WatchUSeek Watch Forums banner
1 - 20 of 48 Posts

·
Registered
Joined
·
503 Posts
Discussion Starter · #1 ·
I think this is a candidate for the how to form. If the moderators agree please copy it there.

How to adjust the accuracy of a G-Shock. (maybe other watches too)

Materials - tools to open the G-Shock case, screw driver, or watch back removal tool, stop watch can be a 2nd G in stop watch mode, note pad, pen, a know base time that will not change, internet time or an atomic recently updated, G-Shock to be adjusted.

Step one. First set the gaining / loosing watch, that we are going to call set watch, to a known correct time source, we are going to call base time, that does not gain or loose time. You can use an atomic watch that has updated in the last 12 hours or you can use the internet time at www.time.gov. Either one will work.

Step two. Then you must figure how close the set watch is to the base time. Look at the base time and start the stop watch when you see the seconds change to some thing that ends in 0 such as 08:55.30 seconds. Then look at the set watch and stop the stop watch when you see the set watch seconds change to 5 seconds after the time you started it on the base time. In the example above you would stop the stop watch when the set watch changed to 08:55.35. Then you look at the stop watch and for example if it said 5 sec and 20 hundredths then you subtract 5 seconds and the set watch is 20 hundredths of a second slower than the base time. If your stop watch said 4 sec and 80 hundredths then the set watch is 20 hundredths of a second slower that base time. Write it down. This step is just getting the exact difference between the base time and set watch. (Side note in reality as long as you know the exact difference between the base time and the set watch (it can be 20 minutes 5 seconds and 43 hundredths) you do not need to set them to the same time it just has a bit more math involved.) Some of you might say why not just start the stop watch when base time goes to a number then stop the stop watch when the set watch goes to that same number. If the difference is very small you can not start and stop the stopwatch fast enough to get an accurate reading. I tried to see just pushing the buttons on a stop watch as fast as I can to start and stop it. I can not get less than a 12 to 14 hundredths of a second to show up. Then there is looking from base time to set watch ect. Just add a few whole seconds and then subtract them at the end.

Step three. Then you wait 24 hours, 23.5 is OK and so is 24.5 just do it at about the same time each day. I do it within +- 30 minutes of the same time each day. (Side note set the alarm on the G you are wearing to remind you.) You find the exact gain or lose by using a stop watch and starting it when the base time goes to 08:55.30 and stop it when the set watch goes to 08:55.35. Then look at your time if it is say 5 second and 90 hundredths your set watch has lost 70 hundredths of second. If it said 4 seconds and 10 hundredths then your set watch gained 70 hundredths of a second. What you are doing is subtracting the 5 seconds and 20 hundredths that you got from step 2 when you determined the exact difference between the set watch and the base time. That is why it is best to use the same number of seconds delay when checking the difference between base time and the set watch each time. In our example 5 seconds, the 20 hundredths is making the measurements more accurate. You do not need to find the time difference to the hundredths but it may take a couple of days to see with the naked eye a one second difference between the set watch and base watch.

Step four. This next step is very important!!!!!. Write down the date, time, and how much the set watch gained or lost. This is very important!!!!!!! If you do not do this you will not know if you are making the set watch more or less accurate after an adjustment.

Step five. Next open the set watch and locate the trimmer screw. It is in the square in the pictures. I have seen it in at least 2 different locations maybe more I can not remember. I am going copying a couple of pictures from Buzzbait, hope you do not mind. LOL





You will notice that it is different from the screws holding the module together. There will be only one and there is no need to remove the module from the case to get to it. I only turn it about 1/32 of a turn, or a very small amount at a time. Write down witch way you turned the screw. I read some where that Clockwise speeds up the watch. Counterclockwise will slow it back down. I do not remember if that is true or not but make an adjustment. If after the next 24 hours the set watch gets worst then adjust in the other direction for the next adjustment.

Next go to step one and set the watch to the base time again.

Step two next. Figure out how far off the set watch is from base time and write it down.

Then go to step three. Wait 24 hours and check if the set watch got more accurate or less by comparing the gain or loss to the gain or loss from the day before. (Side note you may have to wait 2 or 3 days to see if the watch is gaining or loosing time depending on if you are calculating to the hundredths of a second or not. That is OK just make sure you know weather it gained or lost from the last adjustment and how much)

Now go to Step four. Write it down!!!!!!!

