In the ten years that followed the release of the world’s first production quartz watch in December 1969, the cycle of rapid development and refinement transformed this new time-keeping technology from artisanal science-project into slick, mass-production high-end state-of-the-art haute-horology. The Seiko Astron’s 35A movement used a quartz crystal oscillator operating at 8192 Hz but by the following year had developed into the 3502 running at twice that frequency. The 3823A VFA from 1973 employed special condenser-controlled temperature compensation and was specified to lose/gain no more than 5 seconds per month.
By 1975 the 38 series was being phased out and replaced by the Suwa-produced 48 series and the Daini-produced 08 series, both of which now running at the industry-standard 32768 Hz for stepper-motor controlled quartz watch movements. The 48 and 08 series were specified to a monthly rate of ±10 seconds per month in King Quartz variants, improving to ±5 s/month in the 48 series Grand Quartz and up to ±1 s/month in the Quartz Superior models.
The 48 series ruled the roost until late 1978 when the Seiko’s first twin quartz temperature compensated movement was released. The introduction of this movement marked the point at which Seiko measured accuracy not in terms of a monthly rate but in terms of an annual rate. The Grand Quartz 9943 was rated at ±10 s/year and the Superior Twin Quartz 9983 at ±5 s/year.
It is notable that this rapid development, coupled to improvements in performance, came with a reduction in mechanical design complexity. Those early, pioneering quartz movements are dazzling examples of brilliant engineering ingenuity, requiring some serious chops to set up and adjust correctly but by the time the 99 series was released in 1978, the construction and design of these movements presented far fewer challenges to the watchmaker at service.
A quartz oscillator plays the same role that a sprung balance does in a mechanical watch movement in regulating the distribution of impulses to a gear train. However, the much, much higher frequency of oscillation of a quartz oscillator renders it essentially immune to external perturbations and, furthermore, eliminates the positional variation that afflicts a mechanical watch balance. A quartz crystal is also much less sensitive to temperature than a sprung balance, but it is not entirely immune. Typically, quartz watches are set to run slightly fast at a typical wrist temperature. This is because the rate variation with temperature plot is a parabola whose maximum is tuned to wrist temperature and so variations in temperature above or below that point will both result in a reduced rate (see figure below, left). Typically, a 10 degree reduction or increase in temperature will result in a change of up to 100 seconds per year in rate. It is for this reason that watch manufacturers specify the typical loss/gain assuming that a watch is worn on the wrist for 8 or more hours per day, thereby maintaining a fairly constant temperature.
A number of different strategies have been used to compensate for temperature but perhaps the most effective and least demanding in terms of compromise and power consumption is the twin quartz approach (see illustration below). In movements employing this design, the temperature compensating circuit consists of a compensating oscillator circuit equipped with an auxiliary quartz crystal oscillator and an oscillator regulating device. A detective circuit determines the watch temperature by monitoring the difference in oscillating frequency between the master and auxiliary quartz oscillators. The compensating circuit then corrects the timing rate according to the detected temperature (above, right).
I think it is time to introduce the subject of today’s entry: a very nicely patinated example of the Seiko Grand Quartz 9943-8000, whose 1978 catalogue entry (shown in the catalogue photo above, top row, second from left, model code GQB824). Those of you who have scrutinised the catalogue image may have noted the eye-watering prices of all of the 99-series models, but in particular of the two Quartz Superior watches. These two were priced at twice that of the celebrated 6185 Grand Seiko VFA watches that appeared in the special luxury catalogue published just 6 years previously in 1972.
The original price of the Grand Quartz 9943-8000, fitted with a bracelet, was 120,000 Yen. My example is fitted to a properly ancient and dried up brown crocodile leather strap complete with an authentic GQ branded SGP buckle. It may be that the strap was provided as an additional accessory in the original packaging for I can find no reference to strapped examples of this watch in the Seiko catalogues of the time. The case is gold cap rather than SGP and all the better for it.
