WatchUSeek Watch Forums banner

21 - 40 of 49 Posts

·
Vendor
Joined
·
27,721 Posts
Generally speaking, isochronism is a physics phenomenon, which for any oscillator such as the balance wheel in a watch movement, can be affected by numerous different factors.

In the field of watchmaking though, the term isochronism seems to have taken on the more specific meaning of how well a movement can keep regular timing as the mainspring winds down.
^^^This.

Last night's exchange led me to reconfirm my own understanding of isochronism, as my understanding was solely related to watchmaking.

Indeed, as far as I can tell, my understanding was and is correct. All mechanical watches, even highly accurate COSC chronometers, will lose amplitude as they wind down, causing them to become less accurate as they lose power.

This is normal, and the phenomenon is generally called isochronism. It's not so much considered a "problem", only an important thing to understand in the context of discussing accuracy, particularly as it relates to how accuracy is affected by the power reserve.

It's interesting, and likely not coincidental that the STP runs within spec only when it's at least at half power, and the COSC testing requires full winding every 24 hours. The explanation I read for that is the chronometers are not expected to run within COSC spec at less than half power.

Unless I completely missed the point, I believe the OP was demonstrating how knowing the number of turns necessary to reach the varying states of power is useful, and again, it would be just as useful in a COSC chronometer. To know that number, someone at some point must have done the complex math our OP did in this thread.


New travel-sized Tapatalk is more discrete, and fits in your pocket without that embarrassing bulge.
 

·
Registered
Joined
·
1,609 Posts
Discussion Starter · #22 ·
Unless I completely missed the point, I believe the OP was demonstrating how knowing the number of turns necessary to reach the varying states of power is useful, and again, it would be just as useful in a COSC chronometer. To know that number, someone at some point must have done the complex math our OP did in this thread.


New travel-sized Tapatalk is more discrete, and fits in your pocket without that embarrassing bulge.
Exactly my point for the original posting.

Thank you.

RFG
 

·
Registered
Joined
·
8,365 Posts
Gread read here. :-!

But on the other hand, it makes me want switch over to quartz, full time, lol. :-d
 

·
Registered
Joined
·
4,268 Posts
Would this explain why that when I first put on an auto watch, which I tend not to wind. The next day their time acccuracy is out by sometimes a lot. But it's by then fully auto wound up. So I reset the watch to the correct time, and the following day it's remained accurate ,within reasonable parameters.

This has always confused me. So is it the fact that by being fully wound, the watch is at the optimum tension to remain accurate.
 

·
Registered
Joined
·
8,365 Posts
Would this explain why that when I first put on an auto watch, which I tend not to wind. The next day their time acccuracy is out by sometimes a lot. But it's by then fully auto wound up. So I reset the watch to the correct time, and the following day it's remained accurate ,within reasonable parameters.

This has always confused me. So is it the fact that by being fully wound, the watch is at the optimum tension to remain accurate.
Yes, correct. :)
 
  • Like
Reactions: catsteeth

·
Registered
Joined
·
1,609 Posts
Discussion Starter · #26 ·
Next up is a Miyota 9015. Be aware there are photos and math involved. I'm a watch geek's watch geek. I took my passion for watches further and learned service and repair.
Some facts first, just like all internal combustion engines work on the same principal, the same is true about movements. It's just gears and springs.

It takes 8 full turns of the mainspring arbor to get a full wind. The auto winder only winds in one direction, counter clockwise when viewed from the dial side.

The stem winding system consists of 5 gears, the stem wheel has 16 teeth, followed by a set of 3 gears with the last gear being a compound gear powering the mainspring gear that has 50 teeth. Here is the math, 18/16=1.125, 13/18=.72, 24/13=1.846, 50/15=3.333. 1.125x1.846x.72x3.333=4.979. 5 full turns of the stem winds the mainspring one time. 8x5=40, 40 turns of the stem gives a full wind.

Stem gear.
901586.jpg

3 gear cluster.
90156.jpg

Mainspring winding gear, upper left.
901582.jpg

The auto winder has 7 gears with compound gears and the click. Here is the math, 46/50=.92, 37/18=2.055, 26/10=2.6, 37/7=5.285, 35/37=.946, 50/8=6.25. .92x2.055x2.6x5.285x.946x6.25=153.6. 153 rotor rotations winds the mainspring one time. 153x8=1224. 1224 rotor rotations gives a full wind.

Photos in order from rotor to mainspring gear.

