32xx movement problem poll and data thread

[Schnaps]

My results









Bought in May 2021

[Dirt]

Originally Posted by dannyp

Yes, it's an interesting theory. The fact that the whole barrel mechanism is meant to be replaced could also mean that fewer of them are taken apart and examined at service (making the problem harder to discover and diagnose originally).

Perhaps the thought at Rolex is simply that SOP will be to replace when watches come in for service until the bad batch/batches have been depleted.

Speaking of... One or more bad batches (that are presumably weak in the same way) would explain a few things:

1. Why some watches go bad and others don't.
2. Why the problem recurs on some but not others.
3. Why the problem always seems to present around the same point in a watch's life.




Good point. However, defects can have systemic effects. Is it not conceivable that a spring problem could manifest itself as wear on a seemingly unrelated part due to a chain reaction of sorts? It would also make sense then why the lubrication wasn't a permanent fix. Analogous to:

I get frequent lower back pain (muscular). Try lots of different things. Then I realize: when working out I have a tendency to overuse/not stretch a muscle in my leg that attaches to those in the lower back. The result is a tightening in the leg that pulls at my back, making it hurt. But it's not a back problem, so the back muscles aren't what to address.

Not saying it is the case, but just a possible reconciliation of different ideas.

Some interesting points that are possibly worth more consideration
But I imagine that Rolex know more about their springs and the longevity of them than they are getting credit for.
After all, they've been at it for a very long time and I fully expect the replaced parts are often sent back to the mothership for some degree of analysis to discover if there is a deficiency in quality.
Especially given they are already devoting a disproportionate amount of resources to the issue.
Of note, i have also seen a youtube video of a service to a 32xx movement that apparently restored the timekeeping of the watch and for some reason the Mainspring and Barrel weren't changed. They were serviced in the usual manner and re-used to good effect. This suggests there is nothing wrong in that regard when taken at face value

We already know from someone who is extremely hands on with these things, that one factor which may be compounding the issue is the metalurgy of the pinions in this modern era. It should be something that's rather easily addressed if actually deemed to be a part of the problem rather than being symptomatic.
I note the latest fix for the Omega 3861 as an example of how a seemingly inconsequential change can make a rather profound difference. Certainly enough to potentially see the movement conceivably go the full distance between what is deemed to be a reasonable service interval as far as Omega are concerned. But that's a rather fluid thing in it's own right

I do have a theory which I think has a degree of plausibility that i have held now for quite a long while, that also has its roots in the automotive industry based upon my own experience which I think may be outside of the realms of experience of Rolex or perhaps even the Horological world in general which may apply to the woes of this movement series.
But for all we know, Rolex may have found a solution that meets their criterior anyway

[Dirt]

Originally Posted by Schnaps

My results









Bought in May 2021

Ok, we can see what's going on there.
How is it performing on the wrist in a day to day scenario?
Has timekeeping been declining over a period of time?
Thanks

[Schnaps]

Originally Posted by Dirt

Ok, we can see what's going on there.
How is it performing on the wrist in a day to day scenario?
Has timekeeping been declining over a period of time?
Thanks

Strange thing is that is has been performing -2 sec/day since new 2021. Latest 4 months on my wrist 3-5 days/week and in beetwen in winder.

1 october i changed the date+set the time.
noticed already same evening it was slower.
Now its steady -6 sec/day since the date change

[shedlock2000]

Originally Posted by Dirt

The pinion wear is a common feature with one of the warranty/service fixes being to put lube on it but it doesn't prevent it from coming back when it eventually goes south again.
I think it's routine to replace the pinion on these
Even lube can't prevent the premature wear

Hmm, ok. But that’s not necessarily the cause of the issue — it’s just often correlated with it.

I’ve seen the photo, but does anyone have an exploded diagram of the movement? I don’t know on what the pinion is rubbing (or what is rubbing on it) without knowing where the wheel sits in relation to the pinion.


Sent from my iPhone using Tapatalk

[shedlock2000]

Originally Posted by Dirt

But I imagine that Rolex know more about their springs and the longevity of them than they are getting credit for.
After all, they've been at it for a very long time and I fully expect the replaced parts are often sent back to the mothership for some degree of analysis to discover if there is a deficiency in quality.

