The Great Moisture Battle (Part 1)

So my Night Sky Pi has been happily clicking away in the garden for a while now, and while the software side is running smoothly, I’m fighting a losing battle against an enemy I didn’t anticipate: moisture.

The good news? The housing is doing its job keeping rain away from the electronics. The bad news? I’m getting this annoying condensation inside the camera dome that’s ruining shots. It’s not even the moisture itself that’s the problem - it’s what happens to it. Throughout the day, as temperatures change, that moisture either fogs up the dome like a teenage bathroom mirror or forms little water droplets that make my photos look like they were taken during a monsoon. Hello, dew!¹

Dew Point (AKA The Enemy)

Here’s what’s happening: the air inside the dome is holding a bunch of water vapor, but as the temperature drops at night, that air can’t hold as much moisture anymore. There’s this magic temperature threshold - the dew point - where the air basically says, “That’s it, I’m full!” and starts dumping its moisture onto any available surface (in this case, my camera dome).

You see this effect most mornings with those little droplets on grass blades. Fun fact: grass gets extra dewy because some species are really good at radiating heat away, creating their own little micro-climate that’s cooler than the surrounding air.

My Battle Plan Options

After diving into forums and research papers (yes, I went down that rabbit hole), I’ve found three main approaches to fight this moisture menace:

First up: seal everything completely and toss in a bunch of silica gel packets. Nice in theory, but practically impossible in reality. Creating a truly hermetic seal outdoors is a pipe dream, especially with something constantly exposed to the elements.

Option two is the heating approach, which seems to be the go-to solution for most astronomy folks. The idea is to monitor temperature and humidity inside the dome, calculate when we’re approaching the dew point, and then kick on a heater to warm things up just enough to keep the moisture in the air and off my lens. It’s like having a tiny environmental control system.

The third option is ventilation - basically creating constant airflow through the dome to keep swapping the humid inside air with outside air. This hopefully prevents any significant difference between the two environments, reducing the chance of condensation.

Further Investigation

Alright, so before I dive headfirst into any solution (my usual approach that gets me in trouble), I want to really understand what’s going on. I need hard data on the differences between the outside world and my camera dome’s little ecosystem.

My plan is pretty straightforward: I’m installing two environmental sensors on the Night Sky Pi setup. One’s going inside the dome right next to the camera, while the other will be outside but protected in a little Stevenson screen (fancy name for “louvered box that keeps rain and direct sun off weather instruments”).

These sensors will take readings every 5 minutes - temperature, humidity, and calculating the dew point. I’m curious if there’s a massive difference between inside and outside, or if they’re tracking pretty closely. I’m planning to let this experiment run for at least a month to capture all sorts of weather - dry days, rainy days, still days, windy days… you know, the full English weather experience.

While my sensors are doing their thing, I’m also going to see if I can cheat a little. The Night Sky Pi already pulls data from the Open Weather API. If that data ends up being close enough to my outdoor sensor readings, I might be able to skip installing external sensors on future builds. One less thing to maintain would be lovely!

Next Steps

I’m going to let the current Night Sky Pi keep doing its thing for now, even with the moisture issues. Meanwhile, I’m tinkering with another instance to test out different circuitry without messing up the one that’s already deployed.

This whole experiment is on hold until I can get my hands on another environmental sensor of the exact same model as my current one. I want to make sure my readings are consistent and I’m not just comparing apples to oranges here.

Once I’ve collected enough data, made some pretty graphs, and convinced myself I actually understand what’s happening, I’ll share the results and my plan of attack.