Rotor.
90151.jpg

90152.jpg

90153.jpg

90154.jpg

90155.jpg

90157.jpg

901582.jpg

Here is the breakdown of stem winding compared with timekeeping. Test subject is NTH Nacken, dial down, 51 degrees lift angle.
This one, https://www.watchuseek.com/f71/nth-miyota-9015-a-4587809.html

1/4 wind, 10 full stem turns,
240 degrees amplitude, +7s.

1/2 wind, 20 full turns,
274 degrees amplitude, +6s.

3/4 wind, 30 full turns,
286 degrees amplitude, +7s.

Full wind, 40 turns,
302 degrees amplitude, +6s.

RFG


 

·
Vendor
Joined
·
27,721 Posts
Just two individual samples of each movement, but it's interesting to see the difference (or lack of it) in timekeeping accuracy at low versus full power for each movement, despite the similar amplitude numbers for both movements.


New travel-sized Tapatalk is more discrete, and fits in your pocket without that embarrassing bulge.
 
  • Like
Reactions: Ratfacedgit

·
Registered
Joined
·
1,609 Posts
Discussion Starter · #28 ·
NH35A, this movement uses a push-pull system connected to an eccentric powered by a bi-directional rotor. I did not come up with a good mathematical equation so I rotated the rotor and counted the number of turns to achieve one rotation of the mainspring. It takes 7.25 rotations of the mainspring for a full wind. The rotor takes 150 rotations for 1 mainspring rotation. 150x7.25=1087.

Here is a video showing how it all works, first part is stem winding, then eccentric, then with the rotor installed.
https://www.youtube.com/watch?v=n1Oeu-Hq-hY

Stem winding has 6 gears with one compound gear. Ratios are 62/15=4.133, 15/10=1.5, 15/16=.935. 4.133x1.5x.935=5.8, it takes 5.8 full turns of the stem for one rotation of the mainspring, 5.8x7.25=42, 42 turns through the stem for a full wind.

seiko0.jpg

seiko1.jpg

seiko2.jpg

seiko4.jpg

seiko5.jpg

Here are readings on the timegrapher, 21600bph, 53 degrees lift angle, dial down.

1/4 wind, 10.5 stem turns,
192 degrees amplitude +2s.

1/2 wind, 21 turns,
220 degrees, +5s.

3/4 wind, 31.5 turns,
242 degrees, +3s.

Full wind, 42 turns,
254 degrees, +1s.

RFG


 

·
Registered
Joined
·
1,609 Posts
Discussion Starter · #29 ·
I forgot to mention the NH35 runs on average just over 250 degrees amplitude. Some can reach 280.

RFG
 

·
Registered
Joined
·
882 Posts
I consider myself above average when it comes to watch knowledge (although probably below average by F71 standards), but this thread makes me feel like I'm reading a foreign language. I learned a lot, and I'm fascinated by it, but Ratfacedgit is definitely many steps above anywhere I'll ever be.

Doc is obviously as well, but as a micro owner, he better be :)

Fascinating stuff Rat, keep it up, as crazy as it is, I do enjoy learning!
 

·
Registered
Joined
·
1,480 Posts
I consider myself above average when it comes to watch knowledge (although probably below average by F71 standards), but this thread makes me feel like I'm reading a foreign language. I learned a lot, and I'm fascinated by it, but Ratfacedgit is definitely many steps above anywhere I'll ever be.

Doc is obviously as well, but as a micro owner, he better be :)

Fascinating stuff Rat, keep it up, as crazy as it is, I do enjoy learning!
Hear hear!
 

·
Registered
Joined
·
10,147 Posts
Great job - and great photos, too.

It seems like that it's normal to have that kind of roughness all over the gears? Seems like in all these movements, there's so much bumps, scratches and roughnesses on every gear... I had not expected it to be so "un-smooth".
 

·
Registered
Joined
·
4,922 Posts
This is awesome. Hope to see more, particularly a 2824-2 and a SW200.

Has anyone researched similar studies to see if these numbers are similar across the same movement?

Lastly, have the Japanese mastered isochronism?

Sent from my PH-1 using Tapatalk
 

·
Registered
Joined
·
1,609 Posts
Discussion Starter · #37 ·
Great job - and great photos, too.