Next Step five. Make another small adjustment. You may have to make it in the opposite direction or make it very very small you have to use you judgment here based on the last adjustment and how much it got better or worst.

You continue to do this until you are happy with the watch. I have take a couple that were losing 45 to 60 seconds a month and now they gain or loss is less that 4 seconds a month. I figure if I am going to do this I may as well see how close I can get it. It may take 3, 4 or more weeks depending on how much time you want to spend on this and how accurate you want you watch. I use 15 hundredths as a goal. If the watch is gaining or losing 15 hundredths of a second or less a day then after 30 days it will gain or lose about 4.5 seconds. That is where I set my cut off.

Below is the original post by dudegalea that made me try it.

"look for the trimmer capacitor - it's like a screw head on the circuit board. Should be obvious that it's not an actual screw.
Synchronise the watch to atomic time.
Turn the trimmer cap counter-clockwise just a tiny bit, and see how the accuracy goes over the next day.
If it's better than before, then re-synch to atomic time, and turn the cap a little more counter-clockwise.
If it's worse than before, then re-synch, and turn the cap a little more clockwise.
Each movement of the cap should be really small.
Gradually, it'll get more and more accurate, to the extent that you can't see any noticeable difference in a day. So then you leave it for 3 days at a time, and make even smaller adjustments to the cap."

Hope this helps.

Stan
 

·
Registered
Joined
·
11 Posts
Any confirmation for the clockwise/counterclockwise accelerating/slowing down? I have a dw-5300 which gained 2s a day, now I turned it much clockwise and the end result is that it loses about 2s a day :f, it may be that I broke the thing by applying too much pressure. I'm running a test now, it's near the original position but a bit clockwise. We'll see what happens. Is there any data on temperature change and its effect on the quarz crystal?
 

·
Registered
Joined
·
11 Posts
Now, I turned the potentiometer counterclockwise almost to the original position with good results: from what I can tell ( one day test period) the watch is losing 1/10s a day, so that makes 3 seconds a month. Seems very good but I need to test it longer. Very nice :).
 

·
Registered
Joined
·
15 Posts
my Riseman was about 4.67 seconds/month too fast. While
this was still within Casio's +/- 15s spec, it was much
worse than the 0.5s/month of my G-7700. I read the howto
article about adjusting G-Shocks and decided to try it on my
Riseman, especially since I could put to use some of the
measuring equipment that accumulated beside my desk over the
years.

Those tools are an oscilloscope and - more importantly - a
counter:

I found the clock signal on the third contact pad in the
group above the alti button.




The signal has some intermediate state while rising, so I
set the counter to trigger on the downward slope instead:




This looks like a 32.768 kHz quartz divided by 48. Adjusting
it to a display of 682.666667 Hz would be a bad idea
however, since the counter itself is probably out of
calibration and the load of the probe might also be detuning
the G-Shock's oscillator. Instead, I went for the relative
adjustment of -1.87 ppm = 4.67s/month:


...so the actual adjustment made was 1.15 mHz = 1.69 ppm,
leaving the Riseman 0.18 ppm fast, which computes to 0.45
s/month. Now time has to tell if I got the maths right :)
 

Attachments

·
Registered
Joined
·
2,745 Posts
Please update and let us know if you got it correct. When you check it again in a while, also let us know how far off it is.
 

·
Registered
Joined
·
8,622 Posts
it might have been easier to pick an atomic riseman instead, but nevertheless a nice equipment and a nice tutorial ;-)


regards, holger
 

·
Registered
Joined
·
15 Posts
it might have been easier to pick an atomic riseman instead, but nevertheless a nice equipment and a nice tutorial ;-)
I have to admit that this is an atomic one. Let me explain :). I'm making extensive use of the stopwatch and I chose the riseman because the daytime is always visible in stopwatch mode, so I wouldn't have to switch between timekeeping and stopwatch mode twenty times a day as I had to with my G-7700.

The result is that my riseman spends most of its nights in stopwatch mode, which prevents it from syncing automatically. So before the adjustment my riseman was actually less accurate than my g-7700 unless I happened to have it manually synced a couple of days before.

Please update and let us know if you got it correct. When you check it again in a while, also let us know how far off it is.
Will do.
 

