Has anyone any thoughts on how to house a temp sensor outside.
Its just a ds18b20 on a long wire. So no other components to worry about.
Ive had one hanging outside my window for a couple of months but now Im getting some spurious results.
The one inside is fine and reading true. So Im putting it down to the elements.
Its just soldered heat shrinked and then wrapped in some tape with the plastic case poking out.
Im going to try a replacement and see if that gets rid of the strange values.
Should it be in a closed sealed box or should I alllow air flow and just keep the worst of the weather off.
If its in a closed sealed box I should imagine temp changes would take longer to effect the temp sensor.
But I know it must need some protection.
Should I encapsulate the leads and joints in some hot glue or another compound leaving the plastic case protruding?
I would leave most of the plastic case exposed. I would get some outdoor silicone
caulking and encapsulate the bottom of the case including the package leads and
the solder connections.
How did you solder the wire leads to the package leads? For an application that will
be exposed to the elements it is important to have a mechanically stable connection
prior to soldering. I would either tightly wrap the lead and wire or loop the wire
around a loop you bend in the lead. You could also mount the sensor on a small piece of vectorboard.
Well I have a commercial wireless rain gauge/temeperature unit that communicates with a base unit that also has a temperature sensor and displays the data from both units.
I've notice that if direct sunlight is hitting either unit that the temperature measurement goes way too high. As long as I keep the units in shade they seem to measure temp accuratly and track well, so some kind of box with insulated or relective covering with an open bottom would probably give better results. Keeping wind off the sensor also would probably help.
The temperature sensor needs a radiation shield to protect it from direct sunlight, but still allow air circulation to get an accurate air temperature measurement. Here is an example of one made with sheet metal. Another method involves several plastic dishes for a small flower pot. They are oriented upside down, all but the top one have most of the bottom cut out, and they have spacers between them. IIRC they should be painted white.
Look at some of the commercial weather stations for ideas, then rummage around the kitchen or local discount store for parts.
The sensor grenade gave me the idea that a small white plastic tablet tub upturned would do the job.
One of our cats has continual mediaction for a thyroid problem so Ill grab the next empty one.
Maybe with some foil wrapped over it to reflect the sun and leave the bottom open.
Ive found that even a white box in the sun doesnt reflect enough and warms too much from sunlight.
The box hasnt enough ventilation either.
I got around 22C yesterday in the sun when I know it was around 11C.
It's always going to be difficult to monitor temperature in the sunlight. The best thing to do is mount the probe where it is in the shade. Even if you do successfully shield the sensor from direct radiation, the area the sensor is mounted in will likely be warmer than the general ambient temperature. You will get some back-radiation from other surfaces and you may get convective heating from heat flowing up from objects below the sensor.