Servicing a VR3235 movement (my first attempt at servicing a movement).

[WatchSmith.US]

Turning the brass nut clockwise lowers the bridge, and turning it anti clockwise raises it. IIRC it’s 0.1mm per tooth on the nut which you can turn with a screwdriver tip.

 

[dogwood]

Turning the brass nut clockwise lowers the bridge, and turning it anti clockwise raises it. IIRC it’s 0.1mm per tooth on the nut which you can turn with a screwdriver tip.

How much force should I need to apply to get the height regulating nut to rotate? I’ve tried quite a lot of force and I’m not getting any movement (clockwise or counter-clockwise).

i have the feeling that this might be a difference in design between the gen and VR movement. On the VR, the “nut” doesn’t really look like a nut, it looks like a nut fused to a threaded tube which is then interference fit into a hole in the main plate. On the gen, the nut is just that, a nut which can move up and down on a threaded tube.

pic of the VR regulation “nut” (note the scratches are from me trying to get it to rotate).



The gen regulation nut (part 3135-5120) looks like a flat brass knurled nut.

 

[WatchSmith.US]

Assuming it’s a real nut, it should turn fairly easily.

 

[dogwood]

Assuming it’s a real nut, it should turn fairly easily.

Hmmm… good to know. I’m thinking it’s much more likely to be a fake nut then.

Also, to check my end shake theory, I decided to try using aluminum foil shims as a test. I made two of them, each 0.04mm thick and placed them under the ends of the balance bridge.
with these shims in place there’s now a measurable amount of endshake in the balance wheel, and if I apply light pressure to the balance bridge shock setting with Pegwood, the balance wheel no longer stops (it did before the shims).

But strangely, when I take a slow-motion video of the balance in motion it still only appears to rotate about 90 degrees in each direction. This makes me think that maybe my new pallet fork isn’t giving enough of an impulse to the balance wheel on each tick, or maybe the guard pin is too long and is rubbing on the safety roller causing friction that’s reducing the energy that the pallet fork is delivering to the balance.

I hate to keep bugging the experts (shamelessly tagging oascom ) but I'd love to hear what they might think is going on.

 

[WatchSmith.US]

Without a timegrapher I would be shooting in the dark as far as amplitude. I suppose a slomo video could work but accuracy is the name of the game.

I wish I had better suggestions for you.

 

[dogwood]

Without a timegrapher I would be shooting in the dark as far as amplitude. I suppose a slomo video could work but accuracy is the name of the game.

I wish I had better suggestions for you.

Thanks so much for the help and suggestion of the timegrapher. Also, thanks for the help / feedback / advice in general. The RWI community has been an incredible support on my learning adventure... So, I put the movement on my timegrapher, but the results are so inconsistent that TG0.5.0 wasn't able to even calculate an amplitude.

My gut feeling is that the problem is now in the pallet fork... maybe it's rubbing between the guard pin, or maybe it's compatibility with the escape wheel (draw to lock distance?)... regardless, there's lots more investigation (and learning) in my future!

 

[All_Heart]

Oh boy... update... One of the pallet stones came out when I was trying to apply Moebius 9415 grease to it. I literally just touched the stone with the oiler and the stone stuck to the oiler when I pulled it away. This tells me either the shellac was weak from the start, or I did something to degrade the shellac as part of my cleaning. My cleaning procedure was 10 minutes in the ultrasonic cleaner in a bath of naphtha (lighter fluid), then drying the parts, then another 10 minute bath in a fresh solution of naphtha in the ultrasonic cleaner. Then I removed the pallet fork and balance (the two parts with shellac secured jewels), and placed the dried parts into a 99.9% alcohol solution for another 10 minutes in the ultrasonic bath. I was pretty sure that naphtha was safe for shellac, but maybe not for that duration? Or was it the length of time I left it in the ultrasonic cleaner? Or did I just get a poorly secured pallet stone?

Some pics:

Here's the pallet fork before cleaning (I don't have a pic showing the shellac side)

Here's the pallet stone that came away stuck to the oiler:


A close up of the pallet fork showing the shellac... does the colour seem right? Or does this look like my washing solution destroyed the shellac?

