COSC self-certifying your non-Swiss NH35

Could be interesting to see the results

The NH35 would fail miserably. While you can decent performance out of it if you tune it exactly to your use case, it will never pass the positional variances and will fail to keep spec as it winds down in testing.

The reason for the stress test is any movement can get to the standard, but how well they maintain it, including tough conditions, makes the standard. Lots of folks have calibrated cheap movements to a decent time, but in my experience they don’t hold it well, especially with any stress added to the mix.

The dirt cheap Seiko will never perform like the precise Swiss time pieces they test. The balance it's self has to be tuned perfectly to perform in the other positions.

I agree, the positional is the most challenging…but, read the criteria. It’s -6/+8. That’s a 14 second swing. Between horizontal and vertical. My time grapher is telling me now I get up to a 10 second swing between horizontal and vertical on two different NH35’s.

Yeah, that’s kind of my point. My $25 movements are actually exceeding those specification by my quick timegrapher readings. Two week logged runs in the field seems with 3 second variances imply that it looks good. The condition I put forth though is the power reserve is maintained on the NH35. Naturally watches running through a COSC test would not last 15 days without power maintained either.

I agree that all of the specs would fade during a wind down, though I don’t see that the COSC test has guidelines for allowing wind down. Naturally all watches would wind down over 15 days, so power must be maintained at some interval during a test. One could speculate that all test subject watches would have varying amounts of potential power reserve.

Might help to put the time aspect in perspective too…. COSC is a 1973 standard, developed during a pre-quartz era. NH35 came out 38 years later (2011). Plenty of time for an industry leading company to exceed the standards and minimize production costs. There is a train of thought that says…COSC’s value is primarily for marketing purposes. Since it’s only available for Swiss companies, keeping it attainable but Swiss-only, helps bolster the marketed notion of “Swiss-made” means something.

I don’t know what frequency they’re rewinding at, but I’m sure they let them wind down significantly to test that as well. My Omega AT performs within spec at 25% wind level, whereas my NHs are off the charts different when wound that low.

It’s important to remember COSC isn’t just +\- 4/6 spd dial up at 100% wind on a timegrapher, it’s within spec in a wide array of different conditions. Just because I can regulate a NH35 to 0 spd dial up fully wound on my time grapher, or +1 spd for my very specific office working day, does not mean it will perform the same in different conditions.

NHs are great workhorse movements, but they are not high end and do have significant variance in performance when you tweak use conditions even a tiny bit.

Yeah, I’ve witnessed the NH35’s performance drops off quite a bit when the power is low, though I don’t have any data on how low that is. Tough when they don’t have a power reserve meter, though perhaps one could do a controlled wind to ball park.

It’s important to remember COSC isn’t just +\- 4/6 spd dial up at 100% wind on a timegrapher, it’s within spec in a wide array of different conditions

The wide array of different conditions is spelled out in the original post. Regarding not “at 100% wind on a timegrapher”…that seems an obvious and practical speculation, since it defies imagination that someone would be keeping the test movements at 100% for 15 days straight at each orientation. Criterea 4 in original post point out a difference in horizontal and vertical positions of -6/+8, a quantitative value that can be measured. As mention a little above, that’s a 14 second variance, of which when I measure an NH35, I get about a 10 second variance between horizontal and vertical. …. So it seems we’re in agreement on that point. What is not known is the power level that is maintained for the testing. I agree that NH35’s fall off when power is low, though don’t have a quantitative feel for how low is low. However, we don’t know if that is actually part of the test criterea. 100% as you put forth seems an impractical value, and also an extreme that I doubt is necessary. Is it relevant to pass the test? I don’t see that it’s listed as a criterea to pass the test

Great project @skydave

I’ll probably get a time grapher one of these days, and likely an NH is the only one I’d try to regulate by myself

Like you say - it’s undefined how often the movement is wound - but if one was wearing an automatic watch with the NH continuously, it would remain nearly at full power during day, and perhaps go 8 or 9 hours at night without motion?

When I wear a watch with power meter for a few days non stop it stays perpetually at ‘full’. Only ever so slightly dropping at night, if at all.

So I think ~ 8 hours (overnight) might be a reasonable time frame between windings?

Update us on how it goes 👍

Yeah, I think you bring up a good point. We don’t know that the COSC testing standard says anything about how power level is maintained during testing…. But even if it did, what does it matter in practical terms since the movement we’re talking about is automatic, and would keep itself wound when in use.

What happens after you service these movements? I guess watchmakers only regulate the watch to keep good time but how do you know if it’s holding to COSC Standards? My other question is, do they test all the movements? Or do they just take a sample to certify a lot?

Surely it wouldn’t kill their English translators to just give us the name in a language we understand, even if they were to frustratingly still use the French initialism for some reason

And that why I love seiko movements. They are work horses, I have only ever had 3 movements fail, an omega 8900, eta 2824 elaboure grade, eta 2892.....