Back in Nov 2019, I was chatting with some friends about the old Ga Tech W4AQL Satellite Tracking system from the 1980s and how long it took to load the program from cassette tape. I foolishly remarked, "I could build the whole system today in something the size of a bar of soap with nearly instant software loading time". For one year I have been able to avoid this challenge, because we couldn't find any copies of the firmware. I no longer have that excuse. Let the fun begin.
My initial enthusiasm upon discovering the stash of old units has been dampened by a sudden and unfortunate memory. While I'm not 100% sure, I think the programs were distributed in a compiled format (there were two programs, one for satellites, and another one just for terrestrial antenna pointing based on the remote station's radio callsign prefix). I don't think this was done for copy-prevention reasons, but rather program speed. My recollection is vague, but I think we discovered that the compiled program ran faster and/or loaded faster. With normal languages, that makes sense, but I'm not sure how Basic was compiled.
Regardless, I will proceed and try to get the floppy images extracted to a more modern media for analysis. And continue racking my brain and see if I can come up with any place to look where the source code might be stored.
Seems like a Raspi or a Beaglebone running GPredict would satisfy the size and loading requirements. Depending on your rotator interface, the rest might be as simple as an RS232 level shifter.
Yes. However my plan was to reproduce as faithfully as possible the original. The stumbling block has been the original software, but I’m 90% sure I’ve tracked that down (pun intended). It’s being held for me and hopefully once this Covid dies down I can arrange to pick it up.
The original rotor interface was simple and dumb — open loop control using relays. I’m pretty sure I can squeeze the necessary relays into the space. And because of the modern and lower cost electronics, I’m pretty sure I could upgrade the system to accept feedback. In fact, I have a vague recollection that feedback was in the software but just not used.
I have not messed with the old school rotator control stuff. I remember from the 1970s, it was some wierd, clicky, electromechanical thing. I just used a variant of the EZCom protocol on a serial interface for the rotators that I have built.
Seems like a Raspi or a Beaglebone running GPredict would satisfy the size and loading requirements. Depending on your rotator interface, the rest might be as simple as an RS232 level shifter.