Thursday, December 26, 2013

Oil Spill

The last thing I tackled this last quarter was to change the battery on an oil-filled quartz watch.  Whoever had attempted this in the past has let an air bubble into the case and the watch had stopped.

After installing the battery the watch was still not working.  I am used to working on these movements.  It's essentially the same watch movement we had to complete on our SAWTA 2 tests.  So i took everything apart and during the process I dinged the coil and ruined it.  Replacing the board was not an option because this is a special 3 volt board with a custom battery holder.

No big deal, we have plenty of extra coils around.  So then began the painstaking process of melting off the existing solder and epoxy and soldering on a new coil.  This is difficult because the wires are VERY fine.  After about a half an hour I was able to get a new working coil installed.

This process is incredibly messy.  There was oil everywhere, all over my bench and my tools.  It was a very strange feeling getting oil all over a newly cleaned movement as that's usually exactly what you are trying to avoid.  At least putting the movement back together was quick because you didn't have to worry about oiling anything.

Once back together and ticking once again I refilled with oil, vacuumed most of the air out and we now have a watch that only has almost microscopic bubbles suspended in the mineral oil.  I've not figured out how to remove those yet, I'm guessing a stronger vacuum pump would help with that.
Oil-filled watch, complete with bubble

Even with a new battery the watch did not run, it needed an ovehaul

My vacuum platform, made mostly from scavenged materials helped pull some bubbles out of the oil.  The air compressor's pump does not draw a very efficient vacuum though so it was not as effective as it should have been.  We need a proper vacuum pump.

Thursday, December 5, 2013


Three down,  one to go!
I passed my SAWTA 3 exam this week.  The test involves 4 parts:
1. Overhauling, repairing and casing of a 2824
2. 15 question theory
3. Case refinishing
4. Bracelet refinishing
Things went very smoothly this time around.  The only critique the examiner had for my watch was that I over-oiled the crown wheel and was a little light on the oiling for my lower balance pivot.  Besides that I got full marks!
The quarter is nearly finished, the last project I have on my plate is a cell change in an oil-filled quartz watch.   I am devising a vacuum system to make this simpler.  I will post soon.

Tuesday, November 26, 2013

Fixing a Levin Toolrest

The detent screw on one of our Levin lathes was stripped out rendering the part useless.  For my school volunteer hours this quarter I made a new post and installed it into the slide attachment.

The primary difficulty here was turning down the chunk of steel on a lathe using a 3-jaw chuck which didn't give me the best grip.  Unfortunately we don't have a lathe that can use a regular collet to accept these larger diameters.  Despite the cutter digging in and dislodging the workpiece on a couple of occasions I was able to turn the lower diameter down to where I could use a normal collet to hold the piece and finish forming the top.

To cut out the core I first used a end mill followed by some reamers on the Bridgeport milling machine in the basement. Finally, the detent hole and threads were formed using a drillbit and a simple tap.
The finished post, pressed into position

Friday, November 22, 2013

Overhaul/SAWTA 3 Practice

During the last couple of weeks I have been busy practicing overhauling and troubleshooting automatic movements.  Most commonly we've been working on the ETA 2824 as that's most likely the watch we'll receive for our SAWTA 3 examinations.

During the examination it is our task to troubleshoot issues with the watch which have been introduced by the examiners.  During dis-assembly you identify and fix the problems, then all that remains is to clean the watch and recase it with a high level of accuracy and cleanliness as both will be factors in the grading.

Besides the overhaul there is a written portion of the exam and a polishing exercise, both of which are proceeding nicely for.

I have taken some pictures of a practice overhaul and will offer a brief explanation of some of the steps below. 

The caseback has been removed
The rotor and stem have been removed, the movement can now come out of the case
Removing the hands while protecting the dial and hands with some plastic
The dial-side with the dial and date ring removed, nothing to repair on this side!
The hairspring is not centered, the coil spacing near 11oclock are wide and near 5 they are narrow, which will need to be corrected by bending the spring
Another shot, the dogleg in the spring near 12 o clock seems is where we will perform the correction
The entry pallet stone is not level in the fork, it has dipped out of its holder and needs to be reset so it can engage the escape wheel
The tilted stone
Using an electric heater and this expensive gauge it is easy to move the stones in or out of the pallet fork micro-metric amounts for precise depthing correction which is essential for a well-running watch
The endshake (up and down movement) of my pallet fork was too small, so here I am pressing the jewel back into its bridge.  Once the jewel was flush with the bridge the endshake was back within tolerance

Here I am adjusting the barrel endshake, which was also much too small.  Too little endshake reduces freedom of the wheels and introduces extra friction into the mechanism.
Here I have corrected the centerdness of the hairspring, looks slightly off here due to the balance being loose, once installed it is quite good.  Once the watch was reassembled the instantaneous rate and running characteristics of this movement were within 100% tolerance for this movement

Wednesday, October 30, 2013

Forming a hairspring

This week I've been working on forming my own 'balance complete'.  All you get are the parts you need: balance wheel, balance staff, collet, roller table, hairspring, and pins.  First one must rivet the balance wheel to the staff, then press on the roller table and make sure they are true (flat).  Then using a ruby-jawed vice you must 'poise' the balance to make sure the weight is evenly distributed along the whole diameter of the wheel.

