, , , ,

My daily cycle into work takes me across one of four Strays, each of which the remainders of common enclosed areas of grazing land allotted to the Freemen of each of the original four wards of York.  The entrance into and exit from this Stray requires passage over a cattle grid.  On rare occasions, my route across the stray into work is aided by a tail wind that allows me to attain a dizzying pace that facilitates a traverse over the grid with barely a perceptible impact on my nether regions.  However, sometimes, usually when battling into a headwind, I find myself behind a ditherer who approaches the cattle grid with unreasonable caution.  With my pace forced to something below a walking pace, I bump and jigger, fillings rattling, and clangers impacting upon the smooth leather of my Brooks saddle.

An analogous set of considerations lay at the heart of the drive towards the development of high beat movements in the 1960’s.  In a typical contemporary low-beat mechanical watch movement, the balance swings back and forth at a rate of 5 beats per second (bps), or, more familiarly, 18000 beats per hour (bph).  Each beat moves the escape wheel forward by one tooth, with two teeth’s worth of progress being accomplished by a complete back and forth swing of the balance.

Animated gif adapted from an original video published by YouTube user 장지웅. (see youtube.com/watch?v=dao0mbAbNLE)

The beat rate of any balance is proportional to force constant of the hairspring (its stiffness) but is inversely proportional to the moment of inertia of the rotating balance wheel. The moment of inertia is a measure of the resistance of the wheel to rotation and is related to the mass being swung about during the rotational motion but also to the square of the radius of the wheel.  So low beat watch movements are characterised typically by larger balance wheels and floppier hairsprings.

While the absolute accuracy of a watch movement is not, in isolation, determined by the rate at which the balance swings back and forth, the ability of the moving balance to recover quickly from perturbations is.  Perturbations can range from something as mild as a change in orientation of the movement, typically arising from changes in angle of the wrist position of the wearer, to shocks, such as those experienced by a cyclist traveling over a cattle grid.

Watch movements are designed to run, most commonly, at integer numbers of beats per second, notionally ensuring that the motion of the seconds between the minute markers on the dial is broken into an integer number of steps.  Thus an 18000 bph movement runs at 5 bps, the seconds hand traveling one fifth of the way between each marker per tick; a 21600 bph movement runs at 6 bps; a 28800 bph at 8 bps and so on.  Curiously there are exceptions to this rule such as the 19800 bph movements used in early Grand Seikos and the Rolex 1570 and 1575 movements which run at an awkward 5.5 bps.

By the mid-1960’s, virtually all mechanical watch movements ran at a steady 18000 bph, 19800 bph or less commonly 21600 bph.  Inevitably, these incremental nudges upward in beat rate served as a driver towards a more ambitious stretching of the envelope and in 1966 Girard-Perregaux developed the automatic 32A movement whose beat rate doubled that of the mainstream standard to a heady 36000 bph (10 bps or 5 Hz).  This movement was an evolution of the manual wind AS 1687/1688 with the automatic mechanism a joint development between Doxa, Eberhard, Favre-Leuba, Girard-Perregaux and Zodiac.

An example of a 36000 bph 42.4 calibre, an evolution of the 32A, in this case fitted to a GP 9086.  (photocredit: Crown Comfort on Watchprozine).

This huge jump in rate represented a very significant technological achievement that relied on much better quality lubricants and a more robust gear train capable of withstanding the greater forces being generated by the powerful mainspring required to drive the balance.  Thicker mainsprings also meant reduced power reserves but in an automatic watch, that is no great sacrifice.

A Girard-Perregaux HF Gyromatic Observatory Chronometer (photo credit: Wikipedia)

Girard-Perregaux submitted 662 cased examples of their ground-breaking movement to the Bureaux Officiels de Contrôle de la Marche des Montres for a 15 day chronometer test.  All examples submitted were judged to have achieved not just chronometer grade performance but results surpassing that benchmark.  40 of these watches were submitted to the Neuchâtel Observatory for a longer test period of 45 days and in recognition of that achievement, Girard-Perregaux were awarded the Neuchâtel Observatory centenary prize.

