weemba@garnet.berkeley.edu (Obnoxious Math Grad Student) (07/13/88)
In article <1222@thumper.bellcore.com>, karn@thumper (Phil R. Karn) writes: >Good point. But how many applications really require six second response >time? Voyager seems to have been highly successful without humans on >board, despite round trip times measured in hours. Yes, it has been highly successful. But not without some extreme diffi- culties. [The following is based on the 10/86 S&T account.] The first major nasty for Voyager 2 was the radio receiver short-circuited eight months after launch. The backup proved incapable of changing fre- quencies. All this time mission control has had to estimate, within 100 Hz, the effective receiving frequency (the instrument is very sensitive to temperature). This is not easy, and a misguess as V2 nears Neptune could prevent any essential last-minute corrections from taking place. The second major nasty for V2 was the gearing that controlled the optical istruments went berserk and then jammed to a halt shortly after reaching the far side of Saturn. Numerous photo-opportunies were lost. The prob- lem took several years for JPL to diagnose. It was rather fortunate that simply heating and cooling got the gears unjammed. For once, the long time scales proved a boon. The third major nasty for V2 was related to the reprogramming of the error correcting code from a wasteful Golay code to a bit-spartan Reed-Solomon code. Six days before U-day the JPL monitor pictures began to go blooey. Two days were wasted trying to find the problem in JPL software, before it was generally realized that it was V2. A bit-by-bit check showed that a single 0 had flipped to 1, and it couldn't flip back. A JPL hacker figured out--overnight--how to program around this. There were other problems, and only a combination of good luck and ex- cellent talent has made V2 a success. I've got my fingers crossed for Neptune. ucbvax!garnet!weemba Matthew P Wiener/Brahms Gang/Berkeley CA 94720