The dial and handset are in very good condition, although there is a little corrosion evident around some of the applied minute markers and a splodge at the edge of the dial at the 44 minute marker. The case is in excellent condition on its upper side but the hooded lugs on the underside have not fared so well, that leather strap having provided a cosy spot for accumulation of moisture and consequent corrosion of the steel base.
I am not too worried about the rust though. I plan to clean that out and passivate with boric acid. The serial number on the case back dates the watch to June 1978, pretty much bang on the first appearance of this model in the second volume of the 1978 domestic market catalogue.
The case construction is rather unconventional. The movement is housed, from the top, in its own steel monocoque sub-case, which is pressed into a separate mid-case from beneath. Initially, when tackling the dismantling process, it is easy to be fooled into thinking that this is a conventional case with a press-fit case back but when you lever the case back up from the gold cap case, the whole sub-case assembly housing the movement comes free.
The next question that presents itself is how to remove the movement from the lower monocoque.
The crown and stem are still attached and initially there appears to be no obvious means to free the crown. However, removing the rubber gasket that sits around the dial edge reveals the secret to gaining entry.
Depressing the stem release lever whilst pulling on the crown frees the crown and stem and allows the movement to be partially lifted from the monocoque.
The movement, dial and hands can now be separated from their home for the past 43 years.
With that head-scratcher out of the way, we can focus on the relative routine of removing the hands and dial and taking a quick peak to the rear of the dial to see its date of manufacture.
The numerals 83 indicate that the dial was produced in March 1978, three months before the rest of the watch. The calendar side initially presents no difficulties in peeling away the layers to access the setting mechanism. The only slightly usual feature is the presence of two equal sized dial guards, sitting on top of one another, one hosting the day jumper spring and the second, lower plate, the date jumper.
In contrast to my experiences grappling with the gear train sides of the very complex 38 and 08 series movements, the 9943 could not have been more straightforward. The magnetic shield is removed from the coil, then the plastic cover from the circuit, then the circuit itself, revealing the train bridge and coil.
Removing the train wheel bridge and train wheels reveals the centre wheel bridge, under which we find the centre setting lever.
With all the remaining parts removed, the movement and all mechanical parts could be transferred to the cleaning baskets for cleaning. The plastic parts, electronic and electrical components and the magnetic step rotor were excluded, the latter to avoid debris from the cleaning process adhering to the magnet. I should also note that the two arms of the rotor stator (the two parts that provide the electrical impulse to the step rotor) are fixed in place with pins and consequently, we can avoid the adjustment headache that seems part and parcel of setting up 38 series whose stators are both removable and adjustable.
While the parts are cleaning, let’s break down the case. The construction of the upper mid case is similar to that of the 0853 King Quartz that we met a little while ago. The crystal is mounted separately in a removable bezel assembly which is pressed into the case aperture, sealed with a nylon gasket. Consequently, we have to remove the bezel plus crystal as a unit by pressing it out of the mid case from the rear using a crystal press.
The crystal is sealed into the bezel surround against a rubber o-ring and held in place by a separate ring sitting to the rear. To remove the crystal, we need to press it and its securing ring out from the top.
The crystal is flat tempered mineral with a frosted bevel and benefits from anti-reflective coating to its rear.
The crystal needs replacing though and I will need either to secure a correct Seiko replacement or improvise as I had to in the case of the 0853.
While I’ve been occupied with the case, the movement parts have been cleaned and dried and are ready to be reassembled. The process is simply the reverse of that documented above, with one or two hops, skips and jumps to comply with instructions as directed by the technical manual: starting with the centre wheel, second setting lever and centre wheel bridge, we then flip over to fit the cannon pinion and the stem, clutch and intermediate wheel for calendar correction. You will notice that I’ve set the stem unlocking lever in position in the top right photo below in preparation for then fitting the setting lever and yoke (bottom left). Finally, the setting and minute wheels are placed and the whole lot are fixed into place with the setting lever spring (bottom right).
