, , , ,

Between 1924, when K. Hattori & Co., Ltd. first used the Seiko brand on a watch, and 1950, when it introduced its first centre-seconds wrist watch, the movements that powered many of the company’s wristwatches were sourced from the Swiss watch company, Moeris.

The first Seiko wristwatch (1924). Photocredit: watchtime.com

Those earliest two-hander, small seconds sub-dial wrist watches used 8 or 9 ligne, 7-jewel Moeris movements, with their characteristic banana-shaped bridge design.

Adapted from Yahoo Japan and eBay auction images

Subsequently, a larger 10.5 ligne version of this movement was used in watches produced from the mid- to late-1920s.  This movement was known as the cal 10 and although branded SKS, its Moeris DNA is still pretty clear.

A new calibre 10A was introduced in 1946 and produced in the Daiwa Seikosha Suwa plant, the evacuation factory used by Seiko to re-establish its post-war wrist watch manufacturing following the destruction of the Kameido Factory in Tokyo.

This new movement used separate bridges for the escape, second, third and fourth wheels but its gear train layout is still recognisably carried over from the cal 10.  From 1948, a further development of the cal 10, the 10B, was produced in the new Kameido factory.

It is not entirely clear to me whether all of these movements were simply bought in from Moeris and re-branded or whether they were built by Seiko under licence but I suspect the former.  What is clear however is that they were not Seikosha designs but rather developments of Moeris designs.

In 1950, Seiko debuted the Seiko Super, its first wristwatch with a central seconds hand.

Photocredit: daddyswatch.com

According to the entry in the Seiko Museum website, this new movement “…required a departure from the existing transmission wheel system to a whole new wheel design from scratch.”  To me, this implies either that the Seiko Super was an original scratch design or that the train wheel layout was redesigned but in the context of an otherwise existing architecture.  I am pretty confident that it is the latter, given the close similarity between the barrel bridge layout and design of the Super (below) and the calibre 10b pictured earlier.

Photocredit: Seller yokayan114 on Yahoo Japan

That conclusion resonates more in the context of the received wisdom that the follow-up Seiko Marvel was Seiko’s first wholly original in-house watch.  The Seiko Epson Global site describes the Seiko Marvel as “an originally-designed mechanical watch that became the basis for Epson’s watch business”.  That being the case, it is probably safe to regard the Marvel as the original modern Seiko wristwatch.

The upsizing of the Marvel from the 10.5 ligne of the Super (23.7 mm) to 11.5 ligne (25.6 mm) allowed for a number of improvements in design, with the focus now on accuracy and the development of a product that could compete in a global market.

Adapted from 1956 Seiko News Vol. 1

The larger diameter of the new movement permitted: an increase in the barrel size, allowing for a higher performance mainspring; larger gears with an increase in the number of teeth, separated by greater distances and allowing for a more even transmission of torque.  A larger balance with a larger moment of inertia allowed for a greater resistance to externally-generated forces and an improvement in rate stability.

All of that theory was born out in practice, with the Marvel movement fitted to a whole range of Marvel-branded Seiko watches produced between 1956 and 1959.

From a 1956 Seiko catalogue. Note the model second from left, lower.

A testament to the inherent qualities of the movement is provided by its subsequent up-scaling and transformation into what was to power the first Grand Seiko in 1960.

This lengthy preamble brings us to the traditional point in proceedings where we introduce our protagonist.  However, in contrast to most of the accounts presented here, our starting point this time is not a complete watch but a fragment of what was once a complete watch.  We begin with a dial, a solitary hour hand and an hour wheel.

