Dogwood's Watchmaking Journey

[Financial Mistake modding]

A few days ago I received a beautifully modded 3KF Nautilus with buff dial, hands, and datewheel. This was my first 5711 and I have to say, this watch is not over hyped. It is a stunning design.



However, less than a week later it suddenly stopped working. I know clone movements can be unreliable and expensive to fix since getting spare parts is nearly impossible. But without any options, I decided to open the watch up and take a look at what might be wrong. I decided to keep the movement in the case since I was hoping the problem would be obvious and superficial... and I didn't want to risk marring up the buff dial and hands.



Digging into the movement is pretty straight forward and most of the movement is accessible from the back.



The balance cock and automatic bridge came away easily. At this point I couldn't see anything obviously wrong, and I should have fully removed the movement from the watch to continue disassembly since the click isn't accessible from the bottom of the movement so there was no way to remove power from the main spring. But I forged ahead since I was on a roll, and it turns out, there was no power in the mainspring for reasons that will become obvious in a moment.



When I removed the barrel / train bridge, there was no power in the mainspring. But I did find a random 0.4mm screw jammed under the bridge. This screw isn't from the movement, so I assume it got in there either during assembly, or during the modding process in China. But as bad as an unsecured screw floating around the movement was, the reason the movement stopped was because the 3rd wheel's arbor had broken in half.



In the pic above, you can see the third wheel and the pinion still in its jewel pivot hole and the grey of the metal failure on both sides.

Needless to say I'm a little sad since according to my WeChat dealer, 3KF doesn't sell these movements on their own. So that means I'll have to buy a whole new 3KF watch just to get a spare 3rd wheel for the 324 SC movement. But, before I pull the trigger on that... if anybody out there has a broken 3KF 324 SC movement with an intact 3rd wheel, I would love to buy it form you

Hey mate, if you still need parts for this movement send me a message

 

[dogwood]

Had a couple of Omega Planet Oceans arrive on my bench today. One just needs a keyless reset, the other a full service and a hand swap.



This was my first time working on this particular watch / movement so I took a few pics along the way.

Interestingly, because the gem has a closed caseback and an antimagnetic cover over the back of the movement, the rep movement is completely undecorated. Thos made it a joy to service compared to the VSF omega seamaster GMT.



The antimagnetic cover just needs to be lifted / pried free. This wasn’t obvious initially, but it doesn’t screw down, it’s a friction fit.

Otherwise my only comments on the movement were that they keyless works were a bit fiddly to reassemble but that was because I didn’t have the balance hack spring in the correct location on the bridge side of the movement (note to self for next time).

And I’m really pleased with how stable the time keeping is after regulation. Overall a fun movement to work on.

 

[dogwood]

Another Omega Seamaster arrived on my bench... this one in need of a keyless reset and a replacement set of hands fitted. Unlike the previous Omega Seamaster which was a Noob (see post above), this one was from OMF and had a decorated rotor:



Interestingly, on the OMF version of the movement (which is an ETA 2836) the stem release pusher has a slot (probably so that it can be actuated by a screw driver) through the barrel bridge. The Swiss ETA has the same system (see the pic below), but on the Noob the hole wasn't a slot, it's was round.



What's interesting, is that the issue with the keyless on the OMF movement was that the setting lever wasn't sitting all the way down in this hole. It appears that there was a metal burr or edge that had been pulled up somehow and it was preventing the setting lever from fully seating. I wonder if this was due to the fact that the release could be operated with a screw driver -- and the sharp edge of the screw driver could have pinned over a tiny bit of metal from the main plate which would have tightened the hole?

Regardless, it was a simple enough fix. I simply used a broach and re-opened / smoothed the setting lever's hole. And with that, the setting lever slotted into plate without hesitation.



With that complete the keyless (and dial side) re-assembled easily (the ETA is a very easy movement to work on).



Then it was a case of re-attaching the dial and pressing on the new hands.



Overall, a nice little project and an interesting fault finding exploration of the ETA keyless works.

 

[dogwood]

Today I was finishing up a service on a Noob v3 Daytona. As part of the service on the watch I was doing a bezel swap, and also replacing the old Noob hands (which were a bit marred up) with a new set from AliExpress.

