The Great Moisture Battle (Part 1)

The initial Night Sky Pi tests surfaced a clear problem: moisture inside the camera dome is degrading images, sometimes to the point of making an entire observation unusable. The good news is there is no water ingress that threatens the Raspberry Pi or electronics. The bad news is what the trapped moisture does as conditions change through the night and into the morning.

When temperature drops, water vapour in the air can no longer stay suspended and condenses on the dome’s inner surface. That shows up first as fogging, then as visible droplets. In other words, dew. The underlying mechanism is the dew point: the temperature at which the air can no longer hold its current moisture content. Reach or cross that temperature and condensation forms.

Options under consideration

1) Seal the dome with desiccant

Create a fully sealed compartment with silica gel to keep the air dry. In practice this is not feasible for a system that lives outdoors around the clock. Achieving a truly reliable long-term seal is far harder than it seems, and any small failure would shorten the life of the enclosure.

2) Heat the air in the dome

Measure temperature and humidity, calculate dew point, and use a small heater only when conditions approach the dew point threshold. The aim is simple: keep conditions just far enough above dew point that moisture stays in the air instead of condensing on the dome.

3) Actively ventilate

Continuously exchange the air in and around the dome so internal conditions track the outside environment. The hope is to remove moisture before it settles and to avoid the stagnant micro-climate that encourages condensation.

What to measure before choosing

Before committing to any one option, the plan is to understand how the dome’s micro-climate actually behaves compared to outside air.

• Two environment sensors will run side by side:
  • one inside the dome next to the camera
  • one outside in a Stevenson screen to protect it from the elements
• Sampling rate: every five minutes the system records temperature, humidity, and a calculated dew point for both locations.
• Duration: at least 30 days, long enough to see dry, wet, and still periods and to gather a baseline across different conditions.

In parallel, the same metrics will be pulled from the OpenWeather API already integrated with Night Sky Pi and the Open Weather Dumper. If the API data tracks the outside sensor closely enough, the external hardware could be avoided to reduce complexity.

Immediate next steps

• Let the live Night Sky Pi continue untouched for a while to observe whether conditions ever improve on their own and to build a clean baseline.
• Assemble a second Night Sky Pi to prototype heater and ventilation circuits without taking the live unit offline.
• Acquire another environment sensor that matches the one already on hand so readings are as comparable as possible.
• Collect, check, and graph the data, then publish the results before attempting any fix.

Progress on this work will be grouped under the moisture battle tag, with results and decisions to follow once the data set is complete and reviewed.

Thanks for reading. If you have experience running dew-point-aware heating or low-maintenance outdoor ventilation for small enclosures, share what worked so others can learn from it.