Once the wheel is set you need to take the hairspring and either snap it or pin it to the collet, press the collet onto the balance wheel and then, using a counting platform find the 'counting point' of the spring.  This is the point where the springs length, when paired with a certain balance produces the desired vibrations,  in this case 18,000/hr.  Once this point has been located, you mark the position by clipping off the excess spring 1 coil further along the spring (as the spring is too long for the job anyway) 

Poising the hairspring is the tricky part, it involves lining up the point of attachment of the spring to the collet and the counting point as close as possible on one side of the hairspring.  This is accomplished using 'Leroy's method'.  Basically you take the angle from the counting point to the point of attachment following the spring (a) then take (a/3)=b, then take (b+a)= new point of attachment.  So just take your current point of attachment, move forward along the spring in that many degrees to find your new point of attachment.  If you measured correctly it will poise the spring.  Keep in mind you want to subtract about 60dgs from the above number for your actual cut because you need extra spring to fit inside the collet.

So now you should center and flatten your hairspring on its collet then put a bridge with a pinnable stud in a watch.  Push your hairspring through the stud and loosely pin it there.  Get it into the ballpark, meaning your new counting point (which you have now checked using the counting platform) should be in the vicinity of the regulating pins.  Then use your timing machine in frequency mode while slowly adjusting the amount of spring coming through the stud.  Once you are within 10 vibrations either side of your target drive the pin home and cut off excess spring.

Now you can busy yourself with forming the dogleg which will interact with the pins, lead to the stud and allow the spring to breathe more freely.  I did this using a tool I created for this particular job.

Lastly, install everything into the watch, make sure everything is still flat and centered, set up your regulating pins and then regulate the watch to a good time.  If everything went smoothly and you didn't muck anything up you now have a working hairspring. 
Truing the balance wheel

Pressing on the roller table

Using the poising tool

More poising

Pinning the collet to the hairspring

The vibrating tool

Another shot of the vibrating tool

Here's the first hairspring I made in the watch, it keeps fairly good time for a first try!

This is a neat little steel template I created to help in forming the dogleg accurately.  The desired radius is scribed into the hardened metal and the 120dg desired angle is also scribed.  One simply has to screw the hairspring collet into the recessed hole then begin manipulation.  The steel is hardened to prevent tweezers scratching the metal while working.
Finally a old Elgin I fixed up for a customer. When it came in it had no minute hand, crystal and was not running.  Works great now.

Thursday, October 10, 2013

Overhaul & Refinishing Complete!

I have completed work on the Constellation.  The results are quite nice I think.  The only thing left to do now is run the complete watch on the winder for a few days to make sure it's keeping good time then mail it back.
To press the bezel back on and secure it with those little claw clamps i took a piece of grey PVC, turned it down even on the lathe then filed some notches in the plastic to accommodate the clamps. Then it was just a matter of using the crystal press and voila, reassembled bezel.

On the winder, this simulates the wearing of the watch

Custom made bezel press, the factory version is discontinued so I made this one.

The completed timepiece, just small traces of polishing compound to remove

Wednesday, October 9, 2013

Refinishing Progress

To remove the bezel on a constellation you first have to undo the bezel claws.  These are two gold grippers that fit into a slot on each side of the case and push down on the bezel, securing it.  To remove these (without the proper Omega tooling) I secured the case face down in a case holder.  After securing the case holder in a vice I covered the setup with plastic to minimize damage to the case.  Finally I used a brass punch to hammer down on each of the little claw legs a bit at a time until they all came loose. 

Once everything was disassembled I had work to do on 3 main components.  The case frame, the caseback and the bezel.  The main issue with the frame is that is has a very angular cut to it so one has to be very careful with the hardbuff wheel to keep from changing the dimensions of the case.  I was able to remove most of the scratches in this fashion but a few of the larger dings are still visible.  It's a tradeoff between a flawless finish and possibly ruining the angles of the case.  After brushing the fresh finish is a major improvement.

The back of the case and the caseback where years of hapless battery changers have attacked the area is pretty chewed up.  So I did my best to minimize the level of visable damage.  The damage is so severe you really can't expect to pry up on the case lip with a caseknife anymore.  So I took everything for a spin on the lathe and burnished and sanded out the majority of the damage.  Here again it's a balance between getting rid of all traces of damage, but possibly removing to much material the caseback won't be able to snap down anymore.