New hands just arrived in the mail. Old hands installed for now to test watch and chrono.



When I went to press on the new hands, the hour and minute hands went on easily enough, but the chrono seconds hand, and all of the sub-dial hands from the new AliExpress hand set are incredibly tight. In fact, they are too tight to easily fit. I struggled for a while trying to even get them into position on top of the arbor, and then when I went to press, it was taking a LOT of force and the hands weren't slotting down into place.

I measured the hand arbors to be 0.22mm for the sub-dials on the Daytona. My pin gauge set still hasn't arrived, so I don't have an easy way to check how large the holes are in the new sub-dial hands, but in this digital microscope pic of the new (left) vs old (right) hands, there is a tiny difference.



My smallest broach is 0.60mm to 0.80mm... is this the kind of thing that requires a super fine set of broaches? e.g. :

Extra Fine Broach Set Set of 12

Add precision to your work with the twelve-piece extra fine broach set, featuring a miniature pin vise and twelve steel broaches for enlarging holes.

www.esslinger.com

I have the feeling that I have and N+1 moment about to happen to the number of tools on my bench.

 

[TheSilentSinger]

Great stuff. It's good to know a watchmaker on the Canadian side from this forum. I may look into using your service in the future.

 

[dogwood]

My micro broaches still haven’t arrived to open up the Daytona hands. But I have a nice Panerai 127 on my bench this week with a beautiful 6497-2 movement. The factory did a great job with the bridge decoration on this movement. And I have to say it’s a joy to be working on this movement — compared to working on Rolex, Patek, AP, and even the decorated ETA calibers, everything on this movement is HUGE! I understand why it’s the movement that watch making courses always start on. It’s simple. It’s good everything in the normal places. But most importantly, all the parts are big and easy to manipulate with tweezers.

 

[dogwood]

Also… a set of pin gauges I ordered a few weeks ago arrived. This set goes from 0.10mm to 1.00mm in 0.01mm steps. I’m not sure if this is a standard tool set for a watch maker bench, or if it’s overkill (there’s no kill quite like overkill), but this should allow me to measure inside diameters on things like seconds hands and jewels more easily.

 

[dogwood]

The replacement 3rd wheel for my 3KF 324SC movement in my 5711 arrived today. Many thanks again to @Financial Mistake modding for sending it to me all the way from Germany. I’m going to clean a few of the parts that I removed from the movement and then re-assemble to see if I can get my 5711 working again. Wish me luck.

 

[dogwood]

Also, I received an Asian clone 7718-1A automatic oiler that I ordered from AliExpress. I was skeptical of the quality, but at $25 and free shipping I figured it would be worth the risk. The oiler itself isn’t that bad. The mechanical action isn’t amazing, but it’s acceptable. The problem is the crack in the plastic reservoir. Fortunately I discovered the crack before I filled it with expensive Moebius 9010. I tested it with alcohol and it leaks badly. Lesson learned, you get what you pay for.

 

[dogwood]

3KF Patek 5711 fixed. Thanks again to @Financial Mistake modding for the replacement 3rd wheel for the movement. It’s lovely to have this watch back together an on my wrist.

 

[dogwood]

So, after swapping in the replacement 3rd wheel into the 324 SC, the movement ran for about an hour and a half, but then stopped. I though that was odd so I put it on my time grapher and plugged in the correct lift angle (51 degrees for the gen 324 SC), and found that the amplitude was quite low — around 200.

I figured maybe I had messed up the reassembly and didn’t have one of the pivots in the jewel hole (terrifying but not impossible). I checked under the microscope and everything looked to be in place. Even so, I disassembled the balance side and re-assembled it being extra cautious to see metal pivots in jewel holes before screwing down the plates. The amplitude got better but only by 10 degrees and the watch still randomly stalled.

Then I thought the issue might be that I might have a deformed tooth on the first or second wheel that only came around every hour or so and that was degreasing efficiency and causing the watch to stall only when the deformed tooth was engaged. I disassembled the movement again and checked every tooth carefully under the microscope. The teeth all look good.

