The Citizen Auto, 1958

Introduction & Background:

This article is about an important watch for the Citizen Watch Company – it’s first automatic wristwatch, launched in May 1958. Simply known as the Auto, it had a short production life of just 3 years or so, since Citizen was developing at that time its ‘Jet’movement which became its mainstream automatic from October 1961.

The mid to late 1950s saw rapid development in wrist watch production, with Citizen introducing Japan’s first shock protection system in 1956 (‘Parashock’), and Japan’s first waterproof watch in 1959 (‘Parawater’). The ‘Deluxe’ hand winding models proved very successful, achieving sales worth over ¥100,000,000 following its introduction in 1958, whilst in 1952 Citizen had made Japan’s first date display watch – the ‘Calendar’. However, Seiko, by far Japan’s biggest watch manufacturer at that time, had introduced Japan’s first automatic watch in 1955 and in 1959 launched the ‘Gyro Marvel’, using the ‘magic ‘lever’ mechanism which allowed winding by the weight (rotor) in both directions.

Although Citizen were to launch the popular Jet range in the early 1960s, the company needed  an automatic to offer a rival to Seiko’s range sooner rather than later – and before the Gyro Marvel was brought to the market. And so Citizen’s  first automatic was launched in the spring of 1958. It also featured bi-directional winding, a year ahead of the Gyro Marvel and was competitively priced compared to the Seiko.

Although the Jet automatics are relatively well known in Citizen’s back catalogue, the 1958 Auto which preceded them is much scarcer and very few examples are to be seen, even in Japanese vintage watch auctions.

This article tells the story of my example, from its purchase last year to its return from restoration just a week or so ago. I will say right now that I am enormously indebted to Brian (aka ’31 jewels’) who undertook the restoration with great enthusiasm and interest, and who was able to guide me in my search for a donor movement. However, the story is not just a simple tale of a new addition to the collection –  it also solves a puzzle about the history of the movement, and unearths an unforeseen connection  with a Frenchman called Gerard Langel….

My Citizen Auto:

The 1958 Auto is a model I very much wanted to add to my collection, but I was reconciled to not acquiring one given their rarity and high prices when they had very occasionally appeared in Yahoo Japan’s auctions. Then, in the autumn of 2013, I came across an example for sale (in Japan) which clearly had an issue or two. As far as I could determine from the sometimes flaky on-line translation, this one had no second sweep hand (that was obvious enough from the seller’s pictures though!) , but also maybe some other missing parts. It appeared however to be in rather good condition, with just some tool marks on the back detracting from an otherwise very decent cosmetic condition.

So I decided to bid, partly with a naive hope that all it needed was a second hand, but more realistically so that I could find out more about the base movement and see if I could source a donor movement for any parts that may be needed.

 I was fortunate to win the auction and a few weeks later this arrived:

I certainly wasn’t disappointed with the condition, especially the dial and hands, and the movement wound by hand and ran. But I quickly found that there was more missing than the second hand….

The 3KA Movement:

I had long known that the 1958 Auto movement was called the ‘3KA’, and was produced with either 20 or 21 jewels. I had never seen a picture of the base movement, only the rotor and underside of the automatic winding mechanism. Whatever was to happen with my new acquisition, I would at least be able to have a look at what lay under this rather nicely finished swinging weight rotor:

Other than the 3KA moniker, I also understood that this movement used a hand winding movement as its base, probably the ‘3KS’. The 3KS was first made in 1955 or 1956, and was the first to use Citizen’s improved measurement and machinery capable of 1/1000mm accuracy. As I researched the Auto I had found the 3KS described as a woman’s model. This could explain why it is not particularly known to collectors, who of course focus almost entirely on men’s watches.