That’s possible, but sometimes the simple things can get overlooked — especially if a few which were tested happen to be ok rather than from a bad batch. With the data received in terms of repaired 33xx movements (the data show nothing consistent in terms of repair and much of what is actually changed or inspected is not communicated) it’s difficult to determine what is or is not consistently damaged/faulty on all movements. We basically don’t have enough actual data on faulty components removed from suspect movements to know what is or is not causal or corollary. Throwing parts at a movement to resolve an issue has a habit of spoiling any differential diagnostic process.

Originally Posted by Dirt

Of note, i have also seen a youtube video of a service to a 32xx movement that apparently restored the timekeeping of the watch and for some reason the Mainspring and Barrel weren't changed. They were serviced in the usual manner and re-used to good effect. This suggests there is nothing wrong in that regard when taken at face value

I was led to believe the barrel is not serviceable at all and must be exchanged. Assuming that what you say is correct and that the YouTuber has managed to resolve the issue on a bench in his shed (or wherever), this seems to contradict what you said above about Rolex being well directed and heavily invested in the repair of the movement. Moreover, it seems to suggest that the resolution is one that can be resolved without the resources Rolex have available to them. Of course, this is not impossible (my suggestion is such a fix), but it at least invalidates the argument you used against my fault spring hypothesis, as it implies Rolex *do* actually overlook some things that (random) YouTuber has identified by himself.


Sent from my iPhone using Tapatalk

[dannyp]

Originally Posted by shedlock2000

I was led to believe the barrel is not serviceable at all and must be exchanged. Assuming that what you say is correct and that the YouTuber has managed to resolve the issue on a bench in his shed (or wherever), this seems to contradict what you said above about Rolex being well directed and heavily invested in the repair of the movement. Moreover, it seems to suggest that the resolution is one that can be resolved without the resources Rolex have available to them. Of course, this is not impossible (my suggestion is such a fix), but it at least invalidates the argument you used against my fault spring hypothesis, as it implies Rolex *do* actually overlook some things that (random) YouTuber has identified by himself.


Sent from my iPhone using Tapatalk

My understanding is that the spring barrel wasn't intended to be serviced (something about the thinness of the walls). That doesn't mean one can't service it, just that one (or more) of the following is true:

1. There's a high probability of breaking when servicing.
2. The time needed to painstakingly work on it costs Rolex more than a replacement.
3. There's a chance that, even if it is serviced and works, the structural integrity will be compromised.

[Dirt]

Originally Posted by shedlock2000

That’s possible, but sometimes the simple things can get overlooked — especially if a few which were tested happen to be ok rather than from a bad batch. With the data received in terms of repaired 33xx movements (the data show nothing consistent in terms of repair and much of what is actually changed or inspected is not communicated) it’s difficult to determine what is or is not consistently damaged/faulty on all movements. We basically don’t have enough actual data on faulty components removed from suspect movements to know what is or is not causal or corollary. Throwing parts at a movement to resolve an issue has a habit of spoiling any differential diagnostic process.

Sent from my iPhone using Tapatalk

There is more than enough information available to us on this forum going back a number of years as to the standard scope of work Rolex has put in place to deal with the warranty claims on this movement.
Anything else is based on common sense when the movement is torn down by the watchmaker.

[Dirt]

Originally Posted by shedlock2000

Hmm, ok. But that’s not necessarily the cause of the issue — it’s just often correlated with it.

I’ve seen the photo, but does anyone have an exploded diagram of the movement? I don’t know on what the pinion is rubbing (or what is rubbing on it) without knowing where the wheel sits in relation to the pinion.


Sent from my iPhone using Tapatalk

As I've said before, the pinion wear is probably going to be deemed to be symptomatic or correlated as you say, of a more sinister problem with the movement. At least that's the way I view it and how my own theory applies to the broader problem.

I believe the pinion by nature and at the site of abnormal wear runs in a jewel from memory. Though it could be running in a bush of some kind, but I doubt it.

Metalurgy could play a part in the wear factor and this has been communicated to us quite a long time ago.

[Dirt]

Originally Posted by Schnaps

Strange thing is that is has been performing -2 sec/day since new 2021. Latest 4 months on my wrist 3-5 days/week and in beetwen in winder.

1 october i changed the date+set the time.
noticed already same evening it was slower.
Now its steady -6 sec/day since the date change

Thanks
It's certainly a head scratcher.

Ordinarily we can expect little nuances to occur with a mechanical movement in regard to timing of which i imagine you are well aware. Even one that's a high grade movement
But what you are experiencing is at least double of anything I've ever encountered of a transitory nature with a Rolex Chronometer grade movement over decades of ownership of previous series of their movements with yours being of a permanent nature

It may be suffering from the dreaded 32xx lurgy.
Is there a personal threshold where you would be inclined to pursue it with Rolex?
Or will you let it run its course until just before warranty is up regardless and send it off to be fixed?

