Article by Simon Mitchell – VK2YU

Overview

The aim of this project was to build an ADS-B receiver using the popular PiAware implementation (created by FlightAware).

The installation and configuration of PiAware is very well documented so I will not elaborate on this in this text only to point interested people to FlightAware’s PiAware URL:

https://www.flightaware.com/adsb/piaware/build/

The antenna I have used in this case is one that came with my DVB-T dongle and has been cut to the appropriate length for the frequency of ADS-B signals being 1090Mhz. This seems to work quite well (surprisingly) and will suffice for where the receiver is intended to be located.

I wanted this project to be self contained with its own internet connection and the only external requirement was power (12v DC @500mAh). This removed the need to piggy back onto an existing WiFi and being mindful of data usage (expected to be low in anycase). It also means it can be setup up as easy as putting the antenna on a tin roof as high as practical and plugging it in. The enclosure needed to be weather proof as far as possible however my previous attempt in an ABS enclosure was clearly going to suffer from heat issues.

The components selected in the end were:

• Power Supply – 12v2A (junk box)
• Raspberry Pi Zero W -with a heatsink attached from the last project
• 16GB SD Card flashed with PiAware v8.2
• Cheap DVB-T dongle that had been lying about for a number of years (with antenna)
• Telstra 4GX USB/WiFi Modem (found on sale for $24)
• Variable voltage regulator (similar to Jaycar XC4514)
• 50mm PVC tube and fittings (approx. $25)
• An old plastic A4 binder
• 40mm Fan (12v) – Jaycar
• Waterproof multi-pin plug and socket (junk box)

Construction

I initially built the PiAware solution and had it all housed in an ABS box (apart from the antenna). This seemed to work OK until we had a very warm few days recently…. While testing the solution I found it was running very hot due to the high ambient temperature, the heat from the individual components was pushing the temp in the box up above 50◦C. Looking at the RPi dashboard it was running at over 54◦C constantly. On cooler days below 25◦C the RPi temperature was usually below 44◦C. So the ABS box was not going to work long term. I started to think about alternative solutions, most of which revolved around mounting everything in a larger box which always ended up more expensive and only resulted in a larger box to heat up. This would only delay the inevitable heat build up on warmer days.

Container 2.0

I ended up mounting the components inside a vertically mounted PVC tube and managed to include a fan inside the lower end to force air over the components (Power reg, Pi, 4G modem and DVB-t dongle in that order due to individual operating temperatures.)

This project will likely be mounted outside (under a veranda or awning) however I still wanted to make it as weather resistant as possible without a huge expense. PVC tube was easy to find along with the fittings and in the process, I found some “insect” proof end caps designed for “water harvesting” that had a finer metal mesh rather than the cheap and easily damaged nylon version. These end caps also ended up being cheaper than the nylon version somehow!

Carrier board

The next issue was how to mount the components in the tube that allowed easy extraction in case anything went wrong down the track and a rebuild was required.
Eventually I (literally) stumbled on the answer finding an old flexible plastic A4 binder under the work desk. The front and back covers were an opaque plastic that was rigid while still flexible enough and was easy to cut with a good pair of scissors.

I cut a 40mm x 320mm length to make the carrier base and then worked around mounting each component to allow maximum airflow around each device with the fan mounted at the end of the strip. The carrier base/board slides into the PVC tube and holds all components along the length of tube with a short tongue below the fan left to secure the 12v input wires and allows for ease of removal.

Fan & Power regulator

As the power supply is 12v, the fan connected to the Power regulator input then 5v went out and onwards to the Pi and the 4G Modem.

RPi Zero W

I soldered figure 8 cable directly to the back of the RPi to the 5v and GND test pins PP1 & PP6 located behind the micro USB power socket. See below:

The Pi already had a heatsink attached and this was left on to allow better cooling. It is minimal in size and caused no additional issue for mounting in the PVC tube.

4G Modem

The modem I found when I first started thinking about this project and I happened to find one on sale a couple of days later. This modem is based around the popular E3872H broadband modem and includes the B28 (700Mhz) Telstra band that is found in regional areas where this project is to be located. I picked one up for $24 which included a starter SIM and 10GB of data. (Perfect for testing!)

https://www.telstra.com.au/internet/mobile-broadband/prepaid/telstra-pre-paid-4g-usb-wi-fi-plus

(Update, Sept 2025: the link above is broken, as the Telstra Pre-Paid 4GX USB Modem is no longer available. It seems to have been superseded by the Telstra 4GX MF833v USB. Note that the modifications described below might have to be changed to work with the newer model.)

As I read up on these devices I found the Telstra version has a unique model number: E3872H-608. The “608” seems to be specific to Telstra. Perhaps to designate the B28 LTE band?

As I’m only using WiFi on this modem the USB port is not required, so in the interests of saving space I opened the 4G modem up and removed the USB connector, carefully desoldering all pins and re-soldered on a short length of DC figure 8 cable to the 5V and GND pins.  This saved the end cap and the USB plug length which is just under 3cm. not much, but enough to warrant the effort and to get a look inside the modem as well.
-All antennas are foldable trace antennas stuck to the inside of the back case, so fully de-casing the thing wasn’t going to work. The modem can run a little warm so I did end up removing the front case panel which is really just there to show the carrier logo and maybe to protect the SIM and microSD from being easily removed. I thought at least removing the front panel might let a little more heat out if anything.

DVB-T dongle

The DVB-T dongle I had lying around from past tinkering. This is not a precision version like the RTL-SDR models (which I also own, and think are fantastic). This one had a small window in the outer case which was for an IR receiver and a blue power LED. Opening the case both of these were removed as (IR not required and the LED only uses power for no purpose especially inside a PVC pipe!) The dongle runs fairly hot so it is now fully caseless. Again, I desoldered the USB plug (full sized) as it doesn’t fit directly into a RPi Zero. I could have used an OTG cable but decided to directly wire a micro USB lead to it which kept size down once more. The other benefit in this project was to maximise the airflow rather than choke the inside of the tube up with cables and adaptors. The innards of the dongle were then mounted standing up on the carrier board and secured with an L shaped piece of the carrier board plastic and a zip tie. The antenna socket on the device was pushed through a 5mm hole in the carrier board and by connecting the antenna lead on the other side this helped hold the dongle in place.

The final mounting solution

Shown below is the final mounting carrier and components fixed prior to installing inside the PVC tube:

L to R: 4G modem just in shot, RPi Zero W, power regulator board and fan

L to R: DVB-T dongle, 4G modem, RPi

The top cap was made from 2 x 90 degree fittings and the insect proof breather. A BNC adapter lead was installed through the first PVC bend:

Looking in the top end of the tube. Components on carrier board installed:

The completed mounting
BNC socket mounted under the first top 90 degree joiner connected to antenna (with magnetic base). (the power socket had not been mounted at this stage but sits just lower than the shoulder of the bottom 90 degree elbow on the outside (right side) of the elbow.

Air is drawn into the tube by the fan and forced over the components then expelled at the top “U” turn to prevent water entering. I could have put another 90deg bend facing downwards at the bottom but the fan was high enough inside the main tube not to worry.

Testing and results…..

• Powering up and it seemed no smoke has escaped! The fan is a little noisy but not bad enough to redesign. In any case the results of the forced air cooling seem quite good.
  
  When testing the ambient temp was approx. 16/17 degC and looking at the PiAware dashboard it indicated the RPi CPU running at 20.7 degC as shown here >>.

-No, the weather wasn’t as hot as it had been when the RPi was at 54 degC but in general without the fan it would normally be around 29 – 36 deg C

• Inspecting the DVB-T dongle I found it was a little warm but again significantly cooler than it ever ran without cooling previously.

• The 4G modem was also much cooler than it had been without the fan as well (significantly cooler than in the ABS box even on a cool day)

The project was a great little learning exercise and all in all I consider the project should achieve what I set out to do. Soon the build will find its forever home, somewhere inland on the Mid North Coast of NSW and will hopefully see a long life all going well and improve the ADS-B coverage in the area.

Thanks for reading!

73’s
Simon
VK2YU