WatchUSeek Watch Forums banner
1 - 17 of 17 Posts

Enoslives5

· Registered
Joined
·
224 Posts
Discussion starter · #1 ·
I love my GW9100. It's a keeper for life: perfect size for me, love the "revolver" look, fits every look (can even be stretched to work with a suit if necessary).

Unfortunately it doesn't sync every night because I live in downtown Boston, which i guess is too far from Ft Collins to be reliable at syncing. It only ends up syncing when I take it traveling.

I find that this watch very consistently runs a couple second fast over the course of a week. So the fast time is not noticable over a months, but if it doesn't sync for a couple months then it starts to drift more meaningfully.

The consistency with which the watch is fast is quite remarkable, and I'd venture a guess that when move quartz watches are inaccurate at all, they are also consistently inaccurate in the same way (reliable always a couple seconds slow or fast over a week).

I think it would be cool if Casio launched a watch that using atomic syncing to "correct" a quartz watch error even when the sync fails. When it syncs, my casio knows the time of my last sync, and the amount of seconds that drifted off of atomic time. By a simple equation (current atomic time - displayed time ) / days since last sync, Casio could create an "offset" that should be applied to the time every night the watch fails to sync. That way, even when syncing doesn't work, the watch will always be exactly correct.

Has anyone ever heard of a watch that does this? If not, I hope Casio incorporates something like this in a future high-end model. For people with unreliable sync locations, I think it could be a killer feature for accuracy purposes.
 
I don't think this is possible because of the main reasons a watch looses/gains is temperature so your watch will gain at a different rate then mine because I live in San Diego. One thing they could do is make the module thermocompensated and it wouldn't only loose 5-10 seconds a year and it could use the atomic function to correct itself.

I am surprised it doesn't sync it Boston it should.
 
Discussion starter · #4 ·
I get syncs in the suburbs but not in downtown unfortunately. I live in a little brick box of an apt.

That fact that our two watches would lose (or gain) different amounts is fine with my methodology. The calculation would be in the software of the watch and would be performed on each successful sync, so that way it would be the most up-to-date "offset" for the watch and would be personalized to each watch.

Given the functions of these advanced watches, I think it would be a very simple calculation for these watches to perform. It's just a little bit more software that Casio would have to write for future Atomic models.
 
It's interesting but I doubt any watch will ever do it because in order for it to work the watch would need to make corrections of fractions of a second everyday it doesn't sync and you have to assume the loss or gain will always be consistent and the watch would then need to make calculations.

But if it worked it def would be cool. Or G-shock can just make a satellite corrected model instead of an atomic model.

The seiko satellite gets reception pretty muh anywhere on the globe.
 
I have a Gw-8900 that gains about 16 seconds per day, yes, per day. I am in Denver which is an hour from Ft. Collins. Sometimes I have to go upstairs to sync it but I don't think there has ever been a day that it has never synced.

I tried opening the watch to find the trimmer like the guide mentions but this watch does not have one visible.
 
Just to add to what Fer said, it's not just temperature but also pressure changes that affects watch accuracy. Also, the direction and degree of that change will vary. Thus, programming atomic sync to make a 'best guess' adjustment when it doesn't have a signal can't be done reliably. *Maybe*, if it kept a running average on time drift with each correction, it might be able to apply such an average, but again... it's not reliable guessing.

AFAIK, all CASIO G-Shock watches have a trimmer adjustment screw. I don't own a GW-8900, but I'll bet anything it has one.

I own a GW-9100 that was drifting about +30 sec/mo without atomic sync. Even though I get a good signal, I decided to try adjusting it a bit. Now it gains about 8 sec/mo. An important factor to time adjusting is to do it at the "right time". I've done mine at the end of the warmer months, noted that a couple of watches were showing very accurate but then began to lose time once the cooler months started. It's probably best to do the adjustment in the cooler months and then make only very minute adjustments if the time gain is a bit too much in the warmer months. You simply can't achieve +/-1 sec/mo accuracy and expect it to last, due to changing temperature, pressure, and inherent drift (I had one 20 year old G-Shock gaining nearly 50 seconds per month once).

You should get a good signal in Boston. It may just be your apartment is poorly situated for it (meaning there are many obstacles nearby interfering with the signal). Do you have a westward facing window? Try out a number of different positions. You can get somewhat of an idea if you try manual sync and look at the signal strength readouts.
 
I had no idea any of those affected the accuracy. I guess I never thought about the why enough to inquire. That's pretty interesting.
Even the thermocompensated quarts modules need to be worn consistently to achieve desired accuracy. My chronomaster which is +/-5 sec per year needs to be worn 12 hrs a day to be under spec, haha.
 
Found on Wikipedia. "It is possible for a computerized high accuracy quartz movement to measure its temperature, and adjust for that as well. Both analog and digital temperature compensation have been used in high-end quartz watches. In more expensive high end quartz watches, thermal compensation can be implemented by varying the number of cycles to inhibit depending on the output from a temperature sensor. The COSC average daily rate standard for officially certified COSC quartz chronometers is ± 25.55 seconds per year. Thermo compensated quartz movements, even in wrist watches, can be accurate to within ± 5 to ± 25 seconds per year and can be used as marine chronometers to determine longitude by means of celestial navigation."
 
Atomic sync is effective when there's a good signal and the watch is kept still. Since a human being must sleep at some point during a 24 hour period, atomic sync can be reliably done without much effort. But again, if the environment doesn't provide a good signal it's haphazard. That's why CASIO has been pursuing the Bluetooth technology to interface with an iPhone and synchronize this way. It's more reliable and takes less time. And then you don't need all of that thermal compensation technology that is expensive and not as effective. But you've got to admit, if you've got the bucks, having that thermal compensation functionality is pretty cool. :-!
 
Just to add to what Fer said, it's not just temperature but also pressure changes that affects watch accuracy. Also, the direction and degree of that change will vary.
While technically accurate, the quartz module in your watch is unaffected by changes in atmospheric pressures. That pressure is far too small to have an effect.
If atmospheric pressures did have an effect, then even thermo compensated watches would still be off.

Statistically, 20 seconds a year is within the error margin for thermally compensating a quartz oscillator.

Now atmospheric pressures DO have a measurable effect on electromagnetic radiation (the timing signal from Ft. Collins), mostly in how those changes affect the ionosphere.

Edit : I got interested, and if a thermally compensated oscillator is off by 20 seconds per year (and some have tighter tolerances than that) :

60 sec/min * 60 min/hr * 24 hrs/day * 365 days/year = 2628000 sec/year

20 sec / 2628000 sec = 7.6*E-6 or 0.0000076 or 0.00076% off per year.
 
1 - 17 of 17 Posts