[Dirt]

Originally Posted by dannyp

My understanding is that the spring barrel wasn't intended to be serviced (something about the thinness of the walls). That doesn't mean one can't service it, just that one (or more) of the following is true:

1. There's a high probability of breaking when servicing.
2. The time needed to painstakingly work on it costs Rolex more than a replacement.
3. There's a chance that, even if it is serviced and works, the structural integrity will be compromised.

That is all true, but it seems as though it may not be impossible.
Though it was only a youtube video and who knows how well it reflects real world outcomes.

[saxo3]

Originally Posted by Schnaps

My results









Bought in May 2021

You let us discover the numbers from your photos? OK, your rates data:

DU: -3.7 s/d
DD: -1.5 s/d
9U: -4.4 s/d
9U: -5.1 s/d
3U: -2.8 s/d
12U: -7.9 s/d
12U: -8.6 s/d
12U: -6.2 s/d

All measured after full winding?

The 32xx is regulated in 5 (not 6) positions and 12U is normally not measured but 6U.

The average rate X cannot be determined from your data set since you did not measure in 6U position.

I recommend you repeat the measurements after full winding and again after 24 hours at complete rest, for positions DU, 6U, 9U, 3U, DD.

Lift angle is 53 and not 52 degrees, which you used.

Good timegrapher (Witschi Watch Expert) but too much Schnaps?

————
PS: a timegrapher measurement procedure you can find in this thread in posts #1425 and #771

https://www.rolexforums.com/showpost...postcount=1425
https://www.rolexforums.com/showpost...&postcount=771

[Andad]

Originally Posted by Dirt

Ok, we can see what's going on there.
How is it performing on the wrist in a day to day scenario?
Has timekeeping been declining over a period of time?
Thanks

Was it fully wound?

Unless we know the condition of the PR there is nothing to see.

[shedlock2000]

Originally Posted by avera

I know there is not consensus on the cause or how to fix it but does going in for a stay at the RSC seem to permanently fix the problem or have there been reports of it happening more than once to the same watch?

There are multiple occasions when a watch has been sent in to correct the issue and for that watch to again suffer the slowing down issue post repair.

The general consensus of opinion here, so far as I understand it, is that whatever Rolex seem to do does not rectify the fault and the issue crops up again another few months down the line. Accordingly, it seems that the issue is systemic.

What is not clear is why some movements suffer and some don’t.


Sent from my iPad using Tapatalk

[amanbra]

Originally Posted by avera

As posted a couple of pages back, my 126710BLRO slowed down from a very consistent -1spd to a very consistent -4spd after 13 months on the wrist. I took it to my watch maker and the amplitude was under 200. I then took it into the Los Angeles RSC. Fun experience all around. It came back on Saturday after 5 weeks on the inside. The letter I got back with it said something like “Service performed”. I don’t have an amplitude-o-meter so I can’t check that but it is running at about -1spd again. I have read quite a bit of the ~100 pages on the this topic but not all of it. I know there is not consensus on the cause or how to fix it but does going in for a stay at the RSC seem to permanently fix the problem or have there been reports of it happening more than once to the same watch?

my DJ has been to RSC twice for the same issue. This December it will hit the 1 year mark from the second visit. At the moment it's doing exactly the same as the day I got it back the second time around. No one can say if this is a perm fix and what they changed etc but it's doing very well the second time around.

The issue is though, with the lack of communication from Rolex I don't think I'll ever feel fully at easy with the watch... maybe if it stays like this for the next 5 years... but yeah... who knows...

[S7gpt]

126610ln coming up to a year old, worn ever day and constistant -2spd - fingers crossed it stays that way

[sevykor]

Originally Posted by avera

As posted a couple of pages back, my 126710BLRO slowed down from a very consistent -1spd to a very consistent -4spd after 13 months on the wrist. I took it to my watch maker and the amplitude was under 200. I then took it into the Los Angeles RSC. Fun experience all around. It came back on Saturday after 5 weeks on the inside. The letter I got back with it said something like “Service performed”. I don’t have an amplitude-o-meter so I can’t check that but it is running at about -1spd again. I have read quite a bit of the ~100 pages on the this topic but not all of it. I know there is not consensus on the cause or how to fix it but does going in for a stay at the RSC seem to permanently fix the problem or have there been reports of it happening more than once to the same watch?

