The Watch Site banner
21 - 40 of 70 Posts

·
Super Moderator
Joined
·
3,226 Posts
I just had a thought with another possible solution that I hadn't considered previously. Polyurethane casting may be an option. I can do it on my desk with very little tooling and it should have the detail required for these parts.

I'll acquire some small amounts of the casting medium and some mould making material and try it out. I need some for another project and testing anyway.

Adrian.
I'll be interested to learn how PU does in such a small part, as regards strength. Could it be something that grips very well on the lower wheel's post, but would still expand a little if need-be, so that sizing isn't as critical? How does PU take to cleaning solutions? Looking forward to your impressions after experimenting with it.
 

·
Registered
Joined
·
614 Posts
I like the design of the corrector wheel on the ETA movement, it is a lot better. I think PE would be able to do something like that, laser cutting, depends on the machine. It's certainly much more robust. I'd like to do away with the tension spring around the post, an elegant solution would do that. Also with the post, I'd like to thread it so that it is just held onto the plate with a counter sunk screw, that way it is easier to work on and repair later if required.

I think if we went with something like the ETA corrector that you've shown there we could make a huge improvement on the function.

I was just thinking, the guy that was laser cutting the 4006 springs, he worked for a laser cutting company in the UK that specialised in the micro cutting work and he was able to do finer work than the 4006 spring. He sent me a bookmark that was pretty impressive that was of thicker material. Is he still around? last I heard he was retiring. I helped him out a little with that project in the early stages.

I did a cast tonight of the ring, and I'd say it's been a success. This was just thrown together very quickly to find out if the idea was viable. The ring looks fantastic but I got a bubble on one of the tangs. Here's an image -



So I've got some ideas on how to improve that. Basically with this I just dropped the part in the top of the silicone and let it stay there with surface tension. What I will probably do on the next one is make the silicone mould by attaching the part to the bottom of the vessel and pouring the silicone on top of it, which should get a better and flatter result. In person the form of the part is actually perfect. I really can't complain besides the bubble. It's hard to photograph with the poor lighting here.

As for dimensional accuracy, I wont be sure until I've done more testing. I'm sure there's some generic data on it around. As for cleaning, not too sure. From memory acetone and alcohol will eat it.

Here's the mould -

 

·
Registered
Joined
·
2,201 Posts
Discussion Starter · #24 ·
Excellent guys, great to see some action here!!!

@ Adrian: many thanks for offering to draw up the dxf file. That will certainly be most helpful with the tasks to come.

2. The cast wheel looks great, it would be very interesting to see how it performs (a) regarding slipping and (b) in the long run. Unfortunately, I do not (yet) have a metal corrector but lets see how things on Yahoo.jp turn out. Alternatively, we could procure a pre July 1968 5606 which should have the metal corrector wheel and use this part.

3. ANTHONY has access to laser cutting equipment!! I do however not know in which scale the equipment would work.

@ Noah: great ideas regarding the shape of the corrector wheel - I must admit however that I have not understood fully how this would help us? Would it be an improvement in design (make the mechanism more fail safe) or would it simplify manufacturing?

Another idea (i think we did discuss it early on in the project) after seeing the ETA parts: how about manufacturing a wheel that has only TWO teeth? it would be perfectly fine from a functional point of view but I can see implications that would simplify manufacturing a lot, consider the following steps on a sheet of metal:

1. drill holes for the center post
2. mill out the inner portion of the teeth, which in this case can be done in one stroke (the portion that forms the teeth being perpendicular to the milling axis)
3. form the outer shape of the wheel by means of e. g. PHETCH


On another note, guys: I think it would be a good idea to start individual posts for the different solution approaches we are following up on to keep discussion more flowing, what would you think?

Or do you prefer discussing all ideas/options in a single thread?

