RTC with low temerature


I’ve got a problem with my RTC on Q2686. In trial, the RTC works fine, but now the system is installed outdoor for further tests and now, the RTC stops sometime.

The stoptime is random.

My system is supplyed by Li-Ion akku and Solarzells, so it’s imposible to have peaks on supply voltage. The backup battery is installed (instead the akku is connected at all time).

I supose it’s because of the temperature of -5°C… but i have a Class B (-40°C to +85°C storage and operating -20°C to 60°C)

Does anybody have some experience?
It can’t be the software, it works at room temperature

can it be the same problem?



Your module may be rated down to -40, but is your battery?

I’ve had a lot of experience deploying battery powered devices into extreme cold climates, and I’ve found that the battery usually gives out way before the silicon does.

The Q2686 will automatically shut down if the battery voltage goes below about 3.2V. This is configurable though - look at the AT+WBMI commands in the AT user guide. You can also get unsolicited notifications of the battery voltage through the same AT command.

My first thought would be to enable the unsolicited battery voltage responses and log them somehow to see if your battery is giving out…

Another issue you may have is that the clock crystal (in particular the RTC crystal) may drift signifigantly with temperature - depends on the crystal. So unless you have a way of disciplining your RTC (i.e. NITZ messages or similar), the module Date/Time (AT+CCLK) value is going to drift all over the place.

Please let us know how you get on.

ciao, Dave

Hellow dave

thank you.
I have to test out if it’s the battery or not (but with -5°C only …). But what’s about the backup-battery? The main-battery can not go down for such a long time that the backup runs out of power, it’s imposible.

I can’t find the AT+WBMI command, is it new?
I have the AT Command Interface Guide for OpenAT Firmware v7.4 (rev 011) from April 2009 and can’t find a new one

Edit: Found! AT+WBCM you mean? But there’s a smal problem… i don’t use the Wavecom Battery-Charging-Modul because i charge only with solarcells and the PTS says, that’s important to have a constant voltage, isn’t it?

Yes, that’s it!
Also, the unsolicited indications are +WBCI:

You can still enable the battery monitoring without using the charge control.


Oops - as you found out, I really meant AT+WCBM :blush: Sorry about that.

I’m a little confused about your set-up - is this what you have:

Q2686 supply (VBATT) constantly connected to Lithium-Ion Battery that has an external charge circuit (supplied from your solar cells).
RTC battery (CR2032 or similar) connected to RTCBATT pin?

The RTC battery is only for supplying the RTC - it’s not a backup battery for the module.

The Varta website indicates their CR2032 batteries have an operational temperature range from -20 to +70, but my quick search of the Varta and Energiszer sites couldn’t find temperature derating curves for these batteries.

I have also noticed that the RTC battery drains extremely quickly on some of my modules - but I put it down to cheap, no-name batteries. I have some suspicion that the example circuit supplied in the Wavecom tech docs is not correct, but haven’t checked this out yet.

ciao, Dave


There “battery” is a capacitor with 470uF

correct, but the module is at sleep mode (alarm mode) and don’t use much power.

Yesterday, the module worked and the realtime clock hasn’t stoped. I installed a logger for the supply voltage, let’s have a look what it says


This sounds a little small to me.
Have you checked the Temperature characteriestics of the cap that you are using?

Have you read Section 3.1.2 - RTC Back-up Supply in the Q2686 Customer Design Guidelines (I’m looking in Q2686 CDC rev 3, Feb 9 2006)?

In this document from Wavecom, they state that the following are suitable for RTC backup:

  • A super capacitor
  • A non rechargeable battery
  • A rechargeable battery cell.

and that from a 0.47 Farad supercap you can expect a minimum of 25 minutes before the RTC resets.

I would strongly recommend that you have a look at this document to see what the recommended guidelines are from Wavecom.

ciao, Dave

Indeed - did you really mean micro-farads :question:

That’s 470 milli-farads - which is 470000 micro-farads :exclamation:

Indeed - and also check your orders of magnitude carefully!

Would you like to borrow my low-temperature test chamber? :laughing:

I’ve still got the Problem and now i’m shure it’s not the battery-voltage, the main-battery is at least 3.5V everytime.

and yes, the capacitor is 470 micro Farad … I didn’t think i realy have to use it anytime (because of the main-battery) and those 80s are enough for maintanance interrupts, perhapts it isn’t.

Tja, everybody learn :slight_smile:

@awneil: i’ve got nearly the same test chamber at home, but if you like, you can bring it to switzerland (but with snow on roof) :smiley: