I suspect size won't be the issue, it'll be cost and stability in such a variable environment as the wrist. And of course, power issues.
And lack of interest in making this a general-use consumer device. Especially given that the parts are exotic, like the 200 GHz signal generator. I can't see this becoming mature enough in a 10 year time frame, to consider it for a watch movement at a price anyone outside the top 1% would consider. Figure: if the Citizen 8.4 MHz pans out, and it really does achieve 1 SPY...is there any real value added with a movement that's 10 or 100 times more accurate?
I'm not even sure it'll have broad appeal here. Obviously, the price is gonna be a major player. If it's $10K, still want one?
One thing from that article was the development of the very low-voltage signal generator. I find that very interesting. Integrate that into a simpler atomic clock system and you can probably make this much cheaper with no practical loss of accuracy.
Fascinating development, thanks for the link. I can certainly see the potential applications for many devices but the stated 66 milliwatts suggests it would only be able to run for a few hours on a button cell - so unless they can drastically improve that with much lower powered designs, I doubt that it will be viable for anything on the wrist other than rechargeable smartwatches.
The article's misleading. GPS, WiFi, and screen are all on intermittently; the signal generator would have to be on constantly. A 2032, according to
is rated at ~ 225 mAh at 3 volts, or about 600 mWh usable (before the voltage drops too much). To get a week of use, you'd have to drop the power requirement down to about 3 milliwatts. And I doubt that a solar cell would be close to enough; the recharge wouldn't be *close* to high enough, plus you'd need enough capacity to run the 16 or so hours when it's not in the sun. (Room lighting would do nothing, not with this kind of power drain.)
And I doubt even a rechargeable smart watch would cut it. Again, the killer is the continuous power drain.
Same problem with the CSAC watch: large power usage.
I suppose one possible approach could be to just power the device up for a very short time - periodically and/or perhaps whenever the temperature changes - and use it as a reference to calibrate and compensate the drift of a conventional quartz oscillator.
That may potentially get round the power issue and still provide exceptional accuracy in a battery powered device - using 'Molecular Clock Reference Compensation' instead of direct thermal compensation. Maybe...
The once-per-cycle approach allowed Longines to employ a more power-hungry, higher frequency auxiliary oscillator than Seiko and thus a more accurate and useful reference for calculating corrections.
In both Longines' and Seiko's cases the auxiliary oscillator served to identify variation in temperature by way of measuring the difference against another oscillator of a different temperature/frequency profile. The IC's frequency count against the main oscillator was then adjusted according to a preset calculation.
Now, my dual oscillator VHP has consistently been one of my top performers, coming in at under 3 SPY year in, year out. There is probably a good reason, however, why the industry moved away from dual oscillators and towards thermistors. Besides providing an even more dependable reference frequency, what value would running this new technology as an occasional auxiliary oscillator have over, say, a more finely tuned temperature/frequency reference table and an increased temperature sampling rate (à la Citizen's upcoming 1 SPY cal. 0100)?
Ask me about 4 MHz watches
Firing up only once per inhibition circle the auxiliary 262 kHz oscillator... well, that technically just makes no sense, in my opinion.
We would not get that excellent accuracy-performance from the dual-quartz VHP if it would sample the temperature only once in every 8 minutes (the inhibition circle for that particular watch)! It monitors the temperature full-time and digitally corrects the watch once in every 8 minutes. Temperature monitoring and digital correction are two different actions!
Last edited by ppaulusz; July 23rd, 2018 at 11:20.
There are currently 1 users browsing this thread. (0 members and 1 guests)