The gear train comes next, starting with the step rotor (cleaned with Rodico), fifth wheel, third wheel and then fourth wheel.
I always find it tricky to set the train wheel bridges on quartz movements and this one was no exception, but the process is made a great deal easier if you have access to a stereo microscope to make sure all the pinions are correctly seated.
The battery insulator, coil and circuit block come next.
And we conclude this side of the operation by fitting the green plastic circuit block cover and the anti-magnetic shield plate.
The movement reassembly continues by reconstructing the calendar parts.
At this point, I supplied some juice from my quartz watch testing machine and was gratified to see some action from the fourth wheel. The final testing and regulation will wait until the movement is cased. Satisfied also that the calendar quickset was working and that the calendar changed over correctly at midnight, I refitted the dial. Take a look at the mirror polishing to the sides of those slab-like hour markers.
The movement is now ready to be reunited with the case, but the case is not ready to receive it. Well, in fact, the lower monocoque case is almost ready but the upper case is still in a state of undress. To rectify the latter, we need to reassemble the crystal/bezel unit but to do that, we need to locate a) a new crystal and b) a fresh crystal gasket. The original grey gasket had split and so I set about identifying the closest match that I could from the selection of generic o-ring gaskets from Cousins in the UK. The best match that I came up with was a 30.80 mm outer diameter item with a thickness of 0.5 mm. Somewhat to my surprise, I was able to source a brand new, AR-coated Seiko original crystal (part number 300WB4G00) from a UK supplier and so we are all set to fit the crystal to the bezel assembly. First, we fit the gasket to the inner groove in the bezel, seat the new crystal and then press the retaining ring in from the rear using a crystal press.
You may remember from earlier that the hooded lugs of the upper midcase were in a state of some distress, both coated in a significant layer of surface rust. I was able to clean this up to a satisfactory level and then passivate with some boric acid and I think this should now resist further corrosion as long as the previous conditions are not duplicated (rotting, damp leather strap in a hooded lug + time = rust).
Continuing, a fresh battery hatch gasket finds it place, the hatch is fitted, the plastic movement ring is inserted back into the base of the lower case and we are ready to receive the movement.
The movement (hands refitted) just slides in from the left, a little jiggle to settle, and we have made the first step to returning this to something approaching a watch again.
The original nylon gasket that seals the crystal assembly to the upper mid-case was damaged and so I had needed to source a replacement. I was not able to find a correct Seiko original part and resorted instead to trying a generic part whose dimensions most closely matched the original. A bit of trial and error and I settled on a 1.2mm high gasket (in truth a little tall) with external diameter 32.3 mm and internal 31.5 mm.
If the slightly excessive height of the gasket was going to be problematic, it would reveal itself when pressing the unit home into the upper mid-case.
Everything looks tickety boo and so we press on. The crown gasket for this model is a slightly unusual D-shape with part number DB0060B01. This gasket has a smaller diameter than the far more commonly-used DJ0060B, and sourcing one held up proceedings for a day or two.
With that done, I could refit the crown and finally fit the seconds hand.
At last, we are in a position to complete the rebuild and press the lower monocoque, complete with movement, dial, hands and crown, into the fully assembled upper mid-case.
The complete watch looks great, the extraordinary design details shining through the now clear window to its dial and handset.
I was curious to see how the freshly serviced movement was performing and so with a fresh battery installed, I placed it on the timing pad of my Horotec Flashtest.
+0.4 sec per day is not especially impressive for a twin quartz movement specified to be accurate to ±10 sec/year. However, I still needed to regulate the movement, something which can be achieved by adjusting the trimmer condenser, accessible via the battery hatch.
A small rotation of the trimmer screw anti-clockwise and I tested the watch once more.
Admittedly, a bit of a fluke but 0.00 sec/day will do me nicely. All that is left to do is fit a strap. In attempting to duplicate the look of the watch as received, I selected a brown lizard strap.
