alt Hacker NewsI can't tell if I'm fooling myself or if this actually (partially) worked. Is it possible for only the hour and seconds to sync if the signal isn't perfectly received? How long is it supposed to take to sync the time? Is it possible that this can confuse clocks that get a strong enough real signal while the phone is running this nearby?
I just tried it on a clock that has only ever successfully synced once many years ago, and it's still in the same bad location that never seemed to get a strong enough signal.
Its crappy little LCD animation did indeed seem to dance in sync with when I would turn the signal on and off on my phone. It took a few minutes of trying but then suddenly the hour and seconds updated to the exact time. Had to set the minutes manually :/
It takes a whole minute for the entire data frame to be transmitted. While 1 bps is great for propagation, WWVB (and it's kin) is a really weak signal to pickup using the tiny antennas these kinds of clocks have. There's a relationship between received power, modulation type, data rate, error correction, and error rate so changing one will change others. Thus it can sometimes take at least several minutes for all of the timing data to be gathered as there is no error correction included in the data frame. Also the HF spectrum is incredibly noisy as switch-mode power supplies (pretty much in everything these days) produce a ton of interference so taking a battery powered clock outside can help greatly.
WWVB clocks are great but aren't so good these days. NTP is pretty much as good as anyone will ever need in their home but this has the downside of usually requiring internet access. GPS clocks have been the standard for 30+ years for anyone needing precision timing.