The bezel was easier, there was only minor scuffs and spinning the ring on the lathe and touching up the lined finished with a hard chunk of Artifex did the trick.  The original finish is back and looking great. 

The final challenge is to reassemble everything, and lacking the Omega case and bezel jig I'll need to fashion my own to get those little claws back and seated properly.  I am thinking it'd going to be necessary to fashion a die out of some plastic pipe.  I'll be attempting this operation tomorrow or Friday.  Stay tuned.

Claws, removed.

This is a fun and easy step.  Provides great results too, as long as you don't overdo it.

This is AFTER some burnishing.  I brought it down a bit with sandpaper after this then burnished again.  It's a very minor scar now.

Tuesday, October 8, 2013

Omega Constellation Overhaul

Today I started work on a Omega ladies Constellation.  The watch was experiencing intermittent stopping, the band and case were also quite scuffed up. 

To attempt to remedy the problem, in addition to installing a fresh battery I took the entire movement apart then cleaned and reassembled it.  The calibre of the movement is 1456, but we only had a tech sheet for a 1458 but this proved close enough for most figures and oiling schematics.  I have completed the overhaul of the movement and it has been keeping time for at least a couple of hours so far.  I have it on the watch winder for the night and we will see if it kept time when I return in the morning.

I also hard-buffed the steel link segments, white rouge buffed the gold links then used the bufflex wheel to apply lateral brushing on each steel link.  It turned out quite nicely, nearly as good as new!

The last step in the process is to take the case apart and refinish the frame and bezel.  The tricky part there is that Omega uses 4 small clamps that hold the bezel on and without the proper tool it is nearly impossible to get them back on properly.  I have an idea what the tool looks like though, so I may end up turning out my own on the lathe and fitting it onto the crystal press.  Will be an interesting process.  I will make sure to post final results.

The watch with a link removed to allow access to the caseback.  That little pin sure put up a fight.

Hand remover tool with handless dial

Dial, removed.  Notice the beautiful pearlage finishing to the dial plate.  No one but the watchmaker will ever see this touch.

It's a good idea to cover up the extremely delicate coil with a buff stick when working nearby.  Here I was unscrewing a circuit board screw.

The gear-train, bridge, stator and hack on the mainplate
The delicate electronic components stay out of the cleaning machine, they are cleaned by hand with circuit cleaner

After removing scratches with the hardbuff, the gold links are then masked off to prepare brushing

Brushing complete

Bracelet, refinished.  Just need to mask and brush the locking clasp.

Friday, September 27, 2013

Greiner Vibrograf Poseidon LT-100 Service

In our classroom on the casing table we have a dry watch case waterproofing tester.  It uses very cool technology.  Basically, you stick the cased watch inside the machine on a table then bring down an arm with a micrometer attached to it and rest it on the crystal.  The machine then subjects the watch to (in our case) -0.7bar pressure then a +7.00 bar pressure.  If the micrometer senses the watch case deforming, and remaining deformed you know there is a pressure differential between the inside and outside of the case and it is water tight.  If it does not deform, or does slightly then returns to normal, then you know there is a leak somewhere.

We were having an issue with the vacuum -0.7 bar test on our machine.  It would only work sometimes and we were wondering why.  So I emailed the manufacturer and got a call from the service center in California.  They were very helpful and explained some things I could try to remedy the problem.

After tearing the whole machine apart, servicing it, then putting it back together and ending up with the same result I started fiddling around with the diagnostic mode.  This mode which is built into the service program allows you to activate all of the separate valves in the system manually.  I finally noticed that we only achieved enough pressure in the venturi (which draws the vacuum out of the test chamber, using high pressure airflow) when the compressor was running. 

It turns out the air compressor we are using to feed the testing machine is not charging the tanks at as high a PSI as it used to.  Because of this we were only getting one full negative pressure vacuum test before the machine would mysteriously fail the test.  With this information now I can attempt to service the compressors regulator or just make sure to drain the tanks enough to bring the motor back on to provide me with a high enough pressure to run the test.

Regardless of the extra work, I learned a lot about the workings of the machine.  It is well designed, built and very elegantly put together.

The machine in question

The venturi which creates the vacuum for the system using 'Bernoulli's principle' is the plastic T fitting on the left labelled "S". To remove the top plate and test chamber, you must disconnect the input and bleed lines from the back plate,

And the manifold pressure line,

And the ribbon cable and grounding wire

Here you see the valve stack.  They control vacuum pressure, air in and air out

A solenoid case without its electromagnet

Some small metal shaving on the sealing face of the solenoid, this was removed

The solenoid post, core and spring

These were all removed, wiped clean, blasted with air then reinstalled

The final valve assembly