Under the rotor is a well finished full bridge to which the auto-winding mechanism is attached:

Although there will be more of the auto winder later here is a quick look at the underside of the bridge:

Here is the base movement, as far as I know this might be the first image on the net showing the 3KA base calibre! Clearly it’s on the small side, but I am not convinced it was only used in women’s models since up to the 1950s men’s wrist watches were often small, especially by today’s standards. Although launched in 1958, the use of a pre-1956 movement is evidenced by the lack of shock protection:

Even though I am not a watchmaker, I quickly realised that the gears for the second hand are absent. Time to call in the cavalry! 🙂

The Watchmaker’s Assessment:

I was happy that Brian agreed to take the watch on, and in November 2013 the Auto was with him. There were two main issues: first the gear for the second hand was missing along with a small driving pinion that would press onto the pivot sticking out of the relevant jewel, and second the main (very special) screw holding the winding gear was broken:

  

Brian was absolutely positive that he could repair the screw – so all we needed was a donor movement….

The Search for a Donor Movement

Anyone who knows vintage Citizens will be aware that parts are not exactly readily available 🙂  The Auto is a scarce model to say the least (I think I’ve seen three others in the past five years or so) so finding another one as a donor was unrealistic. However, since Brian could repair the screw and that otherwise the mechanism was in good condition, we agreed we were in the hunt for a 3KS hand winder for the missing parts.

I was able to find a starting point for the search from the ‘Museum of Japanese Made Watches’ which includes many examples of vintage watches from all Japanese manufacturers. Two models are shown, a ‘3K’, with sub-second hand, and a 3KS with centre second – that movement was the target, and the book has a small image:

So followed many hours of patient searching of Yahoo Japan….then I found what might be the correct model. As is often the case with Yahoo auctions, there was no movement picture, but based on what I knew by then, this looked a good bet and it was soon in my hands. Crucially the model / case number looked correct:

 

And the movement:

Although it looked like the right one to me, more importantly Brian said yeah! 🙂

The Restoration:

Once he received the donor Brian confirmed that all the parts were available from the donor, so, in April of this year (2014) he started work – a complete strip down and service as well of course as the fitting of the donor parts and repair of the broken screw.

Last pics of the complete donor watch, with the Auto:

 

Dial sides are the same, manual on the right:

Here you can see an added gear on top of the crown wheel on the Auto, to enable engagement with the auto winding mechanism:

Now all stripped down, and Brian finds that the parts are in good order:

 

I sourced the second hand from a different donor, this Center Second model. Brian later found that this wasn’t an exact match and needed to downsize the tube slightly to fit the pinion:

Out of the case the dial is in excellent condition, with just a tiny outward dimple near the four o’clock marker:

Brian was happy that the gear train was in good order:

Brian now set to in re-assembling the movement and install the donor parts. Note the ratchet wheel has a factory pin hole in it. It’s unique to the auto, and not found on the manual:

The auto winding gear is stacked and screwed onto the main ratchet wheel, and this engages with the auto winding assembly. This is a high torque area for this watch. This, as Brian put it, is where things got interesting. The manual sweep hand’s bridge did not fit – it was too large in the fact that the manual does not account for the larger stacked winding ratchet wheel:

Brian had to carefully machine the side of the bridge down to allow space for the winding ratchet wheel. Although a tight fit, it clears the ratchet wheel. The second hand drive gear was then installed on the movement using the Auto’s sweep hand spring to hold the pinion. Brian finished the assembly of the base movement, fired it up, and found it ran perfectly:

Before taking a closer look at the auto-winding mechanism, one other part Brian fitted was a new crystal. The old one was clearly a replacement and slightly undersized. A domed crystal is needed – here you can see the curve of the dial. There is very little clearance between the minute and second hands, so fitting the right domed crystal is crucial:

So on Brian’s recommendation a very nice new compression fit crystal goes in:

Brian was able to drill out the broken screw and found that there was enough thread remaining to use, and with the addition of ‘JB Weld’ the auto-winding is back in action:

The bridge and rotor are fitted:

And finally Brian signs the inside of the case back 🙂 :

Timing, Winding and Reserve:

Following service and re-build, Brian let me know how the watch was performing on the bench. First is timing (i.e. accuracy in different positions). Brian noted that there is a daily variation of about 20 seconds between crown up and crown down positions, and about 25 seconds between dial up and dial down.