More than once for me… The final fix was to sell then32xx and went back to the 31xx. No more issues now 4 years out


Sent from my iPad using Tapatalk

[alphadweller]

Updating my latest post #2934 with new data from real-life wearing over 10 days.

Model: TT Blue Sub41
Purchase date: Sep 2020
Wearing habit: at the weekend

For this test, I wore the watch over 10 days.
I've been able to play with the positional variance at rest to achieve a deviation close to 0s.

[saxo3]

Originally Posted by alphadweller

Updating my latest post #2934 with new data from real-life wearing over 10 days.

Model: TT Blue Sub41
Purchase date: Sep 2020
Wearing habit: at the weekend

For this test, I wore the watch over 10 days.
I've been able to play with the positional variance at rest to achieve a deviation close to 0s.

Victor, a nice study and data collection, your watch is running very well and you control its timekeping via the rest position, perfect.

A "special feature" of the 32xx movement is that the caliber can remain very accurate for a long time (several months), even with rather low amplitudes, before its accuracy deteriorates and you can no longer compensate by rest position. The key indicator is then a too low amplitude after full winding.

It would be interesting to measure the amplitudes (5 positions) of your watch after full winding to complete the picture for your Sub41. Maybe you know somebody who owns a timegrapher and can do this measurement for you?

[CharlesN]

Excellent and well designed results.
Very informative
A real treat to see such organised results charts and graphs
Thank you.

Just think how much more information you would have with a Timegrapher.

They can be easily bought online for not many Euros simply.

[dannyp]

Originally Posted by saxo3

Victor, a nice study and data collection, your watch is running very well and you control its timekeping via the rest position, perfect.

A "special feature" of the 32xx movement is that the caliber can remain very accurate for a long time (several months), even with rather low amplitudes, before its accuracy deteriorates and you can no longer compensate by rest position. The key indicator is then a too low amplitude after full winding.

It would be interesting to measure the amplitudes (5 positions) of your watch after full winding to complete the picture for your Sub41. Maybe you know somebody who owns a timegrapher and can do this measurement for you?

The one that really shocked me was the 9UP results, where the watch saw the greatest time gain. I thought that 9UP was one of the slower positions. But looks like the watch was only in that position for two hours total, so perhaps those were two hours where the watch was already gaining time from its previous wearing/positioning and didn't have a chance to slow any.

[saxo3]

Originally Posted by dannyp

The one that really shocked me was the 9UP results, where the watch saw the greatest time gain. I thought that 9UP was one of the slower positions. But looks like the watch was only in that position for two hours total, so perhaps those were two hours where the watch was already gaining time from its previous wearing/positioning and didn't have a chance to slow any.

Danny, nothing shocks me anymore with 32xx calibers.

Movement rates depend on how a mechanical watch is regulated and there are no general rules for positive or negative rates in specific positions, i.e. there are no "slower positions". I have seen this in many of my timegrapher measurements.

[shedlock2000]

So I wanted to ask what @saxo3 and others think of the distribution of errors. I think we are roughly at 1/3 of 32xx movements being problematic. This implies that 2/3 are trouble free.

What can be deduced from this is:
1) That those that have reported no issues have not checked properly enough,
2) That some movements don't have the fault,
3) That some owners have not had theirs long enough for the issue to develop.

1) Assuming that some of the checked movements that havent been identified as faulty actually are, we could increase the error rate a bit to account and suggest that, perhaps, half of the movements are faulty -- but this brings us to 2)

If we accept 2), there are two possible causes of failure: build or manufacturing defect? Which also implies that:
a) if the fault is a component issue, then some movements are not fitted with the same batch of components, or
b) that the build of those movements are different than on other movements. a) suggests there is a batch-fault issue which is resolvable by new components while b) is resolvable by a rebuild.

That re-built movements seem to generate the same fault after a period of time implies either that there are consistent issues that specific movement, or that the batch of faulty components is big enough that it is taking time to use them up.

3) indicates that time will identify a greater percentage of faulty movements.

That 2/3 (or 1/2 if we are being generous) of movements are not faulty -- even given some leeway for improperly checked movements or those which are too new to have faulted yet -- implies there is not an inherent design issue (as such a design flaw would present in 100% of movements).