·
Registered
Joined
·
6,894 Posts
Wicked. Just plain wicked. Do all Gs have these contact pads?

If so...maybe you should start a side business adjusting Gs :-!.
 

·
Registered
Joined
·
521 Posts
my Riseman was about 4.67 seconds/month too fast. While
this was still within Casio's +/- 15s spec, it was much
worse than the 0.5s/month of my G-7700. I read the howto
article about adjusting G-Shocks and decided to try it on my
Riseman, especially since I could put to use some of the
measuring equipment that accumulated beside my desk over the
years.

Those tools are an oscilloscope and - more importantly - a
counter:

I found the clock signal on the third contact pad in the
group above the alti button.




The signal has some intermediate state while rising, so I
set the counter to trigger on the downward slope instead:




This looks like a 32.768 kHz quartz divided by 48. Adjusting
it to a display of 682.666667 Hz would be a bad idea
however, since the counter itself is probably out of
calibration and the load of the probe might also be detuning
the G-Shock's oscillator. Instead, I went for the relative
adjustment of -1.87 ppm = 4.67s/month:


...so the actual adjustment made was 1.15 mHz = 1.69 ppm,
leaving the Riseman 0.18 ppm fast, which computes to 0.45
s/month. Now time has to tell if I got the maths right :)
Did you actrually turn the trimmer adjustment screw? You didn't mention about the trimmer like Stan did.
 

·
Registered
Joined
·
521 Posts
Good work, Stan. Adjusting the watch's accuracy has been a task that I had to do with all my G-Shock watches. With some amount of research some 7 years ago, I learned how to do it. Now all my Gs won't deviate by a minute a year!
 

·
Registered
Joined
·
521 Posts
Three minutes/month is good enough if it is the amount of gain. Not loss. That is to compensate for low battery condition...and your watch is not even out by a minute a year!
 

·
Registered
Joined
·
521 Posts
It depends on the watch model. My Mudman G-9000 has its trimmer screw turned clockwise to slow it down, and turned the other way to make it faster. My GL-150, G-2500, G-7300 are just the reverse.
 

·
Registered
Joined
·
521 Posts
Everybody wants an accurate timepiece, man. You can do it yourself!
 

·
Registered
Joined
·
15 Posts
Wicked. Just plain wicked. Do all Gs have these contact pads?
From the pictures I've seen online, all G modules seem to have them.

If so...maybe you should start a side business adjusting Gs :-!.
I'm sure that the better watch repair shops have means to do the adjusting as well. The shop that is near me seems to be rather incompetent though. I had a timex there once for a battery change, and it came back seriously scratched and with the beeper no longer working. They managed to squeeze the beeper contact between the battery clasp and the battery. That's when I started to service my watches myself... can't do much worse than that..

Did you actrually turn the trimmer adjustment screw? You didn't mention about the trimmer like Stan did.
Yes --- without turning the trimmer, the frequency displayed on the counter wouldn't have changed. Sorry for not being explicit about that.

[The direction of the adjustment] depends on the watch model. My Mudman G-9000 has its trimmer screw turned clockwise to slow it down, and turned the other way to make it faster. My GL-150, G-2500, G-7300 are just the reverse.
For the record: My riseman was slowed down by turning the trimmer counterclockwise.

regards,
Andreas
 

·
Registered
Joined
·
521 Posts
Hi, dys! Obviously, you know how to apply oscilloscope and a counter for the purpose of adjusting the acuracy of a G-Shock watch and probably any other quartz watch and I believe that the adjustment (the way you did it) took only once for a watch. So how's the watch's accuracy now?
 

·
Registered
Joined
·
15 Posts
Hi, dys! Obviously, you know how to apply oscilloscope and a counter for the purpose of adjusting the acuracy of a G-Shock watch and probably any other quartz watch and I believe that the adjustment (the way you did it) took only once for a watch.
Right --- this is the main difference to Stan's approach. You can observe the effect of turning the trimmer almost instantly using the frequency counter and don't have to wait days to observe it.

So how's the watch's accuracy now?
It didn't sync for 468 hours and is 0.84 s ahead of legal time at the moment. Considering the leap second at the end of december, the watch is actually running 0.16s/month slow. While this is not the expected result of being 0.45s/month fast, it is more accurate than predicted.

regards,
andreas
 
1 - 20 of 48 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top