428
Voyager 1 loses contact with NASA, turns on retro transmitter not used since 1981
(www.livescience.com)
Share & discuss informative content on: Astrophysics, Cosmology, Space Exploration, Planetary Science and Astrobiology.
The Busy Center of the Lagoon Nebula
π Science
π Engineering
π Art and Photography
Other Cool Links
I'm picturing the Voyager 1 terminal is an ancient computer from the 1970s hooked up to a large parabolic antenna, and everyone is afraid to upgrade it because they might mess something up. I'm sure that's not the case, but its what lives in my mind.
Since I was thinking about it I looked up some stuff: "So Voyager-1 does not βreallyβ have a computer, in the sense that it does not have an operating system or RAM or a microprocessor. It was built in the 60s before any of this was invented and used CMOS-based microcontroller chips from Texas Instruments. Overall, it has a 16-bit processor and a MASSIVE memory of 70 KILOBYTES. That is smaller memory than a thumbnail of a phone image today, but it was enough to send images through which we discovered Jupiter has rings and much more."
From: https://medium.com/towards-generative-ai/voyager-1-what-computer-system-it-has-that-is-still-running-strong-a269aaea316b
Although it doesn't compare to modern systems, the computer systems on Voyager is a computer by all means. It's even the longest running computer that ever existed, having never been shut off. It runs Fortran code.
The image data that the camera made didnn't have to fit in the computers memory. It was written directly to tape, which was then transmitted by the computer. The resolution is 800x800 pixels with only one colour at a time. The colour images or in larger resolutions were combinations of several images. The camera has been shut off by now.
Speaking of not wanting to touch the code, it did have an issue last year, where the code seemed to have stopped or gone into a loop for unknown reasons making it inaccessible for the operators on Earth. Thankfully another part of the computer was instructed to periodically overwrite the main code, so it managed to correct the error by itself. At least that's what I remember reading.
You can cram a load of machine language into 70K. Seems far more than needed, bet a bunch is for redundancy.
It's probably not too far off. The ground station probably uses the same antenna, the computer running it is probably relatively new, but I'm sure there's some kind of emulation for the control software. Like a Fortran emulator, not like WINE or an old DOS VM.