Comment by blattimwind
7 years ago
> So what else could it be? Well, at the heart of every electronic device is a clock. Traditionally, these are quartz oscillators, crystals that vibrate at a specific predictable frequency—generally 32 kHz.
Watch crystals run at 32[.768] kHz because you can divide with a binary counter by 2^16 and get 1 pps to drive the Lavet stepper driving the seconds hand.
Watch crystals are also commonly used by MCUs for their RTC (real time, very low power clock), but never used to produce the main clock of a SoC or something like that. Mostly because that'd need an insanely high multiplication through a PLL (higher frequency multiplication ~ higher phase noise). Base clock crystals are typically 20-50 MHz.
Quartz crystals used to be hermetically sealed.
> Quartz crystals used to be hermetically sealed.
Are they sealed against helium though? It can get through a lot of materials that more common gases can't.
Traditional crystals come in what's referred to as a "hermetic metal can", i.e. a metal box soldered closed with leads going through glass seals. I don't know if that's good enough for Helium.
The sealing is mostly because of humidity. I wouldn't expect quartz crytals to be overly sensitive to some gas, since they're basically tuning forks; the crystal physically vibrates in a resonance mode caused by and inducing an electric current across the crystal.
Hermetically sealed by definition is airtight, i.e. "excludes the passage of air, oxygen, or other gases". [1] With that said, the diffusion rate depends on both the properties of both the seal and the gas.
[1] https://en.wikipedia.org/wiki/Hermetic_seal
Sometimes, definitions aren't the most mechanically sound constructs available.
1 reply →
Very few air seals protect against helium, the smallest gas. Most materials, even metals, are like sponges to helium. It'll seep through slowly.
21 replies →
Well, air is also compound of rare gases of which Helium is one.
I agree it is unlikely to be the RTC clock at issue unless it runs a watchdog of done form. But given that they are using MEMS for the 32kHz clock it seems likely they are using it for other clocks too.
> I agree it is unlikely to be the RTC clock at issue unless it runs a watchdog of done form.
Fairly standard technique to run the watchdog off the RTC clock, because that might still work if the main clock is wonky.
Indeed, and historically, a slower RTC clock was also preferred because it would draw a lot less current during sleep mode. And it allows the rest of the system to throttle the main system clock up and down for power savings.
I said "historically" because I don't know is if these are still important issues on a modern system relative to all of the other things that are probably using the main system clock during sleep mode, and the more generous battery.
2 replies →
Nit-pick - you mean to divide 32.768kHz by 2^15, not 2^16 to get 1 pps.
More nit-picking... according to wikipedia Lavet requires two states for one step, so 2^16 looks good based on this.
Damn... I mean 2^14 looks good. :)
2 replies →