Moteino M0 - Cold weather suitability

[adam.g]

Hi there,

I've been following the development of the Moteino for some time now and I was very intrigued when the Moteino M0 platform was released. I'm interested in possibly using the M0 for a number of low-power research projects in the Canadian Arctic and I'm curious to know how it would fare under extremely cold conditions (-40°C/-40°F). Are there any components that would be particularly susceptible to the cold?

Thanks for your help!
Adam

[Felix]

All the MoteinoM0 main components (chips/crystal/flash) are spec'd at -40°C ~ 85°C (TA), the LDO is -40°C ~ 125°C. I cannot speak for the RFM modules though, because the transceiver chips are packaged in china under unknown conditions and certifications. The original SX1231h transceiver chip packaged by Semtech is QFN 24 Package - Operating Range [-40;+85°C] same as everything else, perhaps it's "safe" to assume the chinese QFN-28 packaging is about the same spec.
https://www.semtech.com/uploads/documents/sx1231h.pdf

FWIW - One huge thing about the M0 is it's sub 10uA low power ability, something you will not find in any SAMD21 competing product. I think this really matters for any battery node, especially in the arctic where you might want to limit service intervals

I think the real limiting factor will be your battery type.
Most common chemistries will not fare that great under extreme cold.

I'm curious what others have to say on this topic. I know that other types of Moteinos (not the M0 yet) has been used in high altitude balloons, and all kinds of extreme scenarios, and they worked great. I can't think of a reason why the M0 should not perform just as well.

[adam.g]

Hi Felix,

Thanks for the quick reply. That's great to hear the Moteino M0 components are rated down to -40°C. Currently, I have several iceberg tracking beacons that were deployed in the Arctic this summer (http://cryologger.org). They are based on the Adafruit Pro Trinket 3V microcontroller, but are really proof of concept as Adafruit wasn't sure how they'd operate at such extreme cold temperatures. I believe the oscillator onboard starts to act funny beyond -25°C, but they've been working well so far with temperatures already dropping down to -20°C.

You are correct that the battery chemistry is definitely a limiting factor. In my beacons I'm using Tadiran lithium thionyl chloride (LiSOCl2) batteries, which are capable of handling the extreme cold temperatures.

I'm a really big fan of the modularity of the Motetino M0, and am looking forward to placing an order once a few of the addon boards are back in stock!

Cheers,
Adam

[Kilo95]

what about using the SuperCap/Solar panel for power like Felix used on a Weathershield?

[redpoll123]

SuperCap/Solar panel work great here in central Alaska in June - daylight 22 hours,
not so good in December (when it is -40F) - daylight 3+ hours.
Alaskan (Canadian) Arctic June daylight 24 hours, December 0 hours.

[Felix]

Adam,
I don't think I've heard about LiSOCl2 batteries, woot - now I know!

If you're looking for the LSM9DS1 accelerometer/gyro/magnetometer add-on - looks like everyone is out of the actual chips until the new year at least
Everything else is available.

[Oyvindes]

A question about the Li-Socl2 batteries. They start at 3.6v and have a cut of at 3v. but the VIN pin on the M0 starts on 3.6v.
Is there a way to bypass this?

[Felix]

Are you looking for the lowest power consumption?

If so then there's not a practical way to do it unless you:

- remove the LDO and charger chips
- solder the "diode-bypass" solder jumper to avoid a schottky diode drop
- bridge the input and output of the LDO pads.
Then you can power from VIN.

Otherwise the LDO will start to eat a lot more quiescent current if you're not above 3.3v+LDO-dropout, and powering from VIN.
The reference MoteinoM0 schematic is here.

[adam.g]

A question about the Li-Socl2 batteries. They start at 3.6v and have a cut of at 3v. but the VIN pin on the M0 starts on 3.6v.
Is there a way to bypass this?

Hi Oyvindes,

I run my LiSOCl2 batteries in a series/parallel configuration so my operating voltage is ~7.2 V. My packs are also specially made with a hybrid layer capacitor (HLC) so that I can operate a high draw Iridium satellite modem (~400 mA) without damaging the packs, which are typically suited for low-draw applications. I haven't yet used this pack with the Moteino M0, but I'll be removing the charge controller if I do and avoiding connecting the battery and USB at the same time.

I know with the ATmega328 boards, which can operate at quite low voltages, it's usually just a simple matter of removing the LDO.

Here's the battery pack information if you're curious:
http://www.tadiranbat.com/pulsesplus.html
http://www.tadiranbat.com/assets/tlp-93121-b-al1.pdf


Cheers,
Adam

[Oyvindes]

Are you looking for the lowest power consumption?

If so then there's not a practical way to do it unless you:

- remove the LDO and charger chips
- solder the "diode-bypass" solder jumper to avoid a schottky diode drop
- bridge the input and output of the LDO pads.
Then you can power from VIN.

Otherwise the LDO will start to eat a lot more quiescent current if you're not above 3.3v+LDO-dropout, and powering from VIN.
The reference MoteinoM0 schematic is here.

It's a combination between long life and extreme temperature toleranse.
The Li-Socl2 batteries are a very well fit for this kind of device.

So a bit of hacking with the soldering iron might solve my problems...?