[russr1123]

So I just got a skeleton watch that has an ST17, handwind movement in it: http://www.tjseagull.com/ArticleShow.asp?ArticleID=112

I've been testing the accuracy of the watch over the past week and I'm finding that it is about +20-30 sec/day. Doesn't this seem a bit high?

Also, the movement has a 40+ hours power reserve. Is there a correlation between accuracy and the amount of power left in the main spring? Meaning, is the watch more accurate at full-wind, half, near empty?

Edit: these results were taken with the watch being wound close to poet reserve depletion. I will post results below with it being wound every day.

 

[Pawl_Buster]

So I just got a skeleton watch that has an ST17, handwind movement in it: ????

I've been testing the accuracy of the watch over the past week and I'm finding that it is about +20-30 sec/day. Doesn't this seem a bit high?

Also, the movement has a 40+ hours power reserve. Is there a correlation between accuracy and the amount of power left in the main spring? Meaning, is the watch more accurate at full-wind, half, near empty?

Is it a Sea Gull branded watch?

Yes, the fuller the wind; the more accurate a watch should be.

 

[russr1123]

No, it's not Sea-Gull branded, but it has the Sea-Gull ST17 movement. It isn't a cheap brand off ebay though. Is +20-30 sec/day normal?

 

[Chascomm]

No, it's not Sea-Gull branded, but it has the Sea-Gull ST17 movement. It isn't a cheap brand off ebay though. Is +20-30 sec/day normal?

That would be within manufacturer's specification (+/-30s/day) however that represents the outside limit of what is considered acceptable accuracy.

[edit: actually I'm not sure if the manufacturer's specification is for +/-30 or +/-45. Does anybody know?]

The ST17 was designed originally as an automatic. As such it would be operating with near to full reserve in normal use. As a hand-winder, it will be running down through the day, resulting in torque variations and consequent reduction in accuracy. To get best performance it is recommended to wind the watch fully each morning. Due to some moderately complicated physics, this will give better results than winding it at the end of the day.

One question though; does your watch wind fully to a point where winding is not possible? Or does it wind on indefinitely (like an automatic would?)

 

[russr1123]

There are several configurations of the ST17 movement ranging from automatic with date complications, down to a hand wind with no date. I would hope the hand wind configuration would have been tweaked to run more accurately knowing that there's no rotor to auto wind it.

Anyway, to answer your question, the crown never stops winding, but since it is a movement made for skeleton watches, I can see when the main spring is fully wound.

 

[russr1123]

I should also note that I have yet to wear my watch, so essentially all my accuracy tests are with it sitting face up. I've been keeping the main spring more wound up to see if it improves accuracy.

 

[Chascomm]

I should also note that I have yet to wear my watch, so essentially all my accuracy tests are with it sitting face up. I've been keeping the main spring more wound up to see if it improves accuracy.

It could be argued that given the watch's function as a wearable timepiece, the only valid accuracy test is on the wrist. On the other hand, I would expect it to run slower off the wrist than on.

It is more likely that the watch will run faster with the mainspring less wound so it will be interesting to see the results of your next measurements.

 

[russr1123]

I hope it does not run faster while on the wrist because it is already running fast. I did a night wind and found that 24 hours later it had gained 9 seconds. To switch to a morning wind, I let the main spring unwind overnight and found that it had gained 6 seconds overnight. So it seems like the watch does indeed run faster as the main spring unwinds.

So let's see how it does with a morning wind.

 

[CheapThrills]

This topic is relatively close to my own gueation.

Considering accuracy between movements of the same type (mine is ST2130), should they have similar inaccuracies between different positions?

For example, is it possible, that some movement type gains/loses time in certain position (because of gracity) or are all these features just personalities of a certain individual watch?

 

[russr1123]

It depends on the manufacturing tolerances as well as the quality control. If both are good, then the positional variance should be similar between watches.

This topic is relatively close to my own gueation.

Considering accuracy between movements of the same type (mine is ST2130), should they have similar inaccuracies between different positions?

For example, is it possible, that some movement type gains/loses time in certain position (because of gracity) or are all these features just personalities of a certain individual watch?

 

[CheapThrills]

It depends on the manufacturing tolerances as well as the quality control. If both are good, then the positional variance should be similar between watches.

Hmm.... Thanks for the answer!

I was just thinking, because I was reading from here some tests : Review: Sea-Gull 816.362 / ST2130 | Watch Guy
My 816.362 seems to advance about 1-2secs a night even, when it is DialDown.
Furhermore, it seems to advance a second or a bit more in 8-10hours even, when crown down...

I'm a bit confused, since it seems to advance 2-5secs a day pretty much regardless of the position.

 

[tony1951]

I let the main spring unwind overnight and found that it had gained 6 seconds overnight. So it seems like the watch does indeed run faster as the main spring unwinds.

So let's see how it does with a morning wind.

This is a very common feature of watches near the end of their power reserve. I have seen it explained as being caused by the fact that a weakly wound watch is likely to have lower amplitude on the balance swing, because less energy is imparted to the balance by the pallet. This means that each cycle of the balance completes a little more quickly than if it had a larger swing.