With all the plaudits raining down on the Swiss, in a quiet corner of the land of the rising sun, Seiko’s Suwa division were busy developing their own take on the high-beat watch movement.  Curiously though, rather than take the obvious route and install a 36000 bph movement into one of their high end models, Seiko chose instead to take a more cautious approach and used the lower key Lord Marvel as the vehicle with which to deliver the World’s second 10 bps movement.  In contrast to the automatic Gyromatic HF of GP, the Lord Marvel movement was developed entirely in house, an evolution of the existing hand-wind 5740B low beat movement whose roots can be traced back to the Seiko Crown 560 of 1959.

The low-beat Seiko Crown movement of 1959, upon which the 5740C is based.  (Photo credit : http://www.grand-seiko.com)

The Lord Marvel 5740-8000 appeared in 1967 and the success of this reconnoitre into the high-beat space prompted Seiko’s rapid development and deployment of a 36000 bph version of the newly designed 61 series movement into its automatic 61 series Grand Seiko later the same year.  The following year, the Daini division launched their fabulous hand wind high-beat 45 calibre, which was fitted initially to King Seikos from 1968 and subsequently into Grand Seikos.

This extended preamble brings us to our subject, a very tidy, but entirely non-functioning example of a Lord Marvel 5740-8000 from January 1977.  Initially, the horse shoe case back lured me into thinking that I’d landed an example produced right at the start of the production run of this landmark Seiko model because the received wisdom is that horse shoe backs were phased out in 1974.  However, the earliest watches were fitted with dolphin case backs and the printed dial codes at the bottom of the dial used the TAD rather than U suffix.*  I note that the Lord Marvel 36000 maintained a presence within domestic Seiko catalogues to as late as 1978 – an impressive 11 year lifetime for a manual wind, no complication watch produced during a period where mechanical watch technology was in its death throes.

The immaculate case back suggests that the interior has not been interfered with on too regular a basis and indeed, the grubby state of the movement is a testament to that conclusion.

The markings on the inner case back appear to indicate that the movement was last serviced on the 22nd June in the 5th year of the Heisei period, which in the Roman calendar, translates to 1994.  Thus we conclude that 25 years have passed since this watch last saw the attention of a watchmaker.

Two features of this movement are worth noting at this point.  First, it does not run at all and the reason for that becomes immediately clear when attempting to wind in some power. My efforts were met with little in the way of resistance, the most likely cause a broken mainspring.  Second, the micro-adjuster wheel on the balance cock is missing.

Withdrawing the crown is all that is required to release the movement from the case, allowing us a proper view of the dial and hands, thus far largely obscured by tape residue on the crystal.

Both dial and hands look to be in excellent condition, barring a minor flaw in the lacquer just beneath the KO of Seiko.

The crown features an encapsulated gasket, which, given its age, is understandably no longer fit for purpose.  In keeping with my recent approach to such crowns, I intend to attempt to remove the gasket and hopefully replace it rather than locating a new crown (which would have been the standard approach at service).

In order to remove the dial, the movement ring needs removing first but one of the screws was welded in place by rust and I could not get a proper purchase.  So I reinstalled the movement into the case to provide a platform to do so.

The dial feet screws are easily located and loosened, and the dial comes free.  The date stamp on its rear dates the dial to February 1977, one month after the case.

Removing the dial reveals the motion and keyless works, housed in an uncluttered landscape typical of a movement free from complications.  We note the presence of two Diafix settings serving the escape and third wheels and quite a lot of what appear to be salt deposits distributed in patches over the surface of the mainplate.

Dismantling the movement follows a path broadly similar to that charted in previous posts on the Cronos, 4420 and, in particular, the Grand Seiko 5722.  Having removed the balance, pallet and Diafix cap jewels, we can pause to note the jeweled bearing serving the barrel arbor.

Recalling the rusty case ring screw earlier, we note staining resulting from the seepage of rust-infused water between the surfaces of the train wheel bridge and main plate.  This is nothing to cause concern because it is not the plate that has corroded but the screw and that can be replaced.  I particularly like the contrast between the un-plated barrel and the plated bridges and mainplate.

