Practices for deploying in sub 0 environments during winter

Dear everyone,

I am soon deploying my “arduino” (it’s a waspmote, but fundamentally is an arduino) project and I have the concern that some things may be affected in the winter, when here in Denmark goes below 0. Not so often and not for prolonged period of times, but we do reach -24 C sometimes. Generally I’d say most places -5 to -15. But not constantly.

I read a little about this in the forum (suggestion like using multiple resistors, in series/parallel to spread the heat around the edges of the enclosure) but I was wondering if you guys think this is necessary for this temperatures? I am using a IP67 enclosure and I was wondering if you more expert people could help me to compile a list of TO-DOs to optimize the deployment of arduino in my situation because I am certain i would help other too.

Another point was how and what to use to insulate the arduino to maintain heat and prevent condensation.

I’m excited to hear what you guys suggest.

PS: Not sure whether this is relevant but some of my waspmotes/arduino will be in deepsleep mode for MOST of the time (sending data to my server via GPRS module once a day) and others will be doing almost the same but measuring every 15-30 minutes, sending data only once a day, plus whenever some alarm of what I am monitoring is triggered. Just to give you some context here.

Put you project in a refrigerator freezer over night and see if it is working in the morning. Styrofoam is the usual and inexpensive insulating material.

It is guaranteed to be working at -10C, however it should work at -20 too (what they told me). The RTC can measure until -40, its just that not all the components support -40. I'll put it in the freezer over the weekend and let you know guys. Please if someone have some idea especially from experience let me know.

I was thinking about buying some Polystyrene inside my IP67 enclosure edges. Then place some cell foam and cave in forms for the board and battery.

I will update you.

I doubt your electronics will have issues with the cold temperatures. Your batteries will probably be a different story.

I am currently using a JeeNode to monitor the outside temperature. So far it works great where I am. To has gotten down to 2 F with no issues. I did use the freezer method in testing it a while back when it wasn't so cold.

Batteries I have should hold just fine, otherwise I’ll use LiSOCl2 that go to -40 or -60 C :wink: Ill do the freezer test though to see how it goes.

My Arduino project with RTC has been fine in a chicken coop this winter with no enclosure at all. It hit -7°F / -22°C once, with many more days about 10°F / -12°C, no problems.

My Arduino project with RTC has been fine in a chicken coop this winter with no enclosure at all. It hit -7°F / -22°C once, with many more days about 10°F / -12°C, no problems.

Forget the Arduino, how did the chickens cope :)

-40 is a standard temp rating for commercial/industrial components, I would not assume that Arduino uses extended temp components however it seems people have been using them with no problems. I think -20 is standard for most components and that's about all you need anyway.

It will be interesting to see the results of your freezer test.

If the project runs very infrequently there will be little or no self heating, so will insulation help much? It will slow down the cooling but if you have constant -10 eventually everything get to -10, insulation or no insulation.

If there is no moisture in the enclosure you won't get condensation, maybe you can close the box in a dry environment and throw in a packet of desiccant for good measure.


Rob

The arduino that wake up less frequently are gonna be somewhat indoor while the one really outdoor will run every 10-15minutes (or 30 min) so I hope that will help.

I bought some silica gel packet to put inside my enclosure..

I will keep you all posted :)

I have a weather station (http://lar3ry.comlu.com/weather/)powered by NiCads charged with a very small solar panel (about the size of those solar garden light panels). This winter we have seen drops in temperaure to -35C for a few days or a week at a time. I occasionally lose the signal from the unit, but it doesn't seem to correlate with the low temperatures, and signal loss could be due to heavy snow or hoarfrost. It sends a burst of data every 10 seconds or so.

I think that for the most part, you will have no trouble running in your climate. Sealing the box in a dry air environment and using silica gel is likely going to help a lot.

This is a preliminary amount of data I gather in the past 36 hours (sending every 30 minutes to speed up the simulation process, which in real life will send every 24hours) (see attachment)

Although it looks like it dropped a lot in the start, I am 99.9% sure that is because as I have been explained by libelium (the producer of the waspmote) that when the battery is fully charged (i.e. is around 4.2V), the level you read is not so accurate and you have to wait for the battery to drop down to around 3.7V where it will stay for most of its life and it will stabilize.

So over "1 year" after (data sent every 30 minutes for over 380 times) the battery dropped only to 85% form the initial 99, and then first 5% went down right away with the first 6 measurements. so it stayed pretty stable.. :)

This makes me hope good :D