INITIAL DESIGN

The weather station’s original rain gauge was a disappointment. It worked for only a season before it started to give me trouble. When I opened the enclosure to troubleshoot the problem, I discovered signs of corrosion. Water had damaged the circuitry. I guess there was no surprise there, but it wasn’t something I had anticipated. The socketed ICs were no longer making connections in the sockets. This problem seemed easy to solve. I decided to simply “pot” the circuitry in epoxy so no water could get at it, just like they do with automotive modules. Unfortunately, I couldn’t fix the rain gauge right away, so I had to use the “manual” rain gauge for a few years.

During this time, I started another project involving liquid levels. I needed to measure the water level in a cistern. No problem. I simply built a longer version of the variable capacitance rain gauge unit. And that’s when I found a problem with the design that I wasn’t able to solve. When the variable capacitance sensor was allowed to remain in the water, its capacitance changed.

Despite having tried several different sensor configurations and wire insulations, I couldn’t build a sensor that would produce a stable liquid level reading from one day to the next. I ended up abandoning the variable capacitance sensor for that project and using an electronically simpler approach that was mechanically more complicated: a float with a cable pulley arrangement connected to a 10-turn potentiometer. It worked great. There were no stability issues.

By the time I finished the water level project, I had come to the conclusion that if I ever got around to upgrading my weather station, I wouldn’t use a variable capacitance sensor, despite its elegance. I’d have to find something better.

The station’s weather vane has worked fairly well since 2000. I’ve had only one minor problem with it. The magnet that rotates over the reed switches to activate them fell off the aluminum disk to which it was epoxied. Apparently, epoxy doesn’t adhere well to magnets. So, I had to reattach it with another adhesive. It has stayed in place ever since.

Eventually, there will be a problem with the reed switches because the vane constantly waves back and forth in the wind. The trees may be too close to my weather tower. The waving constantly activates several reed switches and will certainly lead to switch failure.

The rain skirt that’s supposed to keep water out of the top bearing that supports the vane arm will also cause a problem in the future. It’s just a thin plastic cup (a plug for a cardboard mailing tube) that will surely fail in the sun and rain. It’s still intact, but I’ll have to replace it if I want the unit to remain in service.

There’s an aesthetic problem with the weather vane too. The lower part of the unit is housed in a steel box that has holes bored into it to allow air to circulate. The lower box also houses the relative humidity sensor. So, even though the box is painted, rust has blossomed from the holes and it’s extremely unattractive. The lesson here is to use aluminum because it doesn’t rust.

The anemometer in the original system has been a stalwart unit. I initially made a serious mistake with it, however. In an effort to have it work in the cold, I took the shields off the bearings and washed out the grease. After a year or so, I noticed the bearings were noisy. It turns out that when there isn’t any grease on the bearings, they are subject to condensation and rust.

The unit has continued to work though. Oil has been the only lubricant on the badly worn bearings. This anemometer has the same rain skirt as the weather vane.