Then I thought the issue might have been old oil / dirt on the main plate jewels. I only cleaned the gears and bridges since I was being “lazy” and wanted to keep the watch cases up. I used a couple of foam tipped cleaning sticks to peg out the jewel holes on the main plate. One of the holes was a bit dirty, but all the jewel holes were round and smooth after cleaning. I reassembled but only improved the amplitude by another 5-10 degrees to 220. My two other 324SC movements from 3KF run at 280-300 amplitude.

Then I thought it might be end shake on the 3rd wheel being slightly different from before (since that’s the only part I’ve swapped). There is quite a noticeable amount of end shake on the 3rd wheel, so I wonder if I should press it in a bit to keep that wheel more tightly held.

I also wonder if I accidentally damaged the center seconds damping spring somehow. This spring seems to be the issue with reliability on many of the BBR clone Richard Mille movements. The 3KF 324SC movement has the same design — center seconds driven using a tiny pinion held down with a spring to minimize stutter. But the spring tension, if too high, can reduce amplitude.

Anyway… those are my musings on this movement for tonight.

And a few pics:

 

[dogwood]

Ok... so I've had the 324 SC running on my time grapher for about an hour and it didn't stall out. And the amplitude seemed to be improving ever so slightly to 230 degrees.

But during that hour, I thought about some features I'd really love to see in the TG0.5.0 software that I'm using to analyze the tick-tock of the watch... and as luck would have it, TG is an open source project with a long thread over on Watch You Seek. The original project is by GitHub User vacaboja and goes up to version 0.5.0. But interestingly, GitHub lets you see everybody who has forked the code, and user xyzzy42 has made some improvements to the original code base including the ability to view the time graph horizontally. Unfortunately, if you're a windows / mac user and don't like compiling software from source, you can only download version 0.5.0 in binaries from the original author's website (https://tg.ciovil.li/), if you want the latest and the greatest, you'll be compiling it from xyzzy42's source on github.



The process isn't too bad as long as you're comfortable with *nix shell commands and can follow instructions and google error messages. Long story long: I now have a fancier version of TG time grapher software that let's me see the chart horizontally, and let's me stretch said chart out to expand the scale of any variability.

What we're looking at here (top chart) is the standard time grapher that we all know and love from QC pics. The red vertical lines are spaced 1 minute apart, so the chart shows a little over 5 minutes of the software recording the movement's sounds. The thing that jumps out to me (red circles) is that the movement seems to lose time keeping on a periodic basis almost exactly every minute. Unfortunately the software doesn't allow me to see a chart of the amplitude plotted over time (not does it have a log file where this data is saved)... but interestingly, the amplitude during the "red circle" sections drops to around 220, but during the "non-red-circle" sections the amplitude hovers around 230.



This periodic wobble in the time grapher is making me even more suspicious that the issue is that one of my train wheels (specifically one that turns once every 60 seconds, i.e. the 4th wheel) has some damage or roughness. But when I look at the 4th wheel under the microscope and record a video of the wheel moving through a full rotation over 60 seconds, I don't see any teeth that are damaged or have a bad profile.

While this isn't satisfying enough for me, at least we're making progress. I'm not pretty sure that something funky is going on with the 4th wheel. Although now that I think about it, maybe the issue isn't with the large wheel of the 4th wheel, but rather the pinion of the 4th wheel where it's driven by the 3rd wheel. Unfortunately, I can't easily see the 4th pinion leaves since they're hidden by the bridge. But that gives me something to investigate.

 

[dogwood]

Hmmm... the 4th wheel pinion leaves look undamaged. Although maybe they got sticky somehow from me over-oiling the jewel. I've put all the train wheels and bridges back into the cleaner to get them oil free to try again.

 

[Kurtoseas]

Mate, this is a super interesting thread to read.
First of all, congrats on your incredible watchmaking journey. I am understanding that you are a self-thought watch smith now and experimenting/ learning on the way. Such a beautiful endeavor, something that I don't have the patience for..

Secondly, such a nice way to record the journey and give some insightful and entertaining for us watch folks.

