In the world of the watch cognescenti, Seiko are routinely not assigned quite the credit that they deserve. On watch forums heated arguments flare out of nothing at the suggestion, for example, that a higher-end Seiko model might somehow be spoken about in the same breath as watches emerging from the giddy heights of the Swiss Jura. Horological innovation though does not emerge necessarily just from the top of the market: the position occupied in the public consciousness by Rolex has its roots in Rolex’s record of making no-nonsense, superbly engineered, yet elegant divers and sports watches in the 50’s and 60’s, many of which were widely used as tool watches, instruments designed to serve some sort of purpose other than simply as a status symbol. Firsts widely credited to Rolex include the first waterproof watch, the first self-winding movement and the first watch to display the date. Equally, brands such as Seiko whose centre-of-mass occupies a market position somewhat south of that of Rolex, are responsible too for their fair share of landmarks, often deriving from innovations made in mass-market watches designed for the every-man and woman.
The 1960’s were a fruitful decade for Seiko: I’ve documented some of the truly iconic divers watches made by them through the 60’s and 70’s as well as the automatic 6139 and 6138 chronographs, both of which represent significant landmarks in the development of the wristwatch. In addition then to these more high profile contributions, the Bell-matic slips rather under the radar although to those in the know, they are beautiful, dignified and fantastically engineered watches.
The first Bell-matic, the 4006-7000, was introduced to the Japanese domestic market in mid 1966, fitted with a 27 jewel version of the 4006A automatic movement, the first automatic alarm movement fitted with a centrally pivoted full rotor. The alarm complication at that time was hardly a novelty though, having been first introduced in a wristwatch in the early 1900’s by Eterna.
That watch though was essentially an adapted pocket watch and the first ‘genuinely operational alarm wristwatch’ was the celebrated Vulcain Cricket of the late 1940’s whose alarm was notable in serving its intended purpose of being capable of waking its wearer from their slumber. The first automatic alarm wristwatch was the Memovox produced from the early 1950’s by Jaeger-Lecoutre which used a bumper rotor to wind the movement. In its modern incarnation, the Memovox has been hugely refined and features a full rotor, a feature Seiko introduced with the 4006-7000 in 1966.
These days, of course, an alarm complication on a mechanical watch seems a bit prosaic, but one should not underestimate the power to delight of a chime or tring produced from the movement of a hammer on a resonator rather than as the result of an .mp3 file being played through the speaker of a smart phone. As I have mentioned elsewhere here, my own fascination with the Bell-matic derives from the fact that my father wore one for 20 years or more from the late 1960’s, and it assumed a position in defining who he was at that time as much as the endless chain of odiferous Cuban cigars!
So let’s get to the point of this particular entry. Over the past couple of years, I’ve accumulated five or six Bell-matics, four of which have been sitting patiently in the ‘to do’ pile. Having recently tied up a few loose ends needing my attention, I turned to my youngest son for assistance in selecting the next project and he immediately picked out the watch you see below, a 4006-7021 from December 1969.
This particular watch features a 17 jewel version of the 4006A, a reduction in jewel count over that of the earlier watches made out of necessity, apparently because high jewel count watches attracted a premium import tariff in the US market in the 1960’s making the 27 jewel variant too expensive to sell there. The 27 jewels were consequently reduced initially to 21 in a couple of models sold into the US market, but by mid-1969, all Bell-matics used the 17 jewel version of the movement.*
This particular example makes a decent first impression: the crystal is on the tatty side, the case sports a few dings, the bezel in particular a little ragged. The dial and surrounding alarm ring though both look in excellent condition and so on the whole there is plenty of incentive to make a go of this one. Closer inspection of the case suggests that a long time has passed since anyone ventured into the interior. The case back is in super condition, but the backs of the lugs are filthy:
Removing the case back reveals a mid-case bulging at the seams with the 4006A movement.
The movement looks plenty dirty and with what appear to be a few short bristles or hairs distributed fairly liberally about the place. To remove the movement, we need to release not only the winding crown and stem but also the alarm button. The former is released in the usual way by depressing the setting lever axle whilst the alarm button requires the push bar to be depressed:
With these two out, we are greeted by the full horror of what lurks in the vacated hollows:
Probably best not to linger here too long so instead, let’s take a closer look at the movement, with rotor removed:
Apart from the grime, it all looks exceptionally nice under the hood and as you can see, with barely any power in the mainspring, it can’t help but lazily swing into action. Two key features worth pointing out at this point are the alarm hammer and sounding spring, the operation of which we shall investigate in due course.
With the movement out, the first task is to remove the dial, hands and alarm setting wheel. This unrestricted view of the dial confirms the excellent first impressions – it really is in near mint condition, although the hands slightly less so:
The alarm setting wheel which surrounds the dial is removed by unscrewing the two securing tabs at each side and lifting it off:
With the alarm setting wheel removed, we can gain access to the dial feet screws, remove the dial and hands and take a look at the calendar wheels:
Again, everything here looks in tip top shape. Removing the day and date wheels reveals the unlocking wheel which plays a key role in the operation of the alarm.
The unlocking wheel sits on top of the hour wheel and in the comparison below you can