I suspect some of you may be thinking that this is no basis for anything other than consignment to the cylindrical receptacle in the corner of the room.  My reaction to this dial when it first caught my eye was that it is an object of beauty whose potential shines through the liberal distribution of patina that covers its surface.  It is simply lovely and for me there is no question that it should be saved from oblivion.  Those of you who read the prologue to this entry (here), will know already that I encountered this dial as part of a minor haul of dubious treasure obtained a few years ago while on the hunt for 56-series movement parts.  A sub-set of that haul comprised a small number of very early Seiko Marvel and Crown movements, dials and cases.  My original thinking was that within that collection there ought to be a sufficient number of parts to build up at least one complete example of the Seiko Marvel and one of the Seiko Crown.  In filtering thorough the assorted detritus, I identified about one and a half complete 17 jewel Seikosha Marvel movements, the dial and hand pictured above and two or three stainless steel cases.

As it turned out, none of the cases were of the correct size to accommodate the dial and so I would be having to look further afield for a solution of how to house the watch innards should I get that far.  The two partial Marvel movements also presented some difficulties, principal amongt which, a broken hairspring collet and abused balance wheel and staff for the single balance shared between the two movements.

In the absence of any suitable substitutes from my parts stash, my only option, if I wanted to maintain the Marvel spirit for this project, was to source another functional Marvel movement.  The first-generation Marvel movements were not fitted with shock protection but later versions of the Seikosha movements fitted to Marvel and Laurel models benefited from Diashock shock protection for the balance.  And so, in concocting my on-the-fly strategy, I decided to try to find a Diashock-protected Seikosha movement for as little an outlay as possible.  Within a week or two, I’d secured a suitably bedraggled candidate.

Having extracted the donor movement from its case, it appears we have something reasonably sound, but somewhat tarnished and encrusted with salts.

A closer look around the edge of the movement in the vicinity of the balance shows just how badly contaminated it is with crusted deposits.

The lighting does the movement no favours in these photos – it does not look this bad live – but nevertheless, this is a movement in a pretty poor cosmetic state.  Deconstruction proceeded smoothly to the train bridges on the balance side, this being a pretty conventionally-designed hand wind centre seconds design.

We note however that the ratchet wheel is badly corroded on its underside as is the crown wheel to its upper.

I will replace both with clean parts farmed from the other Marvel movements during the reconstruction later on.  The train wheel bridge and train wheels themselves parted company with the main plate easily enough but the same wasn’t quite as true for the centre wheel and its bridge.

The centre wheel was glued to its bridge by congealed and very dirty oil.  There was no obvious damage to either when the oily sludge was cleaned off but again, cleaner candidates will be available from my Marvel parts stash when it comes to reassembly.

This marks the half-way point and an extended agitation for all parts in the watch cleaning machine.  Locating reverse, we start with the mainspring, cleaned and wound into a suitably sized drum.

Some care is required in pressing the mainspring back into the barrel because the tail needs to be aligned with the slot on the inner barrel wall.  Reassembly of the movement itself starts with the setting parts.

The centre wheel and its bridge come next, the completion of which was hampered by my serially shearing the tops off the tiny screws that secure the jeweled sweep seconds bearing plate.

I did this three times without exerting anything other than the slightest degree of torque and in the end resorted to a more robust screw harvested from a Seikomatic movement.  With that done, the centre wheel bridge can be fitted.

You may have noticed that the escape wheel is served not by a Diafix setting, but by a jeweled bearing capped off by a jewel supported at the end of a figure-of-eight metal plate. I oiled this by placing a drop of oil in the centre of the cap jewel and then lowering it into place before fastening down.

The reconstruction continued smoothly right up to the point where I was to refit the balance.

Unfortunately, the hairspring on the balance that came with my replacement movement was distorted out of plane with the result that the spring would not sit flat.  I was unable to correct this problem and, in my attempts to do so, I fear that I may have made matters worse.  However, the correct balance wheel was in stock at Cousins and so I placed an order and waited a day or two for the replacement part to arrive.  Unfortunately, this ‘new’ balance wheel was supplied with the roller and its jewel missing.  To make matters worse, the spring on that one was also bent out of flat.  With gritted teeth, I ordered a second.  Happily, this one arrived with everything present and correct and the hairspring sitting true.