Acetone, Shellite, Butane, Battery contact cleaner, Lighter fluid, are very high on hydrogen amongst hydrocarbons. They can strip the carbon material off many things e.g. oils, shellac, varnish, superglue, grease, oils etc… so these chems are a no-no on anything that burns or melts with heat easily.

Ever noticed how your greasy hands will suddenly become dry if you wash with liduid butane or lighter fluid? That’s the Hydrogen pulling the carbon from oils and evaporating it.

You can reapply shellac but will need a hot laser or a magnifying glass on a bright sunny day to melt it. It would be easier to get a replacement pallet fork if you don’t mind waiting.

Did you notice the VR3235 is a modified 3135 on the date seating side and 3235 on the rotor side? With just an extra setting wheel to change the hand sweep direction when you twist the crown?

Btw these rhinestone pickers work really well picking up watch hands and small parts and they don’t leave marks. Helps a lot if your tweezer game is off and much easier to maneuver than Rodico.

Just make sure to check the brass tip for burs before using on hands. I usually polish them before loading the tip w/ polyurethane goo.

Good luck! Its great to see someone go through the same learning experience.

 

[dogwood]

Acetone, Shellite, Butane, Battery contact cleaner, Lighter fluid, are very high on hydrogen amongst hydrocarbons. They can strip the carbon material off many things e.g. oils, shellac, varnish, superglue, grease, oils etc… so these chems are a no-no on anything that burns or melts with heat easily.

Ever noticed how your greasy hands will suddenly become dry if you wash with liduid butane or lighter fluid? That’s the Hydrogen pulling the carbon from oils and evaporating it.

You can reapply shellac but will need a hot laser or a magnifying glass on a bright sunny day to melt it. It would be easier to get a replacement pallet fork if you don’t mind waiting.

Did you notice the VR3235 is a modified 3135 on the date seating side and 3235 on the rotor side? With just an extra setting wheel to change the hand sweep direction when you twist the crown?

Btw these rhinestone pickers work really well picking up watch hands and small parts and they don’t leave marks. Helps a lot if your tweezer game is off and much easier to maneuver than Rodico.

Just make sure to check the brass tip for burs before using on hands. I usually polish them before loading the tip w/ polyurethane goo.

Good luck! Its great to see someone go through the same learning experience.

Thanks for the tip on the Rhinestone pickers! And for the background on the hydrogen in cleaning solvents pulling away the carbon from shellac... very useful to know.

So a bit of an update on my progress (or lack thereof). I've tried a few things to solve my extremely low amplitude issue. I'm thinking that there must be somewhere between the mainspring and the balance where there's more friction than there should be -- something that's dramatically reducing the force that the pallet fork is able to give to the balance wheel on each tick/tock. Here's a list of what I've thought of and tried to fix:

1. I thought maybe the problem was that there was too much friction in the train of wheels. To test this I first loosened the screws of the train bridge to make sure that I wasn't accidentally squeezing down on the arbors of the main train wheels too much. This didn't make a difference, but I still thought I should eliminate the possibility of friction in the main train of wheels as a possibility. So I disassembled the movement to the point where I could remove and inspect all of the train wheels. I inspected each wheel and pinion's teeth (all looked good), and I inspected the pivots on the ends of each train wheel arbor (none looked damaged / bent). So I re-assembled the main train and moved on. When re-assembled the train of wheels seemed to spin very freely when I manually pushed on any of the wheels.
2. Next I thought the problem might be that the new Clark pallet fork didn't have enough end shake... like maybe the Clark version was 0.05mm longer on the arbor / pivots so it would be held too tightly between the jewel on the pallet bridge and the jewel on the main plate (NB: neither of these jewels have been lubricated, I'm pretty sure they have to run dry). So I cut and folded an aluminum foil shim and placed it under the pallet bridge to add about 0.04mm of additional clearance for the pallet fork arbor. I don't have a good enough sense of feel for end shake measurements for the pallet fork, but when I manually manipulate the pallet fork it seems to snap from banking pin to banking pin with minimal effort. So with this change I moved on.
3. I then thought the other thing that I've messed with in my "service" was the Kif shock mount for the balance wheel on the main plate. This was my nemesis from several post back. With great reluctance I decided to try disassembling the Kif setting again to see if maybe I'd done something wrong. I got the setting out without losing any parts, cleaned the jewels and managed to re-assemble the setting in under 30 mins (PROGRESS!!!!).
4. To test all of these adjustments I re-installed the balance (including the shims from the post above)... and guess what? The same low (90 degrees total) movement on the balance wheel even with quite a lot of energy in the mainspring. I've tried rotating the movement to see if I get better amplitude in a different orientation (I thought this might give me a clue as to where the problem might be hiding), but it seems to be just as muted in each orientation.