I have a VERY CHEAP Chinese auto Tongji watch that has been regulated so that it loses just abouttwo seconds a day when fully wound by hand. When it gets a little slow (say about 5 seconds) I stop winding it and let the rather feeble auto winder wind the watch. Because the auto wind mechanism is not efficient, it never gets more than half wound, and then the watch gains about five seconds in a day. To keep this watch within about five seconds either side of the correct time, I simply wind it for two days (it then goes a little slow) and then I let it run for another day on auto-wind when it catches up and passes the correct time by a small margin. This suits me fine.

By the way, the really important thing about a watch is that it gains or loses the same amount each day. A watch that gains exactly twenty seconds EVERY day is stable. So is one which gains exactly one second. If it is consistent, any watch can be regulated to stay very near the exact time. One which is twenty seconds fast one day and ten seconds slow the next is not so good.

 

[lysanderxiii]

Hmm.... Thanks for the answer!

I was just thinking, because I was reading from here some tests : Review: Sea-Gull 816.362 / ST2130 | Watch Guy
My 816.362 seems to advance about 1-2secs a night even, when it is DialDown.
Furhermore, it seems to advance a second or a bit more in 8-10hours even, when crown down...

I'm a bit confused, since it seems to advance 2-5secs a day pretty much regardless of the position.

That is good. there will always be a variation between positions, however, the better the movement adjustment the smaller the variation.

If your watch runs between 2 and 5 seconds error no matter the position, you have a very good movement...

 

[CheapThrills]

That is good. there will always be a variation between positions, however, the better the movement adjustment the smaller the variation.

If your watch runs between 2 and 5 seconds error no matter the position, you have a very good movement...

In fact, I have kept it now with better wind.
I wind it manually since I at week I can keep it in my hand only when not at work.
For night and work time, I have left it DU. Now after 48 hours, seems like it has only gained 4 secs. I think it is quite little.

Strange, that it does not seem to lose time in any position, only gains a bit more or less.

 

[russr1123]

I thought I would post an update on my watch which uses the ST17 movement. I originally thought that the accuracy varied based on how much power was left in the main spring. I was wrong, it does not. The only time accuracy is affected by the main spring is when it is almost completely out of juice and you can see the fly wheel rotating slower than normal.

My initial measurements were all face up and were all with never having worn the watch. It was about +30sec/day. I was so disappointed that I exchanged it for a replacement only to find that the replacement behaved exactly the same.

Once I started wearing the watch, I found the accuracy to be around +15/day. Most of that time is gained during the 12-13 hours it is off my wrist. Wearing it regularly seems to average out the positional variances and makes it much more accurate. I experimented with face up and crown down off the wrist, but it did not improve my off the wrist results.

 

[tony1951]

@russr1123


You might find that in one position the watch will run slower or even lose a little time. You can try it in six positions obviously: Face up, face down, crown up, crown down, crown right, crown left.

If you test it for a few hours in each position, noting how it keeps time against a quartz watch, you may find that you can leave it in a particular position at night so that it keeps more or less perfectly on time. It is worth the effort of making these observations. At the least, it will tell you which position NOT to leave it in at night.

Good luck

 

[Dan83bz]

my Rodina "Bauhaus" with the st17 auto is VERY accurate, more accurate than most of my watches, I've been wearing it for about three weeks now and the average rate is about 10-11 seconds per WEEK!!

I wear it 10-14 hours/day and then leave it on the desk facing up.

 

[Der Amf]

my Rodina "Bauhaus" with the st17 auto is VERY accurate, more accurate than most of my watches, I've been wearing it for about three weeks now and the average rate is about 10-11 seconds per WEEK!!

I wear it 10-14 hours/day and then leave it on the desk facing up.

I've had mine 2 days. I set it 3 seconds fast and it is still 3 seconds fast. Obviously I'm delighted. I notice that during the day on my wrist it lost a second or two but regained them overnight. I guess a sample of two days isn't much to go on, but it's an excellent start

 

[Dan83bz]

I've had mine 2 days. I set it 3 seconds fast and it is still 3 seconds fast. Obviously I'm delighted. I notice that during the day on my wrist it lost a second or two but regained them overnight. I guess a sample of two days isn't much to go on, but it's an excellent start

Let's see how they hold up over longer periods of time. Unless I happen to land a few too many unexpected "squids" anytime soon and lead me to get the real deal, this Chinese babe isn't staying too far from my wrist.

 

[Dan83bz]

Let's see how they hold up over longer periods of time. Unless I happen to land a few too many unexpected "squids" anytime soon and lead me to get the real deal, this Chinese babe isn't staying too far from my wrist.

Now that 1 year has past and this is still with me, I can come back with a follow-up

The watch still works first of all :-d

I have not worn i too much, maybe 1 month overall thru the last year. Because....you guessed it, too many watches, too few wrists.

The last time I checked its accuracy it was still pretty darn good, still under 10s/day. |>