Squirrel Net Update: Flaky Components

Squirrel net has been running for only about 10 days. (It's my solar wifi node experiment.) There have been 2 failures so far--the solar charge controller failed, and the Raspberry Pi 3B has been flaky.

The charge controller that failed "HQST 30 Amp PWM Smart Solar Charge Controller with LCD Display" gave up the ghost after exposure to high humidity, but no moisture. The controller appeared to still function, but the voltage sensor on the battery terminals failed. It's not clear if it was a physical component failure or something else.

The Raspberry Pi 3B just fails periodically. It seems like it is probably a software failure. In fairness to the people who make the $40 gizmo, it was probably not designed for this application. The WiFi, for example, is relatively slow and low powered. The hardware also seems to keep plugging away in spite of exposure to the humid conditions lately.

Anyway, I'll probably replace the Pi with dedicated outdoor wireless hardware. It'll have better power and better reliability (hopefully) due to the packaging.

Squirrel Net Solar Wifi Network

original concept

I've wanted to do a hands-on solar project for several years and finally got around to it. My original concept was to mount a small panel on an enclosure that also serves as a squirrel or bird house. I built the enclosure and powered it with a small solar panel (25 watts) and a 7 Amp*Hour battery.

With winter weather and sun conditions, the battery and panel were not able to last through a couple of overcast days, so I quadrupled the panel capacity to 100 Watts and jumped to a 100 Amp*Hour battery. The larger panel seems to be a necessity. On a cloudy, snowy day where the sun isn't visible through the clouds, the panel can collect 8-10 Watts.

The Raspberry Pi is only drawing 4-10 Watts, so the system is able to top the battery off every day during peak sun hours.

The concept of attaching the panel and battery to a birdhouse or squirrel house is not practical. The battery weighs too much (70 pounds for a lead acid 100 Ah battery) and the panel is 2 feet x 4 feet, more or less.

current concept

The smaller battery and panel would work for a load that's not running 24/7. However, I wanted to use the Raspberry Pi as a wildlife/security cam system too, which will run 24/7, so the overcapacity is necessary.

From the cost perspective, it probably would make sense to basically have a solar power station back in the woods and power cables running around... but that's also pretty bad aesthetically.

After a trial run of a couple of months with the current design (at right), I'll do another iteration that's back farther in the woods. I think I need four additional hotspots to cover the entire property.