I’ll be honest and say I’ve never much got on with brown straps, but I think this looks pretty good (although I will confess to switching it to a black strap once I’d completed the photography for this project).
This is very much a watch of its time and for a long time this type of mid- to late-70s quartz watch would have been hugely under-appreciated, dismissed as one of a large number of anonymous quartz dress watches churned out by Seiko throughout that decade.
It is only recently that these watches have come to be properly appreciated for what they were and are. Their relatively astronomical prices at the time look ludicrous in hindsight but that assessment ignores the superlative engineering quality, design and quality of materials used in their construction. All of the Seiko quartz watches I have worked on from this era have been properly impressive and hugely enjoyable to discover. The 1970s was a decade of contrasts in terms of tastes and style but these superficially unassuming watches now appear, with the benefit of the passage of nearly 50 years, to be just as remarkable an achievement as the more obvious mechanical Seiko icons from the 1960s and 1970s.
Your correspondent, a convert to the charms of the front line, cutting edge output of the quartz revolution.
Great write up! You have helped shed more light on the inner workings of the 99xx caliber than many others have. Just wondering if you’ve needed to touch the other trimmer (the hex shaped one) to get the watch to 0.00 seconds? This is also probably related to the watch’s ability to keep excellent timing if temperature fluctuates.
The technical manual states that the second trimmer should not be touched and I followed that advice. I was able to regulate to 0.00 with one small rotation of the main trimmer only. As I suggested in the post, a fluke!
Lyndsay Wilson (Perth, Western Australia) said:
Great work, as always, Martin!
Michael Glen said:
I enjoyed the post as always. This watch oozes opulent 70’s quality in a lovely gold and tobacco brown sort of way. The patina is an interesting phenomenon. I thought higher carat gold alloys didn’t really form a patina as such, especially in an object subjected to wear like a watch. This one, however, does seem to have attained a pleasant mottled appearance in places. I wonder what process causes this?
I think any gold alloy has the potential to oxidise to some extent, depending on the conditions. The steel lugs on this watch had quite a lot of surface corrosion and so I would imagine conditions had been ripe to tarnish the gold cap a little. Glad you enjoyed the article!
Hi Martin! Amazing post as usual! I have got this watch too. Did you know that this very model was still sold in 1985… I found it in the 1985 Seiko catalog.
Also, you mentioned that you used the technical manual for this watch => “[…] skips and jumps to comply with instructions as directed by the technical manual […]”, Would you be kind and let me know which one? I don’t think I have it yet.
Hi Jerome, the technical guide I used was for the 9923A. Freely available as a PDF online. Not brilliant quality but adequate. Thanks for the comment!
Thank you Martin !
Hi Martin, thanks for the write-up! I have a 9943-8040, and it has a lightly scratched crystal. Do you by chance know if they place you sourced the crystal from has an extra? If so mind sharing where you located it?
Hi Aaron, I ordered mine from Gleave & Co but I think it was the only one they had in stock. A search of that part number now on Google only finds my article.
Increible articulo rebosante de conocimientos, muy ameno de leer. Felicitaciones!
Terri Piszko said:
Hi Martin! I saw a post from you from 6 years ago on the Seiko Sports 6139-7020. Funny enough by stepdad’s name is Martin and he’s looking to sell his Seiko Sports 6139-7020. It’s automatic and it dies move when on the wrist. I was able to set the date but not the day. It’s set to Sunday. In any case would you know how much he can sell it for?
Hi Terri, that really depends on its cosmetic and operational condition. I’d need to see some photos and know whether all the chronograph functions work correctly and what its timekeeping is like. But an idea of external condition would be a good start. All the best Martin
I have seen these referred to as SEIKO GROUND TWIN-QUARTZ 9943-8000, is that accurate?