And last, keep up the good thread. This is great and I just wished that you were here in Dubai

 

[dogwood]

Ok... so something was bothering me when I thought about the 324 movement... specifically the ratchet wheel that sits on the barrel arbor. As you can see in the two pics below, the ratchet wheel has a round central bit that elevated on one side -- it's sorta like a flying saucer shape. Initially I didn't think much about which way to fit the ratchet wheel. It fits and works in both orientations, but it wasn't obvious which way it should go. And unfortunately I didn't have a picture of which way it was when I first took the movement apart (stupid me!). But upon reflection, I think it makes the most sense for the raised center bit to face down towards the mainspring barrel. Here's my rational: when you wind the mainspring, the ratchet wheel turns (moves) and the barrel is stationary; when the watch runs, the barrel turns (moves) and the ratchet wheel is stationary (held in place by the click). When winding the mainspring there is a lot of force available via the crown or the rotor, but when the watch is running the only for available is from the mainspring. So any design for the ratchet wheel would take this into account and try to minimize friction between parts. If the ratchet wheel protrusion faces towards the barrel bridge, then it would be minimizing contact surface area (and friction) between the ratchet wheel and the bridge, but maximizing surface area contact (and thus friction) between the ratchet wheel and the barrel. If the protrusion faces downwards towards the barrel, then it would minimize the surface area contact (and friction) between the ratchet wheel and the barrel, but increase the contact area between the bridge and the ratchet wheel... however, on inspection of the bridge, there's an elevated rim on the bridge that is there is minimize contact. So clearly the protrusion should face down towards the barrel... or so goes my logic.

TLDR; I flipped the ratchet wheel and saw a small improvement in amplitude to around 230 degrees. It's not the silver bullet I was hoping for, but it is an improvement.

Protrusion facing up:


Protrusion facing down:

 

[dogwood]

New tool arrived — a set of 12 extra small cutting broaches (0.05mm to 0.55mm).

 

[dogwood]

I got to put the super fine broaches to work right away opening up the holes in the Daytona hands that I have. This is a very delicate process and you don’t want to accidentally open the hand up too much. I did this once and had to use tweezers to squeeze the hole back down a little (a real watch smith, i.e. not me, would have a staking set to close the hole). But going slowly allowed me to get all the hands fitted without much drama.



It’s strange how easy things are when you have the right tools. I’m gonna need more tools.

 

[dogwood]

Finishing up the Daytona. As part of the project I’m doing a crystal gasket swap and a bezel swap. I’ve found the bezel goes on a little easier if I put a homeopathic film of o-ring lubricant on the gasket before pressing on the bezel.

 

[dogwood]

When I was re-casing up the Daytona from the post above, when I re-installed the rotor on the automatic module, I (of course) test the action to make sure everything was working smoothly. But interestingly I found that in one direction, on occasion, I would get the rotor spinning backwards!

I removed the entire automatic module, disassembled it, re-washed all the parts, and examined each of them under the digital microscope. For one of the reversing wheels, I noticed a lot more wear than on the other:



I treated them both (and the little internal ratchet wheels) with epilame, and then re-assembled and lubricated then automatic module. The rotor works as expected now, but I strongly suspect that this Noob v3 reversing wheel might be close to its end-of-life. Fortunately gen replacements aren't that expensive ($75 on ebay). But this is the first time I've seen the rotor spin back so dramatically, so figured I'd share.

 

[dogwood]

Ok, so I know that most of you who come here are here for the macro shots of mechanical movement porn... but I wanted to share something very cool that I've found.

A few posts back I mentioned that there's a great piece of open source software call TG that turns your computer into a time grapher if you have a suitable amplified microphone. And a few posts back I mentioned that I'd compiled it from source to get access to a few newer features. Well, it turns out I wasn't even compiling the latest version and after posting on watchyouseek, the author of the software suggested that I compile from a different branch of the code on GitHub.

For those who don't know:

Ok, so I compiled from a different branch of the code, and now I have a plot of my movement's amplitude (yellow line) on top of the tic-toc pixel graph.



What's amazing about this, is that I can clearly see that the amplitude follows a 60 second cycle. This implies that there's something going on with the 4th wheel pinions or leaves which causes the amplitude to stutter, rise, and then fall... then over the following seconds the amplitude gradually rises back up as the balance wheel collects energy again. The amplitude of the movement is a good indicator of how efficient the power transfer between the mainspring through the train wheels to the balance is... and obviously this efficiency will change depending on which specific wheel teeth are engaged at any given time. I love being able to see that change visualized.