Once more: thanks for the great ideas and work! :rock:

Best
Hermann
 

·
Super Moderator
Joined
·
3,226 Posts
@ Noah: great ideas regarding the shape of the corrector wheel - I must admit however that I have not understood fully how this would help us? Would it be an improvement in design (make the mechanism more fail safe) or would it simplify manufacturing?

Another idea (i think we did discuss it early on in the project) after seeing the ETA parts: how about manufacturing a wheel that has only TWO teeth? it would be perfectly fine from a functional point of view but I can see implications that would simplify manufacturing a lot, consider the following steps on a sheet of metal:
I'm not set on changing the design, but it might be something to consider if we have the luxury. I think since we are changing the fundamental part anyways (making out of metal vs. plastic), then we are already assuming we are not beholden to the original in one key way. A simplified equation (relationships are not equivalent to real life) might be "material property x surface area x tension = resistance to slippage". So if we modify any of the first three elements, the resistance to slippage will be changed. Since I'm not an engineer and cannot calculate the forces of the day and date jumpers (nor of the corrector gear on its post), then I think we will have to experiment with solutions in a real world setting before knowing we have the key to the problem.
I'd caution against changing the number of teeth until we could experiment with how that would work. I thought I could get away with 10 teeth or 4 teeth in the 8-teeth 5619, and found that the travel distance of the wheel is linked to the movement of the date/date dial (or in my case sub-hour hand), and how far that dial/wheel must travel before the jumper takes over and locks it into the next slot. I'm just not sure yet if the 5606 would work with 2 teeth, enough so that I'd want to experiment before deciding to go that route in manufacturing.

On another note, guys: I think it would be a good idea to start individual posts for the different solution approaches we are following up on to keep discussion more flowing, what would you think?

Or do you prefer discussing all ideas/options in a single thread?

Once more: thanks for the great ideas and work! :rock:

Best
Hermann
Yes, I'd be willing to try that approach. Since we may have to discuss two methods against each other, perhaps numbering or lettering them might make it easy to refer to them in discussion. If there is any type of thread modification your account does not allow, let me know and I'll see what I can do/get done.
Excellent guys, great to see some action here!!!

@ Adrian: many thanks for offering to draw up the dxf file. That will certainly be most helpful with the tasks to come.
Yes, I'm glad you'll be able to do that Adrian. On that note, I was looking at Hermann's rendering, and it seems like there is a typo on the inner diameter of the wheel. Shouldn't it be closer to 1.07mm, not 1.25mm? That would also change the thickness of the lower ring part, too. I'm not exactly sure on 1.07, I will need to re-measure the post, and then determine how much of a friction fit we want (if the post is 1.07, do we size the wheel at 1.07, or 1.06, or 1.065, etc.? I think this is the purpose of the split in the remanufactured YJ wheel- it would allow a friction fit that wasn't as critical on initial size of hole, and the necessary +/- tolerance in manufacturing.
 

·
Registered
Joined
·
2,201 Posts
Discussion Starter · #26 ·
Two comments Noah:

1. the measurement of the ID might not be too accurate - in fact, I did not measure it at all. I determined the OD of the wheel (ring, w/o teeth) to be 1.79 mm and the wall thickness to be 0.27, which is how I ended up with 1.25 mm for the ID. That measurement may well contain some inaccuracy, so I would definitely appreciate more (and hopefully more accurate) measurements as a comparison :)

2. Regarding numbers of teeth for the 56xx (sans 5619) corrector: I would think for a standard calibre 56 movement, a single tooth on the wheel would be enough, at least that is my experience with (often not fully functional) glued-together plastic corrector wheels. On these, the height clearance would sometimes not allow all teeth to contact the gear for date QS, sometimes only one tooth would be fully functional - and enough to move the date disk. I do think that this will be fundamentially different on the 5619, where probably two sequential teeth are required to fully move the ratcheting mechanism of the (2nd) 12h-hand.

Best
Hermann
 

·
Registered
Joined
·
614 Posts
I made quite a nice resin cast part today which was a perfect cast.