[amanbra]

Originally Posted by dannyp

The one that really shocked me was the 9UP results, where the watch saw the greatest time gain. I thought that 9UP was one of the slower positions. But looks like the watch was only in that position for two hours total, so perhaps those were two hours where the watch was already gaining time from its previous wearing/positioning and didn't have a chance to slow any.

My explorer the 3 down position is fastest. It read a real head scratcher but it translates to the real world. Leaving it crown down over night speeds the watch up.


Sent from my iPhone using Tapatalk

[dlrolex]

perhaps I will stick with ETA 2824

[Dirt]

Originally Posted by shedlock2000

So I wanted to ask what @saxo3 and others think of the distribution of errors. I think we are roughly at 1/3 of 32xx movements being problematic. This implies that 2/3 are trouble free.

What can be deduced from this is:
1) That those that have reported no issues have not checked properly enough,
2) That some movements don't have the fault,
3) That some owners have not had theirs long enough for the issue to develop.

1) Assuming that some of the checked movements that havent been identified as faulty actually are, we could increase the error rate a bit to account and suggest that, perhaps, half of the movements are faulty -- but this brings us to 2)

If we accept 2), there are two possible causes of failure: build or manufacturing defect? Which also implies that:
a) if the fault is a component issue, then some movements are not fitted with the same batch of components, or
b) that the build of those movements are different than on other movements. a) suggests there is a batch-fault issue which is resolvable by new components while b) is resolvable by a rebuild.

That re-built movements seem to generate the same fault after a period of time implies either that there are consistent issues that specific movement, or that the batch of faulty components is big enough that it is taking time to use them up.

3) indicates that time will identify a greater percentage of faulty movements.

That 2/3 (or 1/2 if we are being generous) of movements are not faulty -- even given some leeway for improperly checked movements or those which are too new to have faulted yet -- implies there is not an inherent design issue (as such a design flaw would present in 100% of movements).

It doesn't imply that 2 out of 3 are trouble free.
It potentially implies that more than 2 out of three are having an issue when we factor in the possibility that not all owners are aware of a problem or simply don't even care

[saxo3]

This thread is running now since 21 months.

Here is a graphical display, together with the corresponding numbers, about the outcome so far.





Two (at least) interesting points, which have not changed since my last statistics update in January 2021 (post # 2229, page 75):

The quantity of 32xx watch owners that observe and report issues with their movements did not decrease over time but remained rather constant at a level of about 28 – 30 %.

There are still 4 times more poll voters than different contributors to the thread. That means the majority voted but did not post in this thread.

[saxo3]

Originally Posted by shedlock2000

So I wanted to ask what @saxo3 and others think of the distribution of errors.

The answer is simple: this poll does not provide a statistically relevant result because there are too many unknowns and any poll interpretation or conclusion (about the percentages) requires a set of assumptions that no one can verify or falsify.

The 32xx problems exist, on a large scale and for several years, there is no doubt about that fact. It is more common than some on this forum want to hear or accept.

[Dirt]

There is no doubt about how abnormally prevalent the problem is.
On this forum and others, there is tons of conbined anecdotal evidence based on many years of ownership of watches which is not necessarily limited to Rolex watches to draw from.
We have deep insight from someone within the system with pictures and more than enough explanations around it going back through the Dark-years until 2018, complete with an account of greatly increased resources being put into it at RSC's.
We have quiet admissions from the retail end that there may be increased warranty claims around this problem.
Also there is some insight from a fairly well known industry insider, that the problem is rather well known among watchmakers generally that are not necessarily attached to Rolex. They do talk with each other.

We will never be able to glean the deeper depths of the actual distribution of errors, of which there are sure to be many but the trend is made patently clear through this thread
I am confident nobody really enjoys this stuff because it reflects poorly on the new offerings from the crown as it currently stands with the exception of the other movements not covered by this thread. The other movements are reportedly(or lack there of) going just fine with or without minor/rountine updates and in accordance with Rolex historical norms(anecdotally speaking).

[HiBoost]

Originally Posted by Dirt

We have quiet admissions from the retail end that there may be increased warranty claims around this problem.


Also there is some insight from a fairly well known industry insider, that the problem is rather well known among watchmakers generally that are not necessarily attached to Rolex. They do talk with each other.

Do you have a link or reference to these two points? I'd be curious to read more. On another forum I was in a discussion with a watchmaker (professionally trained by another large Swiss brand) who indicated his colleagues within RSC itself were overwhelmed and overworked trying to keep up with 32xx related issues. It sounded a bit alarming if accurate.