Top
Best
New

Posted by zdw 10/26/2024

Debugging my wife's alarm clock(ntietz.com)
137 points | 81 commentspage 2
FridayoLeary 10/28/2024|
I had no idea that digital clocks time themselves by by measuring the mains frequency. And that this is an intended, albeit very minor, function of the power grid! It's amazing how things that provide only a tiny benefit to any individual but has an effect on such a massive scale that people actually sit down and spend time making it work.
rob74 10/28/2024||
"Over a decade old" is technically true, but probably an understatement, to me it looks like something from the 1980s or maybe 90s? Apparently it's old enough to not have "planned obsolescence" designed in...
xnorswap 10/28/2024||
You can see a "11/99 T1" sticker on the case, which might indicate a batch date of Nov 1999.

The printing style on the sticker on the case feels later than 1980s too.

rob74 10/28/2024||
Ok, then maybe it's a model designed in the 80s/early 90s that was still being built in 1999?
andai 10/27/2024||
Fascinating. I thought mains frequency was variable, which is why mains hum can be used to date recordings? It seems that there's some drift (why?) but they try to keep the average constant, over some window?

The article mentions the device prefers to use the mains frequency for timekeeping, using the crystal oscillator only as a fallback. I found this surprising, but I don't know much about either of those things.

shultays 10/28/2024|
I am no expert but from what I read & remember: it can have fluctuations and lagging behind or leading "real" time but it gets corrected and synchronized back later