OK it finally happened. My humidity sensor stopped working.
I woke up this morning and when I checked on the tank there was no humidity data. It looks like it stopped working at 8:49pm last night, and I didn't notice until this morning.
On inspection, I could see that one of the connectors to the temperature sensor was very badly corroded. I had used clear nail polish to coat most of the circuit board, but I must have missed that spot. This circuit board has been getting sprayed with fertilizer solution 3x a week during the automatic waterings. Since I missed that spot with the nail polish, I'm actually surprised it lasted this long because the fertilizer solution is probably super corrosive.
I tried to clean it up and get it running again real quick this morning so unfortunately I don't have any pictures of the corrosion. It was on one of the little black connectors on the red temperature sensing circuit board here:
I cleaned it up and plugged it back in but it still didn't work, so I left for work. When I got home I scrubbed it really good, cut off the connector, and soldered in the wires. After testing again I still wasn't getting data from the humidity sensor, so I swapped in the sensor from Tank 0 and tested that. It worked, so I guess the corrosion on the temperature sensor connector actually broke a different part of the circuit. Fortunately I'm not running any humidity control on Tank 0. I was just measuring it for fun, so I was able to use its humidity sensor as a spare. With the sensor working again I coated up the solder joints with a layer of clear coat, and now it's curing. I don't like to put uncured nail polish in the tank because I don't know how the plants feel about the fumes, and it seems to take forever to cure in the high humidity. I'll have to wait until tomorrow morning to get the tank back under control.
Here are the repaired and coated solder joints:
This was pretty typical timing for a failure. I just ordered a batch of 20 custom sensors about two weeks ago, and I was really hoping my system could make it until they arrived.
While it's frustrating to have things fail, it's not all bad. If you're able to track the failure down to the actual cause and fix it at the source, you're left with a more reliable system. This sensor survived almost 6 months of constant high humidity, wetness, and fertilizer spray. When it finally failed it was because I did a sloppy job with the waterproofing, and corrosion caused an electrical failure. The fact that the failure didn't originate in the actual humidity sensing element is an excellent sign for the next batch.
The next batch of sensors have all the electronics on a single circuit board, and there will be no connectors in the tank. The whole assembly will be spray coated with a waterproof conformal coating to protect the electronics. These are still being built, but I should have them in the next few weeks.
---------- Post added at 09:15 PM ---------- Previous post was at 09:08 PM ----------
Plants are still growing, there are a few in bloom but not too many, and I started some new moss:
Linear Mode
