G'day TZners,

I recently scoured my motley timepiece collection in search of another subject for my watchmaking practice. I chose a watch that was a little bit different. Once again, I worked on this watch under the watchful eye of fellow TZner Rob B.

Here is the subject of my latest excursion into the world of watchmaking (watchbreaking?). This is a vintage Omega Seamaster I obtained some time ago with the intention of restoring it, including case refinishing and a re-dial. The case was in good condition already, and I simply polished it up using a rotary motortool with a felt polishing wheel and red Dremel No.421 polishing compound.

I like these heavy lugged Omega cases, although when you get the movement out of them they don't feel quite so substantial. I thought it would be an interesting exercise to service a bumper auto, as it's a little bit different to the normal full-rotor system.

Below is a close-up of the dial. The markers and printing are all in very good condition, but it appears that the lacquer finish had deteriorated and someone hamfistedly attempted to remove it completely. The end result is a dial that is very dull, with many brush marks across it. However, it's certainly wearable as it is, and indeed I have worn it a fair bit. The watch could also do with new hands and a proper crown to replace the generic one.

The case screwback is off to reveal the movement. It's a 17 jewel Cal.344 bumper automatic, an in-house Omega movement. The movement serial No of 14032329 dates production to 1954. This one runs at 19,800 beats per hour. As usual for Omega movements from this era it is copper electroplated, giving a pleasing finish that I had previously thought to be rose gold gilding. The movement has a no-nonsense look about it, and on the wrist imparts that distinctive "bump" as the rotor operates.

Although these bumper automatics are not as efficient winders as the later full-rotor movements, I have found this one to reach full power reserve very efficiently with normal wear.

Uh Oh...the movement seems to have disintegrated into it's many, many, component parts! This pic was taken after complete disassembly and ultrasonic cleaning of the parts. I have grouped the parts into their major functional group, and attempted to allocate the myriad of different sized screws where I believed they should be - a task I find difficult, but essential to getting the critter back together without too many bits left over. Surprisingly, I didn't lose any of the tiny screws. Note that the mainspring and barrel were not included in the ultrasonic cleaning. I did not remove the mainspring from the barrel - it was clean and required only an oiling.

Here is the gear train about to pull out of town. I had some major frustration in attempting to replace the train bridge whilst guiding all four pivots back into their respective jewels/holes. I'd get the 3 wheel pivots in, only to have the escape wheel pivot pop out, or vice versa!

It was at this point that a major problem was discovered. At some time in the dim, dark, past, someone had attempted to close up the hole for the center wheel pivot by using a punch around the hole in the train bridge to drift the metal. Under the microscope I could see at least 4 large punch marks. It hadn't worked, and the center wheel was left with significant slop. You really are at the mercy of how well previous service work has been done. In this case I had the options of putting the watch back together as is, having the pivot hole rebushed, or obtaining a replacement center wheel & train bridge. As I don't intend to wear the watch much until it gets a redial, I decided to leave it as is for now. Rob inquired with his supplier about the replacement parts, and to my surprise, they are available from Omega at reasonable cost. (I'm always a bit amazed that parts are still available for such old movements).

Below is a pic of the movement well on it's way to completion, with the balance assembly, and bumper auto mechanism and rotor to be added.

This is a detail pic of the bumper auto winding components - the rotor itself, the bumper springs, and the autowind ratchet.

The rotor is upside down, and you can see that it has a weight attached to the underside to add mass, and thereby increase the momentum of the rotor to assist in winding. I am not sure what metal the weight is made of, but it is not soft enough to be lead.

The bumper springs are rectangular in section. These are sometimes commented upon as a weak point in bumper movements. That may well have been the case in some, maybe earlier, movements, but these springs showed no sign of wear, and do their job admirably. It is important when reassembling the rotor and ratchet to ensure the teeth are meshed in the correct position, so as they do not disengage at either end of the rotor's travel.

In the close-up pic below you can see how the bumper springs fit into the movement. They are simply pushed into the rectangular holes and are retained by the bottom plate of the autowind mechanism. The rotor buffers directly against the springs.

I can't overstate the usefulness of a microscope, both in actually working on small parts, as well as just having a good old gawk at the movement, making it much easier to see how things work, and checking out the finish. This is Rob's 10X stereo microscope, which I found very helpful in working on the balance jewels, and in oiling of all the jewels. Under the scope you can see just how much oil you are applying, and can be sure that you haven't over or under oiled. I did find it a little difficult to judge my movements under magnification at first, but in delicate maneuvers, such as unhooking the tiny arms of the incabloc shock protection springs, it proved invaluable.

Looked at under the microscope, I found the finishing of this movement to be very good. The flat copper surfaces are straight grained with mirror polished angled edges, wheels are circular grained, steel parts snailed, straight grained or oilstone greyed depending on the item, and all screw heads are polished "black". The visible non-functional edges of the plates are anglaged as well. The smaller individual parts, the pallet fork for instance, showed clean finishing at 10X, and the gears are well cut. The only obvious flaws I noticed were artefacts of previous watchmakers gentle ministrations. All-in-all, it appears to be a well made movement without too much pretension to fancy decoration.

Here is the movement completely reassembled. The balance, and bumper-wind, assemblies have been added. It took me a long time to get to this point - many, many hours, with breaks only for food and rest! Rob gave me "hints" on how to go about the reassembly, otherwise I'd still be sitting there now!

Below is a close-up view of the swan-neck regulator. These might look good through a glass display back, but they really are very beneficial in the fine regulation of the timing of the movement. This arrangement allows very fine adjustments to be made via the screw on the regulator.

So, everything was back together, what needed to be lubricated had been lubricated, and I didn't seem to have any bits left over. What could go wrong from here? Well, the darned watch ran about 6 minutes fast, that's what! I was pretty sure it hadn't run like that before, so there had to be a culprit somewhere. Surely it couldn't have been my skillful work?

Rob pointed out that the balance assembly allowed for seperate adjustment of the position of the hairspring curb pins. This adjustment operated independently of the regulator, by allowing the pins to be moved along the hairspring, effectively lengthening or shortening its travel. It is quite likely that I could have moved this adjustment whilst manhandling the balance assembly in & out of the movement. A few adjustments later, and a bit of fine regulation via the swan-neck and the movement was getting some very decent readings off Rob's Vibrograf timing machine. It was great to see those straight traces appearing on the paper!

Here is a summary of the timing machine results in five positions:

Position	Rate/24hrs	Bal. Amplitude
Dial Up*	-5 sec	300 deg
Dial Down	-10 sec	-
Crown Down*	+6 sec	-
Crown Right	+3 sec	-
Crown Left	+2 sec	-

Beat Error 0.2ms
* The most natural wearing positions of Dial Up (-5 sec) and Crown Down (+6 sec) should cancel each other out to result in quite an accurate daily rate when worn on the wrist. I will wear the watch to see if this theory is correct.

I was very happy to obtain such results after servicing the watch. Whilst I found certain aspects of the reassembly to be quite a challenge, it was very enjoyable, and I learned a lot. The bumper wind aspect of the movement was straightforward and held no great surprises. My thanks go to Rob for his inestimable assistance and encouragement, and the use of his tools and timing machine.

Hmmm...now, what shall be my next victim?

Copyright 2000 Paul Delury.