32xx movement problem poll and data thread

[enjoythemusic]

Vote... Can't recall.

Will do on readings... around Sunday or so. Let's give her a few days of daily wear.

[Easy E]

Thanks for the update. My 619LB was exactly a month to the day. It’s currently running great. All in all one month is tolerable, but that is the second one I have sent in this year.

[enjoythemusic]

Thanks for your patience, as wanted to give her time to settle/etc from the fresh service last week. Accuracy on-wrist (and off-wrist) seems to be a very consistent -2 sec/day.

Here she is after ~48 hours of initial 24/7 wrist wear, then fully hand-wound just to be sure.





.
.

Kept wearing her, and then set her down and let her run for 40-ish hours, so this is half-ish empty (from Saturday at 5:20pm to Monday at 12:10pm just sitting on my desk).







.
.

Immediately after the above photos, the CHNR was hand-wound to full and re-tested.

[Easy E]

Thanks for the update. Seems solid.

Question for you, nothing to do with the movement. In regards to the two that I sent in this year, in both instances I called Dallas and was told specifically that 32 warranties go to Long Island. Did you send yours to LI or Dallas? I'm really just curious more than anything.

Thanks again, glad your watch is back safe and sound.

[enjoythemusic]

Sent it to RSC Dallas, came back from RSC LI like yours.

[Easy E]

Interesting, thank you.

[Dirt]

Originally Posted by enjoythemusic

Thanks for your patience, as wanted to give her time to settle/etc from the fresh service last week. Accuracy on-wrist (and off-wrist) seems to be a very consistent -2 sec/day.

Here she is after ~48 hours of initial 24/7 wrist wear, then fully hand-wound just to be sure.

Attachment 1393021

Attachment 1393022

.
.

Kept wearing her, and then set her down and let her run for 40-ish hours, so this is half-ish empty (from Saturday at 5:20pm to Monday at 12:10pm just sitting on my desk).

Attachment 1393023

Attachment 1393024

Attachment 1393025

.
.

Immediately after the above photos, the CHNR was hand-wound to full and re-tested.

Attachment 1393026

Attachment 1393027

Attachment 1393028

Are you happy with the result?
That's the main thing

[enjoythemusic]

Heck yeah!!! I'd be honored to buy the Rolex watchmaker lunch and bier (or wine / spirits). A fine cigar afterward with good conversation.

[saxo3]

Originally Posted by Easy E

Ok, boys and girls, here it is. I ran 4 days of tests with two different watches - my oldest a BLNR 7/19, and my newest LV 4/23. So that's one older 3285, one newer 3235. The older one well within spec, the newer one is very close to edge. There is zero bias, zero emotion, zero fussiness, just data.

Correct.

Originally Posted by Easy E

As predicted I see no change of significance from day to day.

First view: there is an error in the listed numbers for your BLNR, it probably should read 258° and not 358°. Therefore, I have corrected the corresponding average values. Second remark: it makes no sense (to me) to list average values with 2 or 3 decimal places, see my proposal below.

Originally Posted by Easy E

The Test:
Wind each watch, wait 24hrs, take a reading, repeat for 4 cycles.For sure two variables to consider.
The first, did I wind them up with the same amount of power each day? I believe so, I went 60 turns with each, each day.
You certainly did that correctly, for complete winding of any 32xx caliber not more than 40 full crown turns (360° each) are required.

Issue two, which is just a function of my time to bench just is what it is - not exactly 24hrs between readings, so be it. I still don't think that had a material difference in the results.
I share your view.

Originally Posted by Easy E

Maybe someone with some better statistical analysis can chime in on these results.

Your timegrapher data, each set taken after full winding and during 4 consecutive days, are nicely reproducible. I trust these numbers.

But it triggers are different point, which we rarely discussed in this thread: it is all about understanding measurements and their associated errors.

In more detail: if one measures let's say 100 x the rate, amplitude, and beat error, using the same setup under identical conditions, the result will be a distribution (+/-) around an average value, which is called accuracy.

Accuracy is the difference between your measurement result and a well known reference, for example an atomic clock. Let's say you measure + 3.5 s/d with your timegrapher and repeat this measurement 100 x. Then you will achieve a distribution of numbers, let's say your data points vary between 3.2 and 3.8 s/d (just an example).

