Can end-shake and or side-shake ever be too small?

[VWatchie]

I just serviced a Vostok cal. 2431 in overall very good condition. I've had it since it was new (2014) and it was performing reasonably well before servicing it. It's identical to the cal 2416B but has been modified (by Vostok) to make the hour hand rotate 360 degrees only once per day (24h movement). Anyway, in the process, I carefully adjusted the end-shake to be as small as possible while carefully making sure all wheels were spinning freely.

After the service, the amplitude became extremely erratic. The dial-up position jumps between 300 degrees and 240 degrees in what I perceive to be a random manner. I can't detect any particular pattern for the variations. It appears that the amount of power reaching the escapement varies randomly.

Now, Vostok movements aren't exactly known for their precise tolerances. Could it be that these movements don't handle super fine end-shake as well as modern Swiss movements? Perhaps I should try to increase the end-shake? Of course, and as always, the problem could be something entirely different, but as I did adjust the end-shake to be as small as possible (0.01 - 0.02 mm) that's what I suspect could be the problem, no?

I guess my question is, can end-shake and/or side-shake be too small even when all wheels are perfectly free (Libre Sans Jeu)?

[Shane]

.01 - .02 mm / .0004 - .0008 in, is quite small.  Some mechanical assemblies require a higher degree of clearance do to a wider acceptable tolerance within all the components making up that assembly.

Once you pick a design philosophy, you need to keep with it all the way through the assembly (assembly is being used as both a noun and a verb).

Dimensional and angular deviations abound, total indicated runout of the shafting and the pinions and pivots machined into them, as well as the ID in relation to the OD of the gears those shafts are pressed into, the engagement between the involuted gear pairs, times how many pairs within that gear train.  Not to mention, if the pivots are at the end of a tapered shaft, you have also restrained them all radially as well.

It's hard to say what the problem or problems might now be but if it was running better before you tightened it up, is it really perfectly free now?  It sounds like you will need to loosen it back up.

Let us know how you make out.

Good luck.

Shane 

[nickelsilver]

 Depending on the sideshake and diameter of the arbor at the pivot, you can have a situation where the shoulders at both ends are making contact with both hole jewels at the same time if the endshake is very small. It can be exacerbated by very short arbors too, particularly the pallet fork*. 0.01mm is hard to visualize, I made a sketch of an arbor with jewels and exactly that much endshake to illustrate.

 

*Troubleshooting a new-build tourbillon a few years back I had a situation where the fork was ultra free, with "feelable" endshake, but the watch was having amplitude issues- ended up adding 0.01mm endshake fixed it.

 

 

[VWatchie]

Thanks for your input. Much appreciated! I can't say that I fully get the theory of it, but it's clear to me that increasing the end-shake a bit is worth a try. Reading your messages I now recall that I also replaced one of the 3rd wheel jewels and it could be that the hole diameter of the new jewel was a tad bit small. I'll look into that as well.

9 hours ago, Shane said:

Let us know how you make out.

Good luck.

Will do! Thanks!

[nickelsilver]

Here's a cutaway with close ups of the two ends. The sideshake is 0.01, which is about right for a 0.15 pivot (this size). If the arbor is under tension, it will generally want to tilt- like when you try to wind the watch with the wheel train in but haven't put the bridge yet- everything tilts . With a very small but existant endshake, the shoulders of the arbor can touch both flat surfaces of the two jewels at the same time. With a little more endshake, as the tilt is restricted by the pivots, it will become possible for one shoulder to touch one jewel (and is almost certainly the case 99.999999% of the time), but both shoulders can't contact both jewels at the same time.

 

 

Edited August 15, 2022 by nickelsilver

[Shane]

10 minutes ago, nickelsilver said:

Here's a cutaway with close ups of the two ends.

Nice CAD work.

VWatchie, also remember that Nickelsilver probably also modeled those parts in theoretical perfection.  All dimensions spot on, lines that should be parallel are, all angles are accurate without internal filleting or external rounding, concentricity or T.I.R. zero.  Which is all easy to draw, very (impossible) to obtain.  Watch parts (even "cheaper" mass produced watch parts) are probably closer than most but will (on average) deviate from print.  That's called acceptable tolerance.  If you don't want to hand fit every part in every assembly, you assemble using fixtures adjusted for parts assuming the widest possible "acceptable" deviations.

And as Nickelsilver said, while under torque is not the same thing as when it is free wheeling.

Shane 

[JohnR725]

21 hours ago, VWatchie said:

but as I did adjust the end-shake to be as small as possible (0.01 - 0.02 mm) that's what I suspect could be the problem, no?

I thought I would compared to Rolex so I snipped out some other specifications. When you're looking at the chart you do need to find a watch comparable in diameter size as others have pointed out that will change things. You'll notice that your beating out there specifications you're much tighter than you really should be especially with a watch of that quality.

11 hours ago, VWatchie said:

Reading your messages I now recall that I also replaced one of the 3rd wheel jewels and it could be that the hole diameter of the new jewel was a tad bit small. I'll look into that as well.

When I'm reading this it's reminding me of clock repair typically by a watchmaker but not necessarily. But usually when a watchmaker looks at the clock things seem really sloppy and there's an obsession to put bushings in. Typically only need a bushing if the hole is no longer is round. No matter how sloppy it is if it's still round that's the way it was made. So then obsession with bushings you can end up with clocks that will not run on the weight there designed for because it's just too darn tight. So yes if you get carried away with getting your tolerances super super tight you're going to have a running issue in the watch. Especially if the watch was manufactured with more liberal tolerances that requires greater spacings possibly bigger hole diameters for pivots.

 

[VWatchie]

On 8/15/2022 at 12:19 PM, nickelsilver said:

With a very small but existant endshake, the shoulders of the arbor can touch both flat surfaces of the two jewels at the same time.

Wow, crystal clear, and perfect illustrations!  Thanks a ton!

This is the kind of stuff that should go into your watch repair book (when and if you decide to write it ).

[Knebo]

On 8/15/2022 at 11:40 PM, JohnR725 said:

 

That is super useful! Thanks.

Maybe a silly question, but how to you MEASURE the current endshake ? (I can see how I could measure the CHANGE when I adjust the jewel depth with a micrometric press; but how do I find out the initial value)

[Neverenoughwatches]

1 hour ago, Knebo said:

That is super useful! Thanks.

Maybe a silly question, but how to you MEASURE the current endshake ? (I can see how I could measure the CHANGE when I adjust the jewel depth with a micrometric press; but how do I find out the initial value)

Not easy unless you have the right tools. Probably not practical or convenient for a repairer to actually try to measure it, its not like we are building a watch from scratch. . I'm sure its instinct based on experience, for us as amateurs it cant be much else. For pros for i imagine almost all watches its just a natural ability to know when its right.

[VWatchie]

What I did to get a feel for it was first to set my micrometre to 2/100mm and have a look at the gap in my stereo microscope. Next I looked at a lone mounted wheel to compare from memory. In this way, I think that I can now visually determine the distance quite precisely.

[Neverenoughwatches]

On 8/15/2022 at 12:13 AM, VWatchie said:

I just serviced a Vostok cal. 2431 in overall very good condition. I've had it since it was new (2014) and it was performing reasonably well before servicing it. It's identical to the cal 2416B but has been modified (by Vostok) to make the hour hand rotate 360 degrees only once per day (24h movement). Anyway, in the process, I carefully adjusted the end-shake to be as small as possible while carefully making sure all wheels were spinning freely.

After the service, the amplitude became extremely erratic. The dial-up position jumps between 300 degrees and 240 degrees in what I perceive to be a random manner. I can't detect any particular pattern for the variations. It appears that the amount of power reaching the escapement varies randomly.

Now, Vostok movements aren't exactly known for their precise tolerances. Could it be that these movements don't handle super fine end-shake as well as modern Swiss movements? Perhaps I should try to increase the end-shake? Of course, and as always, the problem could be something entirely different, but as I did adjust the end-shake to be as small as possible (0.01 - 0.02 mm) that's what I suspect could be the problem, no?

I guess my question is, can end-shake and/or side-shake be too small even when all wheels are perfectly free (Libre Sans Jeu)?

I missed this thread when it was originally posted, so this made some interesting reading for me regarding tolerances and the binding up of pivots in jewels.  Something i see not mentioned here would be the addition of another component that can affect the side shake tolerance.  And that would be the lubrication between the pivot and jewel. The oil has a consistency whatever that is for the oil used for a particular bearing. Having a thickness to it means some of the tolerance is taken up. The oil's  molecule beads whatever has now become the bearings and replaced what was once just a space. A perceived measured minimum tolerance of a dry bearing and pivot has now become less so after lubrication. One reason a thicker oil will slow the movement of components within bearings. I thinner oil may well have increased your amplitude watchie, thinner oil, small molecule beads, more sideshake tolerance Thoughts anyone.

Edited July 21, 2023 by Neverenoughwatches

[VWatchie]

End-shake is an interesting and crucial topic, and reminds us quite a bit about shoe sizes. In my experience, too little or too much end-shake is bad, but as it is difficult to get it to be perfect it's better with too much end-shake than too little. The same goes for shoes. It's better with one size too large than one size too small, but any extremes should be avoided, as it will impede your ability to walk and run.

[Neverenoughwatches]

11 minutes ago, VWatchie said:

End-shake is an interesting and crucial topic, and reminds us quite a bit about shoe sizes. In my experience, too little or too much end-shake is bad, but as it is difficult to get it to be perfect it's better with too much end-shake than too little. The same goes for shoes. It's better with one size too large than one size too small, but any extremes should be avoided, as it will impede your ability to walk and run.

I agree, the two extremes also have an impact on our silly walks. I think you've just answered the final puzzle in perfecting mine.  IIIIIIIIIIIII thank you watchie X

https://www.watchrepairtalk.com/topic/15356-why-does-the-rate-start-to-rush-when-i-work-out/?do=findComment&comment=224917

Edited July 21, 2023 by Neverenoughwatches

[Knebo]

27 minutes ago, VWatchie said:

What I did to get a feel for it was first to set my micrometre to 2/100mm and have a look at the gap in my stereo microscope. Next I looked at a lone mounted wheel to compare from memory. In this way, I think that I can now visually determine the distance quite precisely.

That's smart and practical. 

Unfortunately, it would require another big investment for me : a stereo microscope.

[Neverenoughwatches]

1 minute ago, Knebo said:

That's smart and practical. 

Unfortunately, it would require another big investment for me : a stereo microscope.

Any choice of  magnification will work. Its just about differentiating one depth of endshake or one width of sideshake  to another

[VWatchie]

Came to think of it. How to determine side-shake was really a mystery to me until I saw the following video and it hardly requires any magnification.

 

Edited July 21, 2023 by VWatchie

[Knebo]

3 hours ago, VWatchie said:

Came to think of it. How to determine side-shake was really a mystery to me until I saw the following video and it hardly requires any magnification.

 

Yes, for side shake it get it with that method. It's easier to grasp as you see the pivot in relation to the jewel hole.

But vertical endshake seems much harder to observe. 

[Neverenoughwatches]

11 minutes ago, Knebo said:

Yes, for side shake it get it with that method. It's easier to grasp as you see the pivot in relation to the jewel hole.

But vertical endshake seems much harder to observe. 

Its not harder to observe with adequate magnification. More mag equals more observation of focus points. What you can see at x40 you might not see at x20 , what you can see at x20 you might not see at x10 and so on.