I can't help by think that something somewhere has to be "fouling" the movement. I'm not sure where though. The only part of the movement I haven't disassembled (as part of this debugging run) has been the keyless works, and it doesn't look like anything there could be adding friction to the mainspring -> train of wheels -> escapement -> balance system. I don't have the motion works re-assembled, so there's no additional friction coming from the cannon pinion / dial side.

I'm honestly at a bit of a loss for where to try working on next. If anybody has any suggestions I'd love to hear them. At this point I'm pretty close to the Microsoft solution: reboot -- i.e. take the whole thing to pieces, clean it again, and re-assemble to see if I get a different result this time.

 

[All_Heart]

Did you disassemble the barrel and the main spring? At full wind the amplitude reaches up to 300+ if the mainspring is anchored on the barrel groove, otherwise, you would have lost significant amplitude by the time it unravels half-way.

Here is a video and at 11:50 minute mark he re-installs the main spring. Curious if you did the same.

https://youtu.be/N9ee6qqyPhg

 

[Nikz19]

Did you disassemble the barrel and the main spring? At full wind the amplitude reaches up to 300+ if the mainspring is anchored on the barrel groove, otherwise, you would have lost significant amplitude by the time it unravels half-way.

Here is a video and at 11:50 minute mark he re-installs the main spring. Curious if you did the same.

https://youtu.be/N9ee6qqyPhg

This is not the case-
A 3135 is an automatic movement. The mainspring will slip through the whole barrel wall once it’s fully wound. There is no groove where the mainspring tip should stay, only manual winding movements have that.

 

[dogwood]

Update: I've tried a few more things...

1. I figured it was possible that I'd messed up something in the re-assembly of the watch, and since I'd partially assembled / disassembled it a half dozen times, I figured it would be worth taking it all to pieces again and cleaning everything off again in naphtha. I didn't bother running it through the ultrasonic cleaner, just a dip in naphtha to remove any lubricants / dirt from my work. This went WAY faster than my previous tear down and cleaning (I'm getting lots of practice with this movement!)... but upon re-assembly (which was also WAY faster than the first time, and I was able to do it basically from memory at this point), I'm still getting terrible low amplitude on the balance wheel. Without the pallet fork in, the train of wheels spins completely freely when it even touch the winding stem... so I don't think the source the the problem is in the barrel or train of wheels. It really feels like the problem is lurking in the escapement.
2. Then I had the idea that perhaps somehow my hairspring has gotten itself magnetized. If the hairspring got magnetized that would have the effect of increasing the spring constant (i.e. making it stiffer), but also would introduce a huge energy drain since moving a conductor through a magnetic field induces eddy currents which result in energy losses. So I put the whole assembled movement onto my de-magnetizer... but unfortunately that didn't solve anything for me.
3. Next I noticed that with a very low amount of wind on the mainspring, the balance wheel would stop in a certain position after I nudged it to start it moving. I hoped that the place where it stopped would lead me towards the source of the resistance. I can't see anything obvious, but I did notice, that the balance always stops with the pallet fork in this position (see pic). I'm starting to wonder if the Clark pallet fork is just somehow not compatible with the VR3235 movement (or maybe the pallet fork I bought is a dud). Maybe its geometry is every so slightly different and that's causing the lock and release forces to be off.

Regardless, I'm learning a lot, even if I'm not solving anything (yet)... and I'm getting WAY better at manipulating little parts. Also, I managed to disassemble and re-assemble the Kif shock setting in the main plate without too much trouble this time. So that's a small win for today at least. As always, if anybody has any suggestions I'd love to hear them. This low amplitude (like 90 degrees) bug is turning into a really hard nut to crack!