I fitted the part to the rocker arm and tried it out, unfortunately a fail. The star wheel still slips. It's possible I may have been a shade under size in the thickness of the part.. I did manage to break a finger off but it was otherwise OK.

I'm going to get some epoxy as well (much harder part) to cast this from and see what result I get.

My opinion on this is that in any sort of solution, I think we are going to need to bond the star wheel to the gear underneath for any sort of reliable operation. I don't see it as realistic to get a whole assembly with the gear made in any sort of way that you could get an ROI from it.

To add to this, the enjoyment level of messing around with this in general was about a 2 out of 10. This is fairly important in any sort of solution.

I'd strongly suggest we also consider looking at the post, and seeing if we can thread the tube section of the post and replace the staking operation with a countersunk screw. I found that the staking was a bit "interesting" as we say here, and also not entirely repeatable.

 

·
Registered
Joined
·
2,201 Posts
Discussion Starter · #28 ·
Many thanks for the update, Adrian!

A couple of comments:

A. the template you are using for casting is a plastic wheel, correct? Is it an intact one or a glued one? Also, what caused the finger to break off? Is the material not strong enough for standard operation or did you deliberately try "incorrect" operation?

B. I absolutely agree that omiting the staking activity would be highly preferable. It would simplify repair work dramatically. On another note, I would like to briefly come back to the goals we set for ourselves - and we agreed that the corrector wheel per se would be our primary goal. I think it is critical that we stick to that, however, if you have any good ideas on how to modify the post/corrector plate that would be great, too :)

C. Re bonding of the part: again I would like to come back to the goals we set - and that we would aim for full functionality including protection of the calendar wheels during incorrect QS operation requiring slippage of the corrector wheel. I do therefore think that bonding the wheel to the post does not represent a good solution...

Best
Hermann
 

·
Registered
Joined
·
614 Posts
I broke the finger with rough handling. I'm going to try casting it in epoxy maybe later in the week, as epoxy has a much higher shore strength.

The casting impression was made from a wheel that I've joined back together, I don't have any good ones that I'm prepared to remove right now.

I made another one later and tested it, this one was absolutely perfect. I also lightly bent the spring so the would press down more on the gear -



This gear still slipped. It would occasionally bite and flick the date wheel (the one with the most amount of torque on it), but was not repeatable.

I might also get a copper dial washer to put under there and see if that increases the tension, or some other sort of washer.

I'm happy to make more and send them to you guys to experiment with. I think I might have some more success if I took an impression from a metal one, if anyone has one around.

I understand what you're saying about the goals, and agree, I'm just thinking that perhaps from a repeatability point of view, we might need to revisit that.

Also, I'm thinking that a simple square with a hole in it might work just as well as a complex part like this. I should be able to get a square brass profile around that size I suspect to try.

What I purchased the casting stuff for was to give this a try -



I've managed to get a perfect copy of it. Here's a video of it in action -

https://www.youtube.com/watch?v=s5J11l0JjH4

Many thanks for the update, Adrian!

A couple of comments:

A. the template you are using for casting is a plastic wheel, correct? Is it an intact one or a glued one? Also, what caused the finger to break off? Is the material not strong enough for standard operation or did you deliberately try "incorrect" operation?

B. I absolutely agree that omiting the staking activity would be highly preferable. It would simplify repair work dramatically. On another note, I would like to briefly come back to the goals we set for ourselves - and we agreed that the corrector wheel per se would be our primary goal. I think it is critical that we stick to that, however, if you have any good ideas on how to modify the post/corrector plate that would be great, too :)

C. Re bonding of the part: again I would like to come back to the goals we set - and that we would aim for full functionality including protection of the calendar wheels during incorrect QS operation requiring slippage of the corrector wheel. I do therefore think that bonding the wheel to the post does not represent a good solution...