It seems like that it's normal to have that kind of roughness all over the gears? Seems like in all these movements, there's so much bumps, scratches and roughnesses on every gear... I had not expected it to be so "un-smooth".
The gears are not rough. When viewed with the naked eye or 10x magnification they appear quite polished. These photos were taken at 100x to 200x magnification, so every imperfection shows up. Also the important surfaces are highly polished and precisely machined.

Check out this example, the gear teeth are slick and smooth. This is 230x magnification. The gear is 5.2mm diameter and .10mm thick. That's .20 inches x .0039 inches. A human hair is 2-6 thousands of an inch thick. Expressed in decimal that is .002-.006 inch. So this gear is as thick as the average human hair thickness, .0039 or 3.9 thousands of an inch.

wheel.jpg
 

·
Registered
Joined
·
1,609 Posts
Discussion Starter · #38 ·
This is awesome. Hope to see more, particularly a 2824-2 and a SW200.

Has anyone researched similar studies to see if these numbers are similar across the same movement?

Lastly, have the Japanese mastered isochronism?

Sent from my PH-1 using Tapatalk
The numbers should be the same, they are 2824-2 clones, save for the power reserve. ETA, 40 hours, STP, 44 hours, SW, 38 hours.

I don't think the Japanese have mastered isochronism, they just make an outstanding movement.
 

·
Registered
Joined
·
12,807 Posts
Interesting.

Of course things vary individual movement to movement...but my personal observations have been

ETA 2824-2 minimal isochronism

Miyota 9015 moderate

Seiko NH35A moderate to high


Sent from my iPhone using Tapatalk
 

·
Vendor
Joined
·
27,721 Posts
... have the Japanese mastered isochronism?
I don't think the Japanese have mastered isochronism, they just make an outstanding movement.
I was wondering the same thing, and it seems tempting to draw that conclusion, if we assume (perhaps reasonably) that these two examples are representative of them all.

If nothing else, and assuming these are representative, then we can at least conclude that these Japanese movements (the NH35 and 9015) are quite resistant to isochronism, at least down to 1/4 power, which would seem noteworthy, in horological terms.

Isochronism isn't the only complicating factor when it comes to accuracy, and so resistance to it wouldn't seem like the be-all, end-all, as impressive as it is to see that sort of consistency in timekeeping from full down to low power.

I'm not necessarily proposing such a test, but if someone like RatFace wanted to do a real thorough assessment, it would need to include multiple postures (up to 6), so we could see how the movement is impacted by posture differences.

It's possible that in other postures, isochronism might have more or less impact on either, or that we might see more or less accuracy in other postures, with more or less posture difference.

If the results of the one position were to be repeated in the others, with a high degree of uniformity/consistency across postures, that would also be impressive (and, speaking for myself, not entirely surprising, but I'm a fan of the Japanese movements).

Beyond the effects of posture differences and dwindling power, anyone with a real urge to find the edge of their performance envelope could start adding some of the tests used in COSC testing, such as temperature variation, and if you really want to get extreme, adding some shocks to see how the movement likes being bumped about.

***

I was also wondering, as a hypothetical exercise, which performance profile would be preferable in day-to-day experience for the regular Joe, out in the real world?

Just taking these two examples of the STP vs the 9015, the STP is more accurate at full power, but also more prone to lose accuracy as power winds down, whereas the 9015 is less accurate at full power, but more consistent as it winds down.

Assuming neither's performance could be forcibly improved through further regulation or adjustment, and both had similar performance in other positions, would you rather have one's performance or the other, and if so, why?

Without numbers for either movement running at less than 1/4 power, I can only guestimate that the STP's average accuracy from full wind down to empty would be nearly identical to the scary-consistent numbers from the 9015, meaning the STP would continue to lose accuracy at a faster rate than the 9015 as they both wind down.

If you're moving around a lot, I suppose the STP will deliver better accuracy, if your activity is enough to keep it at full power. But otherwise, if it's more likely you're running on low power (under 1/4) more often than not, and if so, the 9015 would actually seem more accurate.

It probably seems nonsensical to say that, but I just saw a discussion on Facebook, prompted by someone asking what most people do to start their watch going before putting it on, and it seems like about half the people don't hand-wind enough to get a watch to 1/2 power.

They just give it a couple of shakes, which doesn't add nearly as much power as hand-winding, or they only give the crown a few turns, then they wear it. If they're not moving around a lot, there's a good chance they're not getting to full power, given the number of rotor rotations that would require.
 
  • Like
Reactions: ChiefWahoo
21 - 40 of 49 Posts
Top