 

[dogwood]

Update: I haven't made any real progress, but I have discovered something that I *think* might be causing my amplitude problem. I think there's some kind of mismatch between my Clark pallet fork and my VR escape wheel. The reason for this is that I'm getting inconsistent locking distance on the entry stone (see pics). I remember in Mark Lovik's course he mentioned a rule of thumb that the locking distance should be about 1/6th of the impulse face width. So based on the pics below, it appears that my locking distance on the entry stone is too big most of the time, but is also very small about 10% of the time.

These pics were both single frames from a slo-mo video that I shot. The first pic shows the "normal" locking distance on the entry stone. It kinda looks like my locking distance is closer to 1/4 of even 1/3 of the impulse face width. Could this explain my low amplitude? Perhaps too much energy is being expended by the balance wheel to "unlock" the entry stone?

This second picture shows the amount of locking distance I'm getting about 10% of the time (perhaps this is due to a few teeth on the escape wheel having slightly different geometry and I'm only seeing it each time the escape wheel makes a complete revolution. Interestingly, there are also parts of the slo-mo video where the entry stone locks in this position, and then after a pause slides into the position shown in the image above.


Here's a pic of the exit stone (I flipped the movement 180 degrees, to get better light on this stone)... but interestingly it looks like the escape wheel tooth isn't even locking on the locking face, it looks like it kind of just jams on the corner of the impulse face of the exit jewel.
I kinda wish I had another VR3235 movement on hand so I could check to see what a "healthy" VR3235 movement's pallet fork / escape wheel locking behaviour looks like. At the moment, I' only have VS3235 and N4130 movements in the rest of my reps. I'm kinda tempted to order a VR3235 from AliExpress (C$200) just to see what this looks like. However, before I pull the trigger on a spare movement, are there any watchsmiths with more experience than me (i.e. everyone) that could chime in on the pics above?

 

[john doe iii]

Update: I haven't made any real progress, but I have discovered something that I *think* might be causing my amplitude problem. I think there's some kind of mismatch between my Clark pallet fork and my VR escape wheel. The reason for this is that I'm getting inconsistent locking distance on the entry stone (see pics). I remember in Mark Lovik's course he mentioned a rule of thumb that the locking distance should be about 1/6th of the impulse face width. So based on the pics below, it appears that my locking distance on the entry stone is too big most of the time, but is also very small about 10% of the time.

These pics were both single frames from a slo-mo video that I shot. The first pic shows the "normal" locking distance on the entry stone. It kinda looks like my locking distance is closer to 1/4 of even 1/3 of the impulse face width. Could this explain my low amplitude? Perhaps too much energy is being expended by the balance wheel to "unlock" the entry stone?

This second picture shows the amount of locking distance I'm getting about 10% of the time (perhaps this is due to a few teeth on the escape wheel having slightly different geometry and I'm only seeing it each time the escape wheel makes a complete revolution. Interestingly, there are also parts of the slo-mo video where the entry stone locks in this position, and then after a pause slides into the position shown in the image above.


Here's a pic of the exit stone (I flipped the movement 180 degrees, to get better light on this stone)... but interestingly it looks like the escape wheel tooth isn't even locking on the locking face, it looks like it kind of just jams on the corner of the impulse face of the exit jewel.
I kinda wish I had another VR3235 movement on hand so I could check to see what a "healthy" VR3235 movement's pallet fork / escape wheel locking behaviour looks like. At the moment, I' only have VS3235 and N4130 movements in the rest of my reps. I'm kinda tempted to order a VR3235 from AliExpress (C$200) just to see what this looks like. However, before I pull the trigger on a spare movement, are there any watchsmiths with more experience than me (i.e. everyone) that could chime in on the pics above?

I have a new one I’m about to swap out for a vsf. Located in Cali. PM if interested.


Sent from my iPhone using Tapatalk

 

[WatchSmith.US]

To my knowledge the VRs are vastly different than the VS3235s.

 

[dogwood]

UPDATE: Thanks for john doe iii I now have an incoming VR3235 which when it arrives I'll be able to compare it to the one I have on my bench (and, in the fullness of time, I'll be able to pull that one apart and make it not-work too). But, I've also invested (*cough* hobby expensed *cough*) in a nice digital microscope (an Andonstar 409); and WOW, being able to see what you're doing on a 10 inch LCD is a game changer -- so much easier then trying to take a picture while holding a loupe in front of an iPhone camera too.