The 1-sigma value of this distribution is called precision. Accuracy and precision are not the same thing!
Below a graph (which I posted in May 2021) for better illustration.

My conclusion: your 4-days data set does not allow to distinguish if (a) the movement has small variations or (b) if the small differences are just given by the precision of your measurements. I think your numbers are precision dominated and that the BLNR movement is very stable.

[CharlesN]

But, ...

Looking at the photos of the watch on the TimeGrapher bracket it does look as if the watch was NOT correctly in the grip.

It looks like the crown was 180 degrees from its correct position. The crown should be against the microphone.

I do know that Weishi have printed some of their manuals with an incorrect picture.

The results however should be almost identical so don't panic.

[Easy E]

Originally Posted by saxo3

Correct.

First view: there is an error in the listed numbers for your BLNR, it probably should read 258° and not 358°. Therefore, I have corrected the corresponding average values. Second remark: it makes no sense (to me) to list average values with 2 or 3 decimal places, see my proposal below.






Your timegrapher data, each set taken after full winding and during 4 consecutive days, are nicely reproducible. I trust these numbers.

But it triggers are different point, which we rarely discussed in this thread: it is all about understanding measurements and their associated errors.

In more detail: if one measures let's say 100 x the rate, amplitude, and beat error, using the same setup under identical conditions, the result will be a distribution (+/-) around an average value, which is called accuracy.

Accuracy is the difference between your measurement result and a well known reference, for example an atomic clock. Let's say you measure + 3.5 s/d with your timegrapher and repeat this measurement 100 x. Then you will achieve a distribution of numbers, let's say your data points vary between 3.2 and 3.8 s/d (just an example).

The 1-sigma value of this distribution is called precision. Accuracy and precision are not the same thing!
Below a graph (which I posted in May 2021) for better illustration.

My conclusion: your 4-days data set does not allow to distinguish if (a) the movement has small variations or (b) if the small differences are just given by the precision of your measurements. I think your numbers are precision dominated and that the BLNR movement is very stable.

You are correct, I went back and checked my original data - the 358 is, in fact, 258. I suppose 358 would be near impossible. Good catch.

Thanks for the feedback.

[saxo3]

Originally Posted by Easy E

You are correct, I went back and checked my original data - the 358 is, in fact, 258. I suppose 358 would be near impossible. Good catch.

Thanks for the feedback.

[Easy E]

Originally Posted by saxo3

Let's say you measure + 3.5 s/d with your timegrapher and repeat this measurement 100 x. Then you will achieve a distribution of numbers, let's say your data points vary between 3.2 and 3.8 s/d (just an example).

I may not be able to fully explain what is in my head at the moment, but here goes....Starting with the notion that a target spec is +2/-2 and one takes a reading one time at any given ET, say 24hrs. Say that reading is +2.2. Technically at that moment you are out of spec. However if you repeated that process the following day you may be +1.8, in spec. Well, which is it are you in spec or not? That was partly why I wanted to give this a go, an perhaps what I imagine (in part anyway) was the question that lead to this.

This also speaks to why I made a thread about how far off spec others would tolerate. I think getting that specific is a bit over the top, but still an interesting point, to a point.

[Yipper69]

Got my watch back today. Sent to Dallas and shipped back from Dallas. Took exactly one month, not counting travel time.
I don’t have one of those machines that measures amplitude. I’ll see how it goes this time.
I do notice that there’s more resistance when setting the time and cant feel the ratcheting after watch is fully wound.

[Easy E]

Good that’s it’s back. Hopefully this time they got it.

[saxo3]

Originally Posted by Easy E

I may not be able to fully explain what is in my head at the moment, but here goes.... Starting with the notion that a target spec is +2/-2 and one takes a reading one time at any given ET, say 24hrs.

I think we must be more specific than the "target spec is +2/-2".

Rolex SA specifies "Precision -2/+2 sec/day, after casing".
They provide no more details about the tolerance for these numbers and how they measure this exactly.

In my view, it is not correct to simplify "one takes a reading one time at any given ET, say 24hrs."
Instead, we should speak about a measured 5-position average rate X after full winding.