Best
Hermann
 

·
Super Moderator
Joined
·
3,226 Posts
Adrian,
Great stuff all around! I know what you mean about a "2 out of 10", though :).

My understanding with jewelry vacuum casting procedures is that there is a 10% reduction in size from the mold to the final product- are you experiencing any shrinkage with your method/material? Also, once the mold is perfected, could you estimate how much time and materials cost it takes to make, say, 10 gears? How about 20? (Is there a direct correlation between # of pieces and effort/cost?)

I am attending the annual state watchmaker's convention in a few weeks, I plan on asking around for experience with casting in metal towards a functional goal like this, as I can definitely see benefits from molding the part vs. machining it, from an economical and DIY standpoint. But some of that will depend on how much time and effort it takes one to spit these out. If you had a list of materials you use for the process, I'd love to see what it takes to make the mold.

Regarding assembly, I haven't had too much issue for concern with the original friction-post design- it has been surprisingly robust against repeated interventions. Are you using a staking tool to take apart and re-assemble? I agree with Hermann that while it would sure be nice if the post was easily removable without special tools, it is outside the scope of our current project, plus there are definite structural limitations. The base of the rocker is very thin already, I don't know that it would take having a counter-sunk screw installed from underneath. The modifications to the existing base and post would be extensive, plus you'd likely need a custom screw made, if not a mating post too.

Very cool on the 6139 indicator gear! As long as the material is strong enough to take the repeated disengagement/re-engagement with the indicator teeth as the crown is pulled out for time-setting, then I think you have a winner on your hands! Exciting stuff.
 

·
Registered
Joined
·
614 Posts
To do these prototypes I've been using pretty basic equipment, it's a 2 part silicone for the mould and 2 part polyurethane for the part. I'm not sure of the amount of shrinkage on the part, I haven't been able to find a datasheet for that just yet. I used to deal with two part urethanes on a daily basis and they range between shrinkage/expansion, but it's mainly from heat generated during the reaction, which isn't really an issue with the size of this part.

I can also cast these in rigid epoxy which has a much higher shore hardness, which I think is a way forward.

To make 10, at present with only a single impression, I can do about one a day, maybe two. If I had multiple impressions I could make a lot more. Cost wise, the material is around $30 per 500g of PU and around the same for the silicone. So the cost of the part is minimal. Time wise to stake it apart and back together is about 10-15 minutes.

So here's what I've done, and note this is "quick and easy" way to do this -

Here's the first mould I made, this has now changed as I've put the part at the bottom of the cup now instead of the top. This improved the part quite a bit. That mould is then removed from the little tub so it can be manipulated.



So then, the urethane is poured in. On these because they are so small, I squeeze the mould so it stretches the cavity open to get the urethane in properly. I've never dealt with stuff this small before, so it's new territory for me.



Yep, I'm using a staking set. My concern it that the brass interface on the post will fatigue after a few dis assemblies. Perhaps that won't be so much of an issue on refurb parts. It's an ancient one, a FAVORITE brand, probably from 1950, but it's a decent Swiss one. I don't have any crap tools anymore which dramatically increases the enjoyment of the work, but this part still isn't fun to work with.

As for metal casting, I could see it being possible with perhaps a spin casting method, which should form the part properly. They would need to be put on sprues. I don't have the equipment or experience with this though.

Adrian,
Great stuff all around! I know what you mean about a "2 out of 10", though :).

My understanding with jewelry vacuum casting procedures is that there is a 10% reduction in size from the mold to the final product- are you experiencing any shrinkage with your method/material? Also, once the mold is perfected, could you estimate how much time and materials cost it takes to make, say, 10 gears? How about 20? (Is there a direct correlation between # of pieces and effort/cost?)

I am attending the annual state watchmaker's convention in a few weeks, I plan on asking around for experience with casting in metal towards a functional goal like this, as I can definitely see benefits from molding the part vs. machining it, from an economical and DIY standpoint. But some of that will depend on how much time and effort it takes one to spit these out. If you had a list of materials you use for the process, I'd love to see what it takes to make the mold.