Further progress brings us to the point of being able to survey the main scene of crime.

As suspected, the mainspring has broken close to the arbor, preventing retention of any power at all.  Before cleaning all of the movement parts, we need to address the issue of the missing micro-adjuster on the balance cock.  We start our investigations by removing the balance and start the process of dismantling the regulator and stud holder.

I had decided that I might as well completely dismantle the balance assembly because it was clearly very dirty and would need more than just a trip through the watch cleaning machine to rid it of the tarnishing.  In order to replace the micro-adjuster wheel, I needed first to identify the part number.

A quick check on Cousins’ site proved unfruitful but I noted from my watch parts interchangeability list that the part is shared with the higher grade 62 series movements.  I did not have any of these going spare, but a quick dip into the parts catalogue entry for the 6218C revealed that it too uses the 797470.  Fortunately, I have a number of junk 6218’s lying around and I was able to cannibalize the part from one of those.

Following some application of elbow grease, in particular to the stud holder, we are ready to plunge the whole lot into a vat of toxic watch cleaning chemicals.  An hour or so later, with everything clean and dry, we are ready to engage reverse gear.

I start by reassembling the balance parts, the first step of which requires my staking tool to fit the replacement micro-adjuster wheel.

With that done, I can fit, first the stud holder, and then the regulator arm before positioning the Diashock frame loosely into place.

The whole lot then needs securing with a gentle but firm tap from my staking tool.

The final step is to place the balance wheel back into position and perform some minor adjustments to the alignment of the end of the outer turn of the hairspring between the curb pins.

The mainspring part number, 401431 (no longer available) is not the same as that used in the lower-beat versions of the 5740 and is not listed in the Generale Ressorts catalogue.  The 401430 used in the 5740B is listed and the suggested replacement for that is the 4007D.  Cousins recommend the latter for the 5740C but that would surely be underpowered, given the additional grunt required to run that 10 bps balance.  So instead, I ordered a GR4010, slightly thicker, quite a bit shorter but hopefully up to the job.  It is this that we see fitted to the cleaned barrel in the photo below.

Here’s a view of the reassembled barrel for no good reason other than I like the photo.

In common with the 5722B Grand Seiko movement, we note identical jeweling of the barrel:  one for the barrel lid and one for the barrel bridge.  Thus the arbor is jeweled top and bottom but the mainplate barrel hole remains unjeweled and is therefore vulnerable to wear.

The business of reassembling the movement proceeded without drama, other than a minor issue with the positioning of the hairspring stud in the stud holder.  The end result looks rather spick and span.

With a full wind, the movement is hitting 220 degrees dial up or down, beat error is negligible and variation in timing with position low. Satisfied that the movement is healthy, I turn my attention to the case and the fitting of a new crystal.  I had managed to find an old stock original Seiko replacement, but this still needed some adjustment to allow the bezel to be pressed into position.

Some determined effort and a not inconsiderable amount of time saw the successful extraction of the crown gasket from the crown.

A measurement of the crown tube and an estimate of the external diameter of the original gasket suggested a couple of possible replacement gaskets, both of which ordered, and one of which successfully fitted.  With that done, the movement, now fitted with its dial and hands can be reunited with the mid case and we can make a final regulation pass having demagnetized the movement in situ.

Surveying the watch from above at this point proves somewhat gratifying.

That, my friends, is most certainly an improvement on our opening position.  What a fine looking thing.  A suitable strap selected and we can take one last look at the 5740C, looking every inch of its languid aristocratic heritage, in spite of the frenetic pace at which it is going about its business.

I rather like the rather peculiar collision of elements that make up this watch.  A pinnacle of 1960’s cutting edge mechanical horology smuggled into a low-profile middle market model that seems determined to hide its light under a bushel.

But what a light.

* Thanks to Olivier for pointing out in the comments that this is a late rather than early example.  I’ve amended the text to reflect his input.