New version VR3135 problems

[Konstantinos_Sv]

Hello guys I face a lot of trouble with the new version of vr(new version I mean the version with decorated balance bridge to look like gen with the 2 legs of time control and beat error control are on the opposite side.... I services a lot old version vr and vs and another movements like eta seiko ... never had that issue that ill explain now.i had to check the automatic winding parts (rotor and the bridge) on a new VR after I remove only those parts(automatic bridge and rotor) and when I put them in place back the movement numbers on time grapher went crazy.... especially the beat error.... time grapher read dots everywhere and when I removed again the automatic bridge and check the numbers on timegrapher everything was fine.... really good numbers... so then I put again the automatic rotor and again everything went crazy... and then I try to make the beat error with the automatic bridge on.... and the movement was unstable... a half of the times I put the watch on the time graphed I saw 0 beat error and the other half times I saw 4 beat error with dots everywhere... I tried everything I could... and nothing can make the movement stable.... the problem is not the time is not that I read bad numbers but the unstable thing..... I mean unstable becouse when I put the watch on timegrapher for 1 minute it has 0 beat error after 1 minute beat error went crazy and after 2 minutes again 0 beat error and then after some time again beat error went crazy.. so even if I get the beat error to zero and I read zero .. after some time again it goes crazy... and the worst thing? I remove on another new VR version movement the automatic bridge and the movement from 290 amplitude went to 352! and the beat error again start to be unstable..... does anybody knows why this is happening? the movements went crazy only with the automatic bridge removed... I never touch or did anything else!

 

[dogwood]

Pics?

 

[dogwood]

Also… what kind of time grapher are you using, if you’re using a computer and running TG0.5.0 can you show a picture of what the beats sound like… eg this:

 

[dogwood]

I had a similar issue on a VS3235. I documented my issue here:

Deus Ex Machina: A tale of two VSF 3235 balances...

I have a beautiful VSF Cookie Monster in my collect. I love it. But it's by far the worst at keeping time. I've tried to regulate it half a dozen times, but invariably, after a few days it'll be running several minutes fast. Worse, when I recently put it on my timegrapher and left it in one...

forum.replica-watch.info

TL;DR: there’s a lot of variability in the quality of rep movement parts. That’s why a vr3235 costs $100 and a gen costs $5,000.

If you have some spare vr parts on your bench, try swapping them in one by one to see if you can isolate the issue.

 

[Konstantinos_Sv]

thanks for the answer bro..
I deleted a lot of pictures and I found only 2 of them..due to a lot of space on my phone... think something a lot worse than those pictures... I mean dots everywhere... even if I get 0 beat error after some time on the time grapher its starts again to go like crazy.. dots everywhere 3-9 beat error... and 0 beat error with dots everywhere... and some times the timegrapher couldn't read anything.... and I never did anything to the watch only remove the automatic bridge !!!

 

[Konstantinos_Sv]

I found another one... I just remove the automatic bridge and the amplitude went from 290 to 350+ and the beat error fails....

 

[derjenigewelcher]

351 is almost 360 degrees....wow.

 

[dogwood]

It’s very likely a calculation error in the time grapher’s algorithm for identifying the correct beginning / end of each tick and tock of the watch movement. Each tick and tock of the movement is actually three distinct sounds:

Red: the impulse pin hits the pallet fork horn and unlocks the pallet stone from the escape wheel.

Green: the impulse face of the pallet fork jewel engages with the escape wheel.

Blue: the opposite pallet stone locks against the escape wheel.



The time grapher software has to run an algorithm to identify each of these peaks. The period of time between the first (red) and last (blue) peaks combined with the lift angle and the movement frequency (28800 bph) is used to calculate the amplitude.

If the algorithm doesn’t correctly ID the peaks then the amplitude that gets calculated can be complete garbage.

The same is true for beat error. Best error is the difference in time between the first (red) peaks against when they should theoretically occurs based on the movement’s frequency (28800 bph). But again if the software misidentifies the peaks then the beat error number that gets displayed can be completely wrong. For example in the image above, there’s an extra peak in the “tock” sound profile. But the software ignored it and ID’d the soft initial peak as the start of the tock not the green peak.

This is why I prefer using a computer instead of a simple time grapher — you get to see WAY more information. Also it’s super cheap to use a computer. The software is free and the usb microphone and amplifier cost $30 on AliExpress.

 

[Rman]

It’s very likely a calculation error in the time grapher’s algorithm for identifying the correct beginning / end of each tick and tock of the watch movement. Each tick and tock of the movement is actually three distinct sounds:

Red: the impulse pin hits the pallet fork horn and unlocks the pallet stone from the escape wheel.

Green: the impulse face of the pallet fork jewel engages with the escape wheel.

Blue: the opposite pallet stone locks against the escape wheel.



The time grapher software has to run an algorithm to identify each of these peaks. The period of time between the first (red) and last (blue) peaks combined with the lift angle and the movement frequency (28800 bph) is used to calculate the amplitude.

If the algorithm doesn’t correctly ID the peaks then the amplitude that gets calculated can be complete garbage.

The same is true for beat error. Best error is the difference in time between the first (red) peaks against when they should theoretically occurs based on the movement’s frequency (28800 bph). But again if the software misidentifies the peaks then the beat error number that gets displayed can be completely wrong. For example in the image above, there’s an extra peak in the “tock” sound profile. But the software ignored it and ID’d the soft initial peak as the start of the tock not the green peak.

This is why I prefer using a computer instead of a simple time grapher — you get to see WAY more information. Also it’s super cheap to use a computer. The software is free and the usb microphone and amplifier cost $30 on AliExpress.

You got AliExpress link for hardware and link for software?

 

[dogwood]

You got AliExpress link for hardware and link for software?

Microphone:
C$ 50.03 | Computer Calibration Instrument TGB Mechanical Watch Pocket Watch Dimension Time 1000 Calibration Test Clock Repair Wire Machine

https://a.aliexpress.com/_mrMCem0

Software:

Tg timegrapher - download page

Tg is an open source timegrapher. The tg program is used to assess the health of mechanical watch movements, and to regulate mechanical watches.

tg.ciovil.li

 

[KJ2020]

Try demagnetizing all the parts. Also One Dip the balance and relube the end cap jewel.

 

[40mm]

I had something similar to this where the end curve on the balance spring was not bent correctly and any attempt to correct the movement timing would cause my 1900 timegrapher to display dots as you describe and give crazy changes in beat error. In my case the balance spring was rubbing hard on the adjustment arm pins instead of passing evenly between them.

Hopefully this is not your problem as it is a finicky fix that requires removing the balance wheel.

 

[centrum]

I'm sorry for the trouble you're going through with a VR3135. My repsmith has told me once before that the new VR3135 with added modification is not as reliable as the one without. I hope you get it sorted out and have a reliable movement.

 

[Konstantinos_Sv]

thanks for the help guys ill try first the clean method and re oil... if it doesn't work ill have to remove the balance wheel and I think yes its good to buy a computer to read clear what's happening...

 

[KJ2020]

I had something similar to this where the end curve on the balance spring was not bent correctly and any attempt to correct the movement timing would cause my 1900 timegrapher to display dots as you describe and give crazy changes in beat error. In my case the balance spring was rubbing hard on the adjustment arm pins instead of passing evenly between them.

Hopefully this is not your problem as it is a finicky fix that requires removing the balance wheel.

This is a very good point. I had one come across my bench that was exhibiting similar instability while attempting to regulate it.

I found that the hairspring had a kink in it near the anchor stud. It was a lot worse than in the first pic below, and probably happened during a slip of the tool used to regulate it (how else?).

I smoothed it out to where you see it in the first pic, then the hairspring end came loose from the stud. The remaining adhesive just crumbled into moist dust. Not sure what rep factories use, it isn't shellac. I've done repairs like this before, slow cure epoxy works perfectly.

After the epoxy cured I was able to improve the curve further and now the movement runs well and is properly adjustable.

I have come to dislike this current configuration of VR balances and am going to buy a bunch of the old style ones very soon, maybe today. They are down to $70 from my favorite Ali Express vendor OULIGUAN. That's a steal, hardly more than a Seagull ETA. No telling how much longer these are going to be available.

US $70.00 | Chinese Super 3135 automatic mechanical movement blue balance wheel Men's watch movement VR NOOB VS Provide assembled watches

https://a.aliexpress.com/_mLgysDA

 

[Konstantinos_Sv]

Hello again guys... after removing the decorated hand to look like gen... the movement starts to get more stable but not for a long time.. after 5 minutes everything went crazy again I don't know what's wrong... even if I read good numbers and then I thought yes I did it it is working properly after some time the movement going crazy... this is the photos after removing this ridiculous decoration... and then after 5 minutes this is what happened without touching anything.. and this is how the hairspring is I know its look crazy that from a healthy timegrapher numbers for no reason going like a problematic movement and that's why I start this thread.... I don't know anything.... ill try the one dip I saw the hair spring is not clear so its my last hope before I change the hairspring with SH hairspring becouse I tried to add SH balance bridge and balance from SH to this VR and it didn't work

 

[Konstantinos_Sv]

I hope the sh hairspring fits there.

 

[KJ2020]

Hello again guys... after removing the decorated hand to look like gen... the movement starts to get more stable but not for a long time.. after 5 minutes everything went crazy again I don't know what's wrong... even if I read good numbers and then I thought yes I did it it is working properly after some time the movement going crazy... this is the photos after removing this ridiculous decoration... and then after 5 minutes this is what happened without touching anything.. and this is how the hairspring is I know its look crazy that from a healthy timegrapher numbers for no reason going like a problematic movement and that's why I start this thread.... I don't know anything.... ill try the one dip I saw the hair spring is not clear so its my last hope before I change the hairspring with SH hairspring becouse I tried to add SH balance bridge and balance from SH to this VR and it didn't work

If this was mine to work on I would straighten out this abnormal bend in the hairspring and see what happens. See my last post above.

I also take the deco arm off.

 

[Konstantinos_Sv]

yes I remove the deco off no reason for that useless thing... ill try to straighten that thing after the cleaning to be sure that is not due to the dirt ill write how it will go... thanks for the answer bro

 

[Konstantinos_Sv]

Tomorrow ill post some pictures about the results... I straighten the hairspring as Kj2020 said and the beat error looks like 80% of the problem solved..... the funny thing is how a movement lost his stability when you don't touch nothing else but the automatic bridge... and how this movement gets his stability when you straight the hairspring!... movement with that bend on hairspring before I remove the automatic bridge worked like perfect and after removing the automatic bridge I need to do work on hairspring ... where the logic goes.. haha this movement is ..... thanks a lot all the guys here for the help ill post the pictures tomorrow and I hope in this thread we will post issues with this new VR version problems.... thanks also for the link KJ2020 for older VR ill order some of them