News for hackers. Stuff that matters.
Re:Casio-F-91W die-shot by sn_master on Sunday January 01, 2023 @03:58PM (#34210795)
> all the technological advances that we have

This is 34 years old technology. You can get GPS watches relatively cheap and they'd sync with satellites automatically and always remain accurate.

Re:Casio-F-91W die-shot by vasqw on Sunday January 01, 2023 @04:23PM (#34211073)
Yes. I ended up buying a GWM5610 which synchronises through radio. It's cool. But still I would have hoped there would be a better "local" (i.e. self-reliant) solution.
Re:Casio-F-91W die-shot by myself248 on Sunday January 01, 2023 @06:08PM (#34212089)
There's software for the Beaglebone PRU so you can generate your own WWV signal, for receivers buried too deep into buildings or otherwise suffering from poor signal.

Of course, with any HackRF you can also generate GPS signals.

But it would be a lot less necessary if the watches had reasonable temperature compensation. Obviously the crystal frequency shifts with temperature, and they're typically adjusted to achieve best accuracy when at "wrist temperature" for 16 hours a day, and "nightstand temperature" for 8. Any deviation from that, routine, and crystal accuracy gets terrible.

Temperature-compensated crystal oscillators (TCXOs) switch tiny capacitors in and out of (or bias a varicap in) the crystal loading circuit to "rockbend" its response and keep it nearly flat across a wider range of temperature, but this requires extra power. A minuscule amount, but it adds up quickly when your budget is nanowatts.

However, there's another way to do it, which I'm astonished not to see: You don't need to alter the crystal's actual frequency if you can just adjust the number of counts that represent a second. Every once in a while (every few minutes would be plenty), take a temperature reading, look it up in your calibration table, and stuff or steal a few cycles from the counter register.

This takes basically no additional power, adds one more calibration step at the factory, and should bring drift down by an order of magnitude, into TCXO territory. Seconds per year, not seconds per month. That would be something to crow about, and it baffles me that Casio and others never went that route.

Re:Casio-F-91W die-shot by erosenbe0 on Monday January 02, 2023 @02:38AM (#34215370)
Nxp sells a PCF2131 with an internal TCXO that runs on 64 nanoamps when not doing anything else. So with a modern CMOS process, the TCXO itself is not a big deal. A clock pin at 32768 cyc/s looks to be way more power hungry, which tells us that an ideal low-power solution is an all-in-one SoC.
Re:Casio-F-91W die-shot by ashwinne on Sunday January 01, 2023 @11:27PM (#34214361)
Last year I switched to the GWM5610 after a long search. It perfectly fit my needs - accurate all the time (syncs to atomic clock radio signals), never needs a charge (solar powered) and is indestructible. What more could a nerd want!