So I've disassembled the movement, cleaned it in one-dip, and re-assembled it a few times now, and still no difference. But, now I'm a lot more confident with manipulating parts (progress!), and I'm getting more and more sure about where the problem has to be lurking (and now I might have the pics to prove it). So here are some pics from my fancy new digital microscope.

Here's the locking distance on the entry pallet stone. It looks a little bit too big -- as in the escape wheel tooth is positioned a bit too far down the locking face of the pallet jewel. It looks like the locking distance is around 1/4 of the width of the impulse surface of the pallet jewel. I've read that the ideal locking distance is around 1/6th of the impulse face width.

Here's a pic of the locking distance on the exit pallet stone. It looks a little too small -- as in the escape wheel tooth is locked right at the corner of the exit stone and is barely on the jewel's locking face at all. I'm both not sure how this will affect the timing of my watch (having variable locking distances), and how I would adjust this... could I change the locking distance by adjusting the banking pins?


Here is a nice comparison between the Clark's 3135 pallet fork, and the VR3235 pallet fork. They really do look very similar. When I was getting the big differences in locking distance on the entry and exit stones, I thought it might be because the geometry of the pallet forks was slightly different, but it really seems like they're the same.


Final super-macro shot. Here's my latest attempt at oiling a shock setting. It doesn't look like I've got the oil centered. In theory the edge of the oil should be visible as a circle centered under the cap stone, but my oil job looks kinda like a wave. Did I over oil or under oil the setting?

Many questions as always... for those answering my newb questions, thanks in advance. For those reading along for the ride, I hope you enjoy the higher quality pics... more to come.

 

[derjenigewelcher]

How old is the watch that it needed servicing already?

It does not need service. Its all about beeing a nerd and wanna improve skills

 

[WatchSmith.US]

In some clone movements I have serviced the oil under the cap jewel doesn’t always make a uniform circle. Oftentimes it disperses unevenly which I attribute to slight inaccuracies in manufacturing. I have not experienced this in genuine movements or jewels. As long as the pivot sits in oil you should be ok, but the service interval might be sooner than normal.

 

[dogwood]

In some clone movements I have serviced the oil under the cap jewel doesn’t always make a uniform circle. Oftentimes it disperses unevenly which I attribute to slight inaccuracies in manufacturing. I have not experienced this in genuine movements or jewels. As long as the pivot sits in oil you should be ok, but the service interval might be sooner than normal.

Awesome. That’s really great to hear. I’ll be pulling this one to pieces again since it’s still not working, so I’ll get plenty of practice cleaning and oiling these shock settings. There really is no substitute for time at the bench fiddling with parts.

 

[SS72]

dogwood megga respect, mate. I keep trying and, unfortunately, not succeeding. Your progress motivates me. Great pics, too!

That tip from Triplock about the cling wrap is gold. Mt balance shock clip is somewhere (possibly inside the movement) other than where it's supposed to be. This would have saved me and will in future. Won't help me with hands, though - I keep messing those up, and also the occasional second hand post.

 

[dogwood]

So a little update on progress. I put the VR3235 movement onto my Timegrapher, and using tg 0.5.0 software, I was able to capture what the tic/toc of the watch sounds like... That's the picture below:

In a word this is ugly... in two words, it's very ugly. In theory the tic/toc graph should look something like this (the VS3235 in one of my subs):

I'm still pretty new at this, by my understanding is that the three peaks of the tic/toc graph are what is used to calculate the amplitude of the movement... and furthermore, there should be three clean peaks for each tic and toc. I'm new at this, but my guess is that the three peaks are:

1. Impulse pin hitting the horn of the pallet fork to unlock the pallet stone.
2. Escape wheel engages with the impulse face of the pallet stone that was just unlocked.
3. Escape wheel locks against the locking face of the other pallet stone (this should be the loudest).

The time it takes to go from the first peak to the third peak combined with the lift angle is how the amplitude is calculated (I think), so that would explain why I'm not getting an amplitude number in TG because there simply isn't an obvious first/second/third part of the beat.

If my understanding of the three sounds of each beat are correct, then it implies that I'm just not getting good distinctions in the movement of the pallet fork, balance wheel, and escape wheel. If anybody has any insight, I'd love to hear it.