To get this watch in a very accurate ball park Brian would need to wear the watch for a week plus – but of course he wouldn’t wear a customer’s watch so that will be down to me. His recommendation is to wear the watch for 2 weeks or so, and I can use the +/- adjuster to improve accuracy. However, we suspect that a daily variation of less then 30 seconds for a 50+ year old watch is not too bad, and may anyway be within the original manufacturing tolerances.

Although an automatic, this watch can be hand wound, which is a nice feature in my opinion to get it started and fully wound quickly. However, Brian recommends a maximum of 40 slow winds. With no spare mainspring (the spring in the manual wind is not suitable)  we do not want to risk breakage. With 40 winds there should be a reserve of about 25 hours – in fact I found it to be 27 hours after my first (slow and careful) winding.

Brian confirmed and tested that the auto assembly wound the main spring. Interestingly, how much it wound it is dependant on how many times you wind it manually to get it running from stopped. As the main spring tightens, so the rotor will slow down too. Brian observed that you can hear this watch wind as its worn, adding that it ‘sounded like a Seiko actually…those pawls and teeth’.

The French Connection:

Before looking at the complete restored watch, let’s take a closer look at the auto-winding design. In simple terms the mechanism allows the oscillation of the rotor in either direction to wind the watch via a pair of pawls which engage either side of a ratchet wheel. Here are diagrams from the original patent registration – first a section ‘through’ the rotor:

And now from the underside:

The patent registration also included an alternative design:

This second design looks very much like Seiko’s ‘magic lever’, introduced in 1959. The inventor described a number of advantages of these designs – better efficiency (bi-directional winding), cheap to produce and easily fitted to a modified existing movement.

And now to the intriguing part 🙂  This design was registered with the USA and other patent offices in 1954, based on an application made in 1951. And here is the French connection – the inventor was Gerard Langel, who worked for the ‘Ultra’ watch company in Besancon, France:

Although the Ultra company closed down in the late 1960s or early 1970s, they produced a ‘Superautomatic’ model in the mid-1950s, using their 703 movement. It is featured in ‘Complicated Watches and their Repair’ by Donald De Carle:

During my research I learned that Ultra watches were sold worldwide including Japan, and the owner of the Ultra company travelled to Japan. The Ultra Superautomatic and the Citizen Auto clearly share the same design of auto winder. This begs questions about the use of these designs by both Citizen and Seiko, which I can’t answer at the moment. So far I have found no references to any relationship between Citizen, Seiko and the Ultra company.

Additional Information, 22nd April, 2020:
Further to my own conclusions about the origin of the automatic winding mechanism, I have recently found that the French Ultra company has been resurrected (it closed down due to bankruptcy in 1972). And on its home page there is reference to their mechanism, used in their Superauto watches, being sold to Citizen for use in the 1958 Auto. It also refers to the magic lever being designed some 15 years before Seiko adopted it for their automatics (scroll down to see the information): https://ultra1911.com/

It’s good to have this history confirmed, rather than leaving it as just my own conclusions/assumptions.

And Finally:

This has been a fascinating project, from the first sighting of the Auto on Yahoo Japan last year (2013), through the process of correctly identifying the base movement and on to Brian’s excellent work to restore it in last few months. As Brian put it to me, we have saved an interesting piece from a parts watch fate. And on top of that has come the information about the use of a design invented by a Frenchman…

The best bit of all this of course has been safely receiving the fully restored watch back from Brian. His skills are second to none, and it has been great to see how he has applied his skills and ingenuity. I am also very grateful that he recorded the restoration photographically so it can be shared here.

To end this page, here are a few of my own shots:

Copyright: Please do not copy or reproduce any part (text or images) of this article without the express permission of the author

© Stephen Netherwood, 2014