To determine this accuracy, some timegrapher measurements must be performed under specific and reproducible conditions, which includes well defined temperature(s).

To find X, one measures in DU, DD, 3U, 6U, 9U positions and calculates the average. But each individual measurement contains an uncertainty (as explained in post 4479), which you have not (yet) determined with your Weishi timegrapher.

Originally Posted by Easy E

Say that reading is +2.2. Technically at that moment you are out of spec. However if you repeated that process the following day you may be +1.8, in spec.

Most owners would think like you, but one cannot draw this conclusion, imho, without reflecting about the measurement error, which is the precision.

Let us imagine the precision for both data points would be +/- 0.4 s/d. Hence you would compare X = 2.2 +/- 0.4 s/d with X = 1.8 +/- 0.4 s/d. These two accuracies are identical within the precision of you measured numbers.

Originally Posted by Easy E

Well, which is it are you in spec or not?

As explained above, a conclusion "in spec" or "out spec" is not trivial if the measured average accuracy X is very close to the specs value.

But, in the absence of Rolex specifying any tolerance for the -2/+2 sec/day values, it is correct to say that X = 2.2 s/d is outside and X = 1.8 s/d is within Rolex published specs.

----
Addition: I did not comment on the results presented for the LV watch, see post #4458. All my remarks are the same as for the BLNR watch, except that the LV is already sick and will be soon outside the amplitude specs (min 200°, 24 hours after full winding). Its X-rate value is close to fall outside Rolex specs (-2 sec/day).

[Easy E]

Most unfortunate really. The LV is barely 6 months old and has been worn maybe 6 times. Will check it again likely end of year.

[countzero]

Very informative thread. Wonder if this is a 32xx issue, or a Chronergy escapement issue?

That is, the new movements in updated Sky-Dweller and Daytona - 9002 and 4131/4132, respectively - now also have Chronergy escapements, as of this year. One wonders if we might see the same issues plaguing these movements as well.

Does anyone already have experiences with these new movements, or insights regarding the matter? Any comments from the forum resident Rolex watchmakers?

[saxo3]

Originally Posted by Easy E

Most unfortunate really. The LV is barely 6 months old and has been worn maybe 6 times. Will check it again likely end of year.

Interesting that your in 05/2023 bought Submariner (126610LV) with 3235 movement is already so bad.

The horizontal and vertical amplitudes degraded very fast, and that happens to a new watch in 2023, 8 years after the introduction of the 32xx in 2015.

[Glidelock]

I have been keeping this thread in mind since its original post, and keeping an eye on my 126610. It's funny, because in the beginning (especially) there seemed to be a lot of criticism over paying to much attention to a second here or there. However it started to become exceptionally clear relatively early in this thread that this spider does in fact have legs, and much like this thread those legs just seem to keep growing as time (literally) starts to "slip" by on these movements. But for me, it was not an issue whatsoever, as my 2.5 year old 126610 has kept excellent time within +1 to -2 sec on average since it was new. That was until it just simply fell off a cliff (no...not literally ) about 2 weeks ago! It currently loses 16 seconds a day on a full wind, but in a 48 hour period it starts losing several min. by 58-60 hours it loses 6-10 min a day! I believe the post indicating at this time it's no longer a ? of if, but when this will happen to a watch running 3235 movements. either way, looks like I will be submitting for warranty work before to long. However this is obviously not thrilling to me, as I own multiple iterations of Rolex watches running 3135's and although I have had to have 3135 movements serviced before, they overall have been absolute Tanks as far as reliability!

I am seeing posts where people have sent them in, and they've come back working well within specs, only to absolutely start repeating the cycle again down the road. Has anyone had this fixed and had it back for a long enough period that they truly feel like it's been resolved?

[LNGSD43]

Mine feels quite different since it came back second time. Winding is different, crown feels a bit different and setting feels different. Hard to describe and whether it means anything I can only hope.

[the dark knight]

Originally Posted by Yipper69

Got my watch back today. Sent to Dallas and shipped back from Dallas. Took exactly one month, not counting travel time.
I don’t have one of those machines that measures amplitude. I’ll see how it goes this time.
I do notice that there’s more resistance when setting the time and cant feel the ratcheting after watch is fully wound.

I have 3 32xx watches, all purchased this year. None have issues yet, knock on wood.

But interestingly one of them (DJ36) has much more resistance when setting time than the other two and no ratcheting after full wind. I did notice that immediately because I've owned quite a few other 32xx watches too over the last couple of years and this one is the only one that has quite a bit of resistance when setting time.

It's also the only one of the 3 I own now that tends to lose time when I wear it over a prolonged period (but still within spec).

[Yipper69]

Originally Posted by LNGSD43

Mine feels quite different since it came back second time. Winding is different, crown feels a bit different and setting feels different. Hard to describe and whether it means anything I can only hope.

I got my sd43 back last week and so far so good. It’s about 2 seconds fast after a week.
I do notice that winding the watch is a lot stiffer than before but the auto wind while wearing is much more efficient. Before I sent it in the last time, i could wesr it all day and if i wind it, it would take approximately 5-10 turns. Now it’s fully wound after wearing it for the day.

[t_serban]

Originally Posted by t_serban

Here are the measurements for my 124060 Sub with the 3230 movement, purchased new on March 1st 2023, using a Weishi 1000.
The watch was rested DU and not moved at all between measurements.

After full wind

Position    Rate      Amplitude      Beat error
DU          +2        281            0.3
6U          +1        243            0.0
9U          +2        246            0.0
3U          -1        244            0.3
DD          +2        272            0.1

Avg         +1.2      257.2          0.14
Stddev      +1.16     16.04          0.13

After 12h

Position    Rate      Amplitude      Beat error
DU          +3        264            0.2
6U           0        232            0.0
9U           0        235            0.0
3U           0        228            0.3
DD          +3        270            0.2

Avg         +1.2      245.8          0.14
Stddev      +1.46     17.55          0.12

After 24h

Position    Rate      Amplitude      Beat error
DU          +2        256            0.2
6U          +2        227            0.0
9U          +2        230            0.1
3U           0        217            0.3
DD          +2        259            0.2

Avg         +1.6      237.8          0.16
Stddev      +0.8      16.67          0.10

Time for an update to these measurements since I'm now into the 8th month of ownership.

[saxo3]

Originally Posted by t_serban

Time for an update to these measurements since I'm now into the 8th month of ownership.

Thanks for returning to this thread
I'm looking forward to your new data.

[whatsthetime?]

Originally Posted by LNGSD43

Mine feels quite different since it came back second time. Winding is different, crown feels a bit different and setting feels different. Hard to describe and whether it means anything I can only hope.

In what way is it different before and after service?

[LNGSD43]

Originally Posted by whatsthetime?

In what way is it different before and after service?

Hard to describe. Smoother winding somehow, more resistance than before when moving the hands. Just feels a bit more solid I guess.

[t_serban]

Sadly I missed the 12h mark, but amplitude still looks good at 0h and 24h after 7 and 1/2 months of ownership.

After full wind

Position    Rate      Amplitude      Beat error
DU          +4        283            0.3
6U          +4        253            0.0
9U          +2        260            0.0
3U          -2        254            0.4
DD          +3        281            0.2

After 24h

Position    Rate      Amplitude      Beat error
DU          +3        265            0.2
6U          +3        232            0.1
9U           0        232            0.0
3U          -1        226            0.5
DD          +3        260            0.3

For reference, these are the values measured after purchase:

After full wind

Position    Rate      Amplitude      Beat error
DU          +2        281            0.3
6U          +1        243            0.0
9U          +2        246            0.0
3U          -1        244            0.3
DD          +2        272            0.1

After 24h

Position    Rate      Amplitude      Beat error
DU          +2        256            0.2
6U          +2        227            0.0
9U          +2        230            0.1
3U           0        217            0.3
DD          +2        259            0.2

[Andad]

An interesting post regarding the 32** movement.

https://www.rolexforums.com/showpost...1&postcount=14

[SearChart]

Originally Posted by Andad

An interesting post regarding the 32** movement.

https://www.rolexforums.com/showpost...1&postcount=14

Nothing new really.