Regarding assembly, I haven't had too much issue for concern with the original friction-post design- it has been surprisingly robust against repeated interventions. Are you using a staking tool to take apart and re-assemble? I agree with Hermann that while it would sure be nice if the post was easily removable without special tools, it is outside the scope of our current project, plus there are definite structural limitations. The base of the rocker is very thin already, I don't know that it would take having a counter-sunk screw installed from underneath. The modifications to the existing base and post would be extensive, plus you'd likely need a custom screw made, if not a mating post too.

Very cool on the 6139 indicator gear! As long as the material is strong enough to take the repeated disengagement/re-engagement with the indicator teeth as the crown is pulled out for time-setting, then I think you have a winner on your hands! Exciting stuff.
 

·
Registered
Joined
·
614 Posts
I've done some more thinking on the topic and I have the idea that the hole in the middle of the part could be a critical dimension.

I was previously under the impression that the spring on the post kept pressure on the gear simply by pushing down on the plastic part and then onto the metal gear, and that a fair amount of movement was required on the plastic part for it to function as specified.

I'm now thinking a much tighter hole may improve the action so it would literally be pressed quite firmly onto the gear. This would still allow slip if excessive torque was put on the system, but allow it to handle the torque required to turn the date wheel.

I'll make another impression probably on the weekend and block up the hole or reduce the diameter of the hole. I could probably achieve a better tolerance if the hole was mechanically added after.
 

·
Super Moderator
Joined
·
3,226 Posts
I've done some more thinking on the topic and I have the idea that the hole in the middle of the part could be a critical dimension.

I was previously under the impression that the spring on the post kept pressure on the gear simply by pushing down on the plastic part and then onto the metal gear, and that a fair amount of movement was required on the plastic part for it to function as specified.

I'm now thinking a much tighter hole may improve the action so it would literally be pressed quite firmly onto the gear. This would still allow slip if excessive torque was put on the system, but allow it to handle the torque required to turn the date wheel.

I'll make another impression probably on the weekend and block up the hole or reduce the diameter of the hole. I could probably achieve a better tolerance if the hole was mechanically added after.
I am of the same opinion that the inside bore of the gear determines whether or not the gear slips if blocked by the day or date dials. This is why a split wheel is no longer effective- the inside diameter is now increased/enlarge-able, causing slippage upon the post below in the face of resistance. I think we talked about the washer earlier on in the project, I am of the opinion that it is there to create just enough drag for smooth rotation/engagement of the corrector teeth when switching between correcting the day and correcting the date. It is too weak to offer any sort of frictional force between the gear and wheel below.

My big hang-up on the hole size and shape is "how tight is tight enough vs. how tight is too tight?" The size of the hole, along with material used, will have some range of sizes which will fall within these two extremes- our task will be to find this through experiment. It may be a very narrow range, which will make precise measurements and manufacturing methods critical. When I made the 5619 gear, my approach was to undersize the hole, then carefully broach it out until I could get a "good" friction fit. This "good" is measured by feel, which is not a good criteria/method for mass-production.
How much stretch is the PU material capable of? What if you made a thin sleeve and inserted it into the cracked gear before molding, so that the hole in the new gear was smaller but still concentric? I can turn one down if need be. Otherwise if we can get one of these YJ metal gears, which are ostensibly sized correctly, might they not provide a good model to mold from? Assuming that the dimensions for a good friction fit in metal hold true for other materials (perhaps a bad assumption).
 

·
Special Member
Joined
·
1,834 Posts
These are some very interesting updates and the casting method seems like it has a lot of potential. It will be interesting to see how the epoxy version holds up.

I also think that the inner hold diameter is quite critical to achieve the correct tension and ability to slip. I think that this is one reason that the YJ metal wheel has the V split in the side, so when too much resistance is met the diameter will expand slightly and the wheel can slip.

