Productionizing My ADS-B Feeder

Overview

Last September I started feeding ADS-B traffic to FlightRadar24 and FlightAware using a Raspberry Pi and a customized image. An antenna was first mounted on our 2nd floor deck and quickly deprecated by a roof-mounted antenna which required snaking the cable through my office window. The Raspberry Pi with its software-defined radio dongle and filter, precariously balanced on my window sill, ran without problems throughout the winter. Cool.

And then it wasn’t: while out of town, the Pi fell off the window sill – I’ll blame the cat – and I was offline for days. No one else to blame, I knew that I needed a more permanent, secured solution but ignored it for too long. A triggering event had occurred, therefore time to investigate possible solutions.

Requirements

Secured: The primary components – Pi, dongle, filter – must be immobile by attaching to some fixed surface, hopefully eliminating the possibility of the ensemble falling apart.
Enclosed: The primary components are protected from inadvertent or intentional damage by enclosing in a case with only the cables (antenna, ethernet, power) visible from outside the case.
Weather-Resistant: The case containing the components will be mounted outside, removing the need to bring the antenna through the window.

I have always been concerned about the risk of the antenna cable becoming damaged by bringing into my office; conversely, external placement means that the risk is now with the ethernet and USB-C but replacing those is less expensive than the antenna cable. A good tradeoff in my opinion.

Enclosure

I discovered the D-Line Outdoor Cable Box with help from a Menards employee. Its advertised use case is running extension cords safely outdoors for gardens or landscapes, and has openings through which cables are run, each opening covered by rubber grommets to protect from dust or water. The IP54 rating implies some level of water resistance but not water tight.

Pre-Installation

I did not want to pierce the box to secure the components which likely would reduce its protection against water. The approach I chose would use wood blocks glued to the box bottom to which the components would be secured.

To each block I stabled cable ties and then super-glued each. I applied the super-glue and then the block and box tightly together for eight hours to (hopefully) ensure long-lasting adhesion.

The two screws on the left block are positioned to match slots on the Raspberry Pi’s enclosure which provides further stabilization.

Installation

I first secured the SDR dongle and filter on the left, followed by the Raspberry Pi. The dongle is then plugged into the Pi using a USB extension cable.

There were a few challenges:

The length and tensile strength of the cable ties made it more difficult to necessary when tying down each component.
I did not determine the optimal position of each cable tie, so the components’ positions had to be adjusted to ensure that a tie would be useful. The Pi cable ties end up crossing the enclosure fan, hopefully not reducing air flow too much.
The USB-C cable port is on the outside edge of the box, which will make it difficult to get the cable plugged into the Pi.

Once all components are secured, I connected the antenna, ethernet cable, and power cable to test. Fortunately, everything worked as expected.

Mount

The box can be mounted horizontally or vertically with one particular advantage for vertical: air flow. When mounted vertically, the bottom grommets can be removed which allows fresh air to enter the box while (hopefully) replacing the air warmed by the Pi. A heavy rain forcing water through the top cable openings should drain out the bottom. without affecting. During winter, the grommets may be required to protect the Pi against higher humidity during snow storms. Despite a recent snow storm, I doubt I’ll need to worry about it for now.

The box’s bottom has two slots for mounting vertically. I used two 2½” x ½” brackets with the vertical leg screwed into the window frame, 9″ apart. Slide the box onto the horizontal legs. tighten all screws to hold the box securely.

[Sorry, I forget to take pics of this!]

After convincing myself it was fairly secured – it still slides on/off more easily than I like – I attached each cable and tested: voila, success!

The final step requires attaching the top of the box and ensuring all cables are secured.

Final Thoughts

Direct sunlight. The box receives direct sunlight for a few hours each day. When the outside temperature reached 76°F/24.5°C today, the Pi reported temp was 122°F/50°C; this evening, the Pi’s temp is 97.7°F/36.5°C. Despite reports of Pi’s operating at temps approaching 176°F/80°C, will Minnesota summers cause problems?
Cable Ties. How do table ties degrade over time as temperatures vary?
Wi-Fi. As addressed in a previous post, a Pi would not attempt to reconnect to Wi-Fi after a drop. I’d prefer not to use an ethernet cable but don’t current trust it.
Air Flow. I assume the enclosure fan will force hot air out to be replaced by fresher, cooler air. Assume. I have not added temperature/humidity sensors inside the box, which may be needed if the Pi becomes unstable.
Something else? What am I missing?