Hi Phill, I don’t know where the word ‘Ground’ would apply. These watches are part of the Seiko Grand Quartz range and so that is the term that I would have thought best applies. They are twin-quartz movements but that term is not used anywhere in the branding displayed on the case but is instead indicated through the double-quartz motif that appears on the dial beneath the centre hole.
Brilliant write up
Great article. I bought a 9943-7000 the other day. Great watch, super condition. But the accuracy is not so good. Perfectly adequate, but ~+1.5 seconds a day is far from perfect.
Would you say that I can try to regulate carefully without timegrapher?
What would you recommend as a starting point? One full rotation, one half rotation anti-clockwise?
Hi ark, regulating without a timegrapher will be a tedious process but possibly worth a try. I would start with very small adjustments and then play it by ear, depending on the magnitude and direction that your adjustment achieves. The accuracy you are seeing may suggest that a service is in order but no harm trying to regulate first.
Thanks a lot! I’ll give it a try.
Short update: The first try resulted in ~+0,2 sec a day. That is fine enough for me to not fish in the dark any further. Thanks again and keep up the wonderful work here 🙂
Good job! Well done.
I truly enjoy reading each and every entry. They allow for an appreciation of watches to a greater degree.
That said, I have been researching (in my wife’s parlance “Geeking Out” on Seiko Quartz movements). Two notes for your validation please:
1. As the watch is intended to be accurate as worn on a wrist (approx. 30 C per Seiko) the watch SHOULD show a plus reading when at room temp (25C). At least that is the case with the 38 series. I am wondering if your original setting of .04 per day was actually the accurate state?
2. There is mention in the Seiko equipment of a need for an adapter when measuring accuracy because the two quartzes may be adding noise?
Once again, enjoy your work and writings.
Hi Michael, on your first point, it may well be the case that timing is best adjusted to run slightly fast at room temp on a timing machine to compensate for the likely slightly slower running on the wrist but if you are going to do to those lengths, you probably also need to allow for how much likely time the watch will spend off the wrist in a given 24 hour cycle and at what temperature. In my case, in the winter, the watch could be spending its off-time in the chilly bedroom temperature of a drafty Victorian North Yorkshire bedroom. Given that I wear my watches on a whim from one day or week to the next, I am not so concerned with chasing absolute accuracy. As long as it tells the time more or less correctly over the few days I may be wearing the watch, then I am happy.
As for your second point, I am afraid that I don’t know the answer – sorry!
Is it possible that gold Grand Quartz have a rare version with a slightly darker dial? I found one in recently, after the 9943-8000 there is an “R” character instead of “T”. Year: 1978
Quite possibly. I do not have an exhaustive knowledge of the different model variations but a Google image search might come up with the same variation.
Good evening Martin,
I have finally found a decent 9943 for a birth year watch, according to the serial number it’s a March 1978. Possible manufacturing stock before release.
Your articles have been a pleasure to read and have helped me with restoration of a weekdater. But this Xmas puzzle is currently hold, as need to borrow a crystal press.
I am glad that my articles have been of interest. Good luck with your 9943. It sounds like a great project
Martin A Jones said:
I have read about the merits of the gold cap, as it is much thicker than the standard gold plate. Is there any way to repair or replate if the gold on the cap has worn through?
I have never seen gold cap wear through. I don’t see how it could happen just through normal wear and tear, even over decades of use. The watch would have to be in some kind of major calamity or to have been regularly machine-polished. That doesn’t answer your question though but I don’t think there is an easy answer. My view is that the whole point of gold cap cases is that you don’t have to concern yourself with such worries. That is a very long-winded way of saying ‘I don’t know’!
I’m looking at an offer of 9940, but I’m a bit afraid that the chip will go defect, and I’ll be left with just an empty watch case. How often does Grand Quartz electronic circuit fail?
I have no information on particular vulnerabilities of any Seiko quartz movement but would venture that most failures will be the result of one of: water ingress, leaking electrolyte from the battery, watchmaker accidents/abuse!
How often does electronic chip fail for 994x series?
Please see my answer to your other question.