Pro-Stick Plus and OrangePi PC at -16C

Got my Pro-Stick Plus and OrangePi PC back up on the roof, just in time for their first winter outside. They’ve been running for just over a week, and in the past 24 hours the outdoor temperature has plummeted. As I write this, the outdoor temperature is -16C and is expected to drop to -26C tonight.

While 24 hours of nasty cold is too short a period to reach any definite conclusions, it appears that as the temperature drops, the noise level drops (-37.8db Avg) and the message rate climbs.

One interesting tidbit is that after the core temperature of the Orange Pi PC drops below 0C, it no longer reports a temperature at all. Using Armbian, where the login screen normally shows the core temperature, the Temperature line is now completely absent, and even armbianmonitor just reports 0.

As the outdoor temperature approached 0C, the core temperature remained around 10C and gradually slid down to 0C as the colder weather moved in.

It will be interesting to see how this setup will behave when it gets REALLY cold out here.

Seems this episode was shot in Saskatoon :smiley:


I would watch this movie :wink:

In all seriousness, it’ll be interesting to see how the Pi works at that low a temperature, I’m sure it’s well outside the specified temperature range of it.

Correction: OrangePi DOES report core temperatures below 0C.

It seems that even though the outdoor temperature dropped and remained at -16C overnight, the core temperature of the OrangePi PC dropped MUCH slower. While it still doesn’t show the temperature on the Armbian login screen, this morning the core temperature is showing in armbianmonitor as -3C and on the Portal Graph (with scale minimum changed to -20C) I can see that after remaining at 0C for about six hours, it HAS finally dropped below zero.

Perhaps the aluminum box that was intended to protect the electronics from the elements is retaining more of the heat from the circuitry than I expected.

Well, a “polar vortex”, as it apparently is called, has settled in over us and we’re in for an extended period of really nasty cold.

The outdoor temperature is now -26C and the wind chill is around -33C. Humidity is still 77% which makes for a lot of ice fog in the air. It’s going to get colder yet over the next few days.

The OrangePi PC and the Pro-Stick Plus outside on the roof continue to do their duty, with the core temperature of the OrangePi having now dropped to -11C. Sensitivity appears to keep increasing and the minimum noise level on the portal graphs has so far hit -41db.

Feel lucky to be in Toronto, -5°C, windchill -9°C (23°F, windchill 16°F)

The OrangePi PC and the Pro-Stick Plus outside on the roof continue to do their duty, with the core temperature of the OrangePi having now dropped to -11C. Sensitivity appears to keep increasing and the minimum noise level on the portal graphs has so far hit -41db.

Great to know that Orange Pi PC and Pro Stick Plus can sustain such low temperatures.
It seems that Orange Pi PC (15) and Pro Stick Plus (Can 20.95) are the good investments I have made.

Great results. I’ve done the same thing with satellite lbs using peltier cooling. It’s like supercharging . A little drop in noise is a big deal the Germans had a nice YouTube on it but it has been taken down for copyright reasons

Hi yxespotter,
You get some pretty extreme conditions over there in Saskaton.
I see that the local airport YXE is open today and flights are generally on time.
Here in the UK we would fail massively at managing these sorts of conditions as we are not kitted out for such extremes.

Reading your post I wondered if trace heating tape may assist with the situation.
It is relatively cheap to buy and easy to install.

A random search for the product in Canada produced this, available in a 24 foot length which may be enough to route outside from an upper floor room to your station pi enclosure/s on the roof. … s-0522584p.

I assume getting to the roof location now will be hazardous and going outside to fiddle in -24 degrees is not my idea of fun… but perhaps a project to consider after the thaw.
Now your area forecast doesnt look to be getting much warmer in the near future.
Take a very large brandy or three and consider your options over the Christmas period.
Tip: For the Brandy, No need for icecube, just snap off one of those icicles for naturally purified H2O .
Stay warm !
Seasons Greetings.

Lower Temperature of electronic hardware >> Lower Thermal Noise >> Weak Signals are now above noise floor and hence detected.

Cooling microwave components and LNAs to cryogenic** temperatures enables significant reductions in the operating noise temperature (Top) of receiving systems. The sensitivity of a receiving system is directly proportional to A/Top, where A is the receiving antenna’s effective collecting area. For example, when Top = 80°K (-193°C, -315°F), an array of four identical antennas and receivers is needed to equal the sensitivity of one such antenna and receiver with a Top of 20°K.

**It is not well-defined at what point on the temperature scale refrigeration ends and cryogenics begins, but scientists assume a gas to be cryogenic if can be liquefied at or below −150°C (123°K; −238°F). The U.S. National Institute of Standards and Technology has chosen to consider the field of cryogenics as that involving temperatures below −180°C or −292°F or 93°K.

At the dawn of satellite communications, it was common to use a MASER as a low noise amplifier (LNA). Since broadband communications are difficult with MASERs, this led to development of parametric amplifiers featuring extremely low cooling using liquid Nitrogen. This followed by development of various types of cooling systems resulting in widespread use of easy to maintain parametric amplifiers using normal temperature or electronic cooling.

Subsequently, most systems started using FET transistor. The amplifier using gallium-arsenide (GaAs) Schottky-barrier FETs exhibit excellent low-noise characteristic, broadband characteristic, and high gain compared to conventional bipolar transistor. The reason why these FETs have better low-noise characteristic is that the source of their noise is mainly thermal noise, which compared to shot-noise of bipolar transistor, has a small rate of increase at higher frequencies. More over, the high electron mobility transistors (HEMT) has made it possible to reduce receiver noise still further.

Hi Radiostationx ,
Although having moved to western Canada from across the pond several decades ago, I still recall stepping off the train without the ‘proper kit’ for the -38C of that day.

Our temperatures in the -20C range are considered normal in the winter, although anything below -30C is extreme, but is expected every year for a couple of weeks. Believe it or not, one actually does get used to it, and it’s a matter of being prepared.

I don’t recall if YXE has ever closed due to the cold, although there have been delays for snow clearing. I’ve often flown out at -30C and would much rather take a heated jetway to a 737 going anywhere warmer, than walk across the tarmac to a King-Air going to somewhere just as cold or colder.

Thanks for the heat tape idea. At this point I’m not overly concerned about adding heat to the mix. With the intent of minimizing noise, and knowing that the electronics would need to survive some extreme weather, I built (overbuilt?) a housing out of 1/4 inch aluminium plate and some surplus milled-aluminium boxes for each component. I’m actually surprised at just how well the internal heat generated by the CPU and the dongle is being retained. Of course, I’ll change my tune in the summer when getting rid of heat will become more important.

Cheers, I’ll keep the icicles in mind.

Flightaware - orange line
Orange Pi - red line (temperature sensor has limit -25C)
Thermometer sensor - blue line. Minimal temperature -36C