One of the challenges in regulating movements with fixed stud holders is that there is no easy way to adjust beat error.  In my first measurements of the running of this movement with its original hairspring, I was seeing a very large beat error of 3 ms.  The beat error is a measure of the difference in the time it takes for the impulse jewel to swing from each of its two turning points to the mid-point between the banking pins.  In other words, it is a measure of the difference between the tick and the tock.  The beat error is minimised by adjusting the position of the impulse jewel so that it sits half-way between the two banking pins when the balance is at rest.  In the photograph below, you can see that at its equilibrium position, the impulse jewel is offset some considerable distance from its ideal position.

In a watch with a balance fitted with a moveable stud holder, this position can be adjusted on the fly in a fully assembled movement as part of the regulation process.  However, for a balance with a fixed stud holder, the only way to adjust beat error is to remove the balance wheel from the balance cock and adjust the position of the hairspring on the staff by adjusting the collet position one way or t’other.  This necessarily requires a degree of trial and error.

Having worked my way through this process with the new balance, I was able to achieve a beat error of 0.0 ms and an amplitude of about 280 degrees.

With the movement now ticking away happily, I turn my attention to the dial, which you will recall was looking a bit the worse for wear.

Quite a lot of the patina is clearly the result of abrasion-marking from its close proximity to other stuff in storage combined with the accumulation of a healthy layer of grime.  My default approach to dial cleaning of glossy or satin dials, particularly those of a light colour is to work the dial gently with virgin Rodico.  With the dial supported on a piece of pithwood, I set to work.  Here we are half way round.

That’s looking pretty good.  Continuing from 6 to 12 results in a presentable dial ready to be fitted to the movement.

I think we are ready to fit some hands.  I had an embarrassment of riches to choose from in the gilt Dauphine hand department to pair with the existing hour hand and had soon selected and fitted a suitable set.

You will remember that at the outset, we were missing pretty much everything, including a case in which to house the watch movement, dial and handset.  I wanted to pair this dial with a steel case but none of the three cases I had from the box o’ bits stash were compatible with the movement or with the dial size or, most importantly, both!  And so I had to turn back to Yahoo to see if I could source a suitable case.

In my attempts to identify the model that this dial would originally have been fitted to, I found one image of a Seiko 14045 with this exact dial/handset combination.  However, that model reference is for a gold-plated case and so I set about seeing if I could find a 14044 that I reasoned might be the steel equivalent. As luck would have it, I located an auction for an old stock 14044 fairly easily and landed it with the starting price valuation.

The only fly in the ointment was that it did not come with a crown, and the crown tube diameter on this case was much larger than is typical for this style of watch.  And so I had to wait for the case to arrive, measure the tube diameter and then try my luck at finding a crown to fit.

Before fitting the movement to the case, I test-fitted the bezel and crystal to the dial to make sure that both hour and seconds hands would clear the inner surface of the domed acrylic crystal.

Notice that both the minute hand and seconds hand have a pronounced downwards curve to avoid contact with the crystal and to follow the convex shape of the dial.  With no contact being made, we are ready to fit the movement to the case.  This involves lowering it in from the dial side, securing the movement into the case with the pair of case screws from the balance side and then fitting the crystal and bezel.

The crown sitting to the right is a period correct Seiko crown but comes with no water proofing.  The best that I can hope for is that it will be dust proof, but in the meantime, I may keep my eye out for a suitable waterproof crown.  With the stem trimmed and fitted to the new crown, we can complete the installation.

The case back snaps into place under determined pressure from my thumbs and it appears we have a complete, ticking Seiko Marvel.

With a suitable strap fitted, the watch is transformed into something really rather classy.

I was half expecting this to end up feeling a little underwhelming, toy-like even, given the typically awful Seiko renderings in their 1956 catalogue, but the reality is that the watch conveys a genuine air of quality, matching the hyperbole of the time.

The arithmetic result of this piecemeal process is comfortably greater than the sum of its parts, with that impression aided, perversely, by the weathered appearance of the dial.  An unpromising premise to this particular project has, at least from my perspective, born fruit.

This then is the origin of the species: the common ancestor of the modern Seiko wristwatch.