 ? just to be sure i understand

thanks again

[fhenryco]

@fhenryco
eventually it also appears that i have ordered step up regulators for nothing unless ... in which case might these be usefull and operate for a longtime ?

[Nca78]

@fhenryco said:

MMM OK, hesitating just because i'm not sure removing this regulator is going to be as easy and safe for board as i would like ...

I never had problem to remove it, try it it's easy !

just to be sure , if i keep the regulator what you say about connecting power directly to Vcc is in case the power is provided by a single CR2032, right ?
Yes, in that case voltage will be a bit over 3V with full battery and go down to 2V (but at 2.4V the battery will be neard dead already.

About your third other points (BOD modification) : this is necessary in case the voltager that the atmega eventually receives (either through RAW without regulator , or through Vcc) is lower than a regulated 3.3 V, do i understand well ?
Yes, this is in case you are using cell only.

about your last point (capacitor and sleep between consecutive transmissions): is this again related to the above harware choices for powering the board ? i mean is it made necessary again by the fact that the atmega is running on a less than 3.3V CR2032 ? just to be sure i understand

No that's necessary because you are using a CR2032. These type of battery has a high internal resistance, that will go higher and higher when the voltage decreases. It means if you draw a lot of current, you will lose power because of this internal resistance. Then it means voltage drops and as it drops internal resistance increases.... etc etc So you need extra capacitor to help the battery by providing a bit of extra current during transmit.
Using 2 cells will not improve that much, you might even not have better battery life life that as the regulator will consume a bit of power, and as your input voltage will be from 6V to 5V it makes a sharp drop in voltage meaning the efficiency of the regulator will be far from it's best.

thanks again

[fhenryco]

@Nca78

Last question : what about 3 AA LR6 in series then :

• also has the problem of a large internal resistance ?
• sharp drop of voltage again because voltage ~ 4.5 V still too much greater than 3.3 ==> bad regulator efficiency , right ?

[mfalkvidd]

@fhenryco 2xAA without regulator is what is recommended in https://www.mysensors.org/build/battery

[fhenryco]

@mfalkvidd

Oh! sorry, i had completely missed this battery section on this site "build" page. All right
There is just something that is not yet completely clear for me for this recommended usage of 2 AA batteries in series.
The boost 3.3 V regulator is presented as a good way to take "all the juice from batteries" for them to last longer, but at the same time, as a regulator it is expected to continuously draw current even during sleeping time ... and in this way reduce lifetime of batteries
so i remain a little bit confused as for the actual usefullness of such step up regulator : is it only usefull in a project where sensors are transmitting very often so that the sleeping current is not the dominant factor reducing battery lifetime ?

[mfalkvidd]

@fhenryco there is no agreement in the community. Some have had success without step-up, some with step-up.

[Nca78]

A lot of people don't change the BOD settings on the prominis and so when using only 2 AA or AAA the sensor will stop running when the batteries are around 1.35V which is a lot of waste. In that situation yes you will get better battery life with the booster.

But if you use a sensor that can go below 2V and set your promini to have 1.8V BOD and bootloader to run at 1MHz (so the pro mini can run below 2V) the batteries can go down a bit below 1V and at that level there's not much "juice" left in them. With that configuration you'll get a much better battery life than with a booster, and you will avoid the main disavantage of the booster: the interferences with the radio that sometimes make the sensors unreliable.

[Nca78]

And no, no problems with AA/AAA regarding internal resistance. You can draw a lot of power from them without problem, so you don't need a capacitor for that. Just keep the capacitor for the radio.

[fhenryco]

Now that i have received pro minis 3.3 i can practice !

following the method in http://www.home-automation-community.com/arduino-low-power-how-to-run-atmega328p-for-a-year-on-coin-cell-battery/
i just tried to cut the trace to the LED with a cutter, ... but the LED remained shining after several deeper and deeper attempts .... so i just removed the LED, but now my mini pro is not responding anymore: dead!

so i'm wondering if i have destroyed something else trying to cut this trace , and considering how easy it is rather to just remove the led, i'm wondering why the guy advices to cut the trace which is probably a more dangerous method for the pro mini

[fhenryco]

eventually, i was more skilled with a second mini pro : after removing both LED and board regulator, the consumption i measured is about 19mA when active and 80 microA when sleeping , i guess mainly because of the step up regulator (from 2xAA to 3.3V)

[Nca78]

@fhenryco I agree with you, I find it more simple to just remove the led. Cutting trace can lead to mistake (like you did on your first board) and there are a bunch of different promini clones, some might have traces using different paths than on the example.