I have checked on YJ and do not see the corrector wheels at the moment. I am assuming that these were from the same seller as the previous auction identified. I will keep an eye out in the future for the next items on sale but if others spot some please let me know.

Otherwise if we can get one of these YJ metal gears, which are ostensibly sized correctly, might they not provide a good model to mold from?
I think that this would be a very good starting point to test some of the different manufacturing options but it does begin to open it up to some potential criticism for being a “knock off” of someone else’s work and using their IP.

I have checked the laser cutter that I have access to and confirmed a few details. It is a Trotec Speedy 300 with both ytterbium fiber laser and a carbon dioxide (CO2) laser source. This allows cutting a wide range of material types.

I am still to determine the exact tolerances of the machine as it does seem to be quite dependent upon the material being cut, kerf and the laser power required.

I have also had another thought on the practicality of using this to work with. It is possible to do engraving on the laser system as well as full cuts through thin materials.

It should then be theoretically possible to do an initial “engraving” pass to reduce the centre section of the wheel and then a second pass to perform a “cut” for the hole and the outline.

The use of the laser cutter is something that I need to do some further investigation on but it will depend on materials chosen and other variables.
 

·
Registered
Joined
·
2,201 Posts
Discussion Starter · #35 ·
Excellent information Anthony, many thanks for sharing!

Please do follow-up on that potential regarding laser cutting and keep us informed!

Regarding the wheels on Yahoo: it is the same seller that regularly offers these wheels for sale. Unfortunately though, they auctions apparently last only like 2 days every time, not enough for my questions to come thru via FromJapan.
Unfortunately, I thus still do not have an answer to the question whether larger amounts of the wheels could be supplied, if anyone has an alternative way of contacting the seller it would be very welcomed...

Best
Hermann
 

·
Super Moderator
Joined
·
3,226 Posts
I think that this would be a very good starting point to test some of the different manufacturing options but it does begin to open it up to some potential criticism for being a “knock off” of someone else’s work and using their IP.
Hi Anthony,
You make a good point about IP, and I certainly would want to avoid a litigious situation at all costs. But isn't it a bit of a grey area since everyone would be copying Seiko's product anyways? In any case, I wouldn't advocate the final product that will be available for sale being based on just a mold of another aftermarket gear, just as something for us to experiment with within the group. When it comes to as simple a product as this, with few variations possible, how does one safely navigate IP and still produce a functioning part, especially if both 3rd parties take their dimensions from the original Seiko product?

On the laser cutting front, that is very promising news! I will have to read up on it though, as I currently expect you made some of those words up :) Shows how well-versed in it I am (not).
 

·
Registered
Joined
·
2,201 Posts
Discussion Starter · #37 ·
Good point about IP Anthony,

but I am with Noah on this, personally I cannot see any infringement here.
Plus: were there is little money (i. e., profit) involved, high litigation costs are usually not an option, IMO :)

Best
Hermann
 

·
Special Member
Joined
·
1,834 Posts
I also agree on the IP issues and do not think that it is a major thing just something that I thought we should keep in mind and consider.

It was more along the lines if we ended up replicating the YJ version with the distinctive notch it "might" lead to some comments. We have seen this before when other parties replicate a 3rd party crystal or band that others have spent the time to develop.

Again I do not think that is a significant issue at all but just one we should be aware of. It would be a little different if these replacement parts were widely available and we were replicating this to just undercut the price but as we all realize this is not something with significant dollars involved.
 

·
Super Moderator
Joined
·
3,226 Posts
Good points Anthony, and I also feel the ethical side of the coin would be enough for us to avoid an outright copy of the YJ piece. Though we will likely arrive at a very similar looking part once we're done, it is the work involved to get there that counts.
 

·
Registered
Joined
·
2,201 Posts
Discussion Starter · #40 ·
21 - 40 of 70 Posts
Top