COLE%USC-ISIB@sri-unix.UUCP (10/22/83)
From: Randy Cole <COLE@USC-ISIB> I respectively beg to differ about the possibility of buying a $100 WWVB receiver and trying to extract the time codes from it. We at ISI bought 4 WWVB clocks (receiver, 8085 micro, and antenna) from an outfit called Spectracomm for about $2500 apiece. You can now buy cheaper ones, maybe $2000 or so. The Spectracomm clocks have the reputation for being the best around. We use them for absolute time measurements for network transit times, etc. The problem is that there is an incredible amount of noise at the 60 KHz frequency of WWVB, from things like switching power supplies, etc. The places where you want the time are generally around computer rooms, which are the noisest places on earth. We have an outdoor active antenna on top of ISI (200 feet up), so I don't know how you could do much better than that. You can watch the clocks synch up and on some days it takes 5 minutes or so, because about every 3rd or 4th bit is noisy. I'm not saying you can't do it, but if you were doing it for a research project where people time is costly, it would be false economy, because it would take a fair amount of experimentation. Also the electronics in LF receivers exhibits a built-in delay, often on the order of 20-30 microseconds. Therefore you need some calibration with an absolute, like an atomic clock. We even got the Naval Observatory people to bring the official USA portable atomic clock around, which was really neat. We spent a lot of time getting accurate time, and it was kinda fun, but it wasn't easy. Randy Cole