DIY Satellite Tracking

Travis Goodspeed has authored a blog post detailing his method of tracking low-earth-orbit satellites. Starting with an old Felcom 82B dish made for use on maritime vessels, he added motors to move it around and a webcam-based homemade calibration system. "For handling the radio input and controlling the motors, I have a BeagleBone wired into a USB hub. These are all mounted on the trunk of the assembly inside of the radome, sending data back to a server indoors. ... In order to operate the dish, I wanted both a flashy GUI and concise scripting, but scripting was the higher priority. Toward that end, I constructed the software as a series of daemons that communicate through a PostgreSQL database on a server inside the house. For example, I can run SELECT * FROM sats WHERE el>0 to select the names and positions of all currently tracked satellites that are above the horizon. To begin tracking the International Space Station if it is in view, I run UPDATE target SET name='ISS';. For predicting satellite locations, I wrote a quick daemon using PyEphem that fetches satellite catalog data from CelesTrak. These positions are held in a database, with duplicates filtered out and positions constantly updated. PyEphem is sophisticated enough to predict in any number of formats, so it's easy to track many of the brighter stars as well as planets and deep-space probes, such as Voyagers 1 and 2."

Re:Ok, maybe

[Anonymous Coward]

Maybe "low" is relative... ;)

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Reads like TDWTF material...

[Greyfox]

Hmm. I didn't put a whole lot of effort into mine. I roll a TCP socket server that pretends to be a HTTP daemon and returns a KML file to Google Earth through a network link. Initially it was set up to read a data file with an ephemeris table for a satellite in it and plot a chunk of the satellite's orbit and its current location as a placemark. I'd just feed two or three interesting satellites worth of interesting data into it, running each one on a separate port.

Lately I've been using the same coordinates library to generate static KML files of GPS tracks in Google Earth. These files don't have the same realtime interest that the satellite tracking one does, since the GPS tracks are being gathered earlier and being batched processed. I have several C++ libraries up on Github, [github.com] if you're curious.

Thirty years ago...

[ctrl-alt-canc]

,,,I tracked NOAA and OSCAR satellites using a polar Earth map upon which I superimposed a orbit-shaped transparent ruler. I then used NASA data, Kepler's equations and a desk calculator to compute the right ascension of the orbits that crossed my reception zone, and I set the ruler accordingly. While I was receiving satellite data I had to read on the ruler the satellite position and move my antennas, while another ruler gave me the amount of Doppler correction. Doing all these tasks alltogether was messy, but it was one of the funniest and more rewarding things I ever did when I was involved with ham radio. I don't think that if I had a computer to automate everything, I would have enjoyed this that much...

Re:Thirty years ago...

[nojayuk]

I worked with the OSCAR Euro command people for a while in the early 80s at the University of Surrey. We had an ex-Navy AA radar X-Y director head on the roof of the building with some Yagis strapped to it to track the satellites and a microcomputer to do the tracking for us as well as sending control commands to the OSCARs as they went by.

The guy in charge was someone you might have heard of, a Ph. D. student name of Martin Sweeting who was busy designing and building cheap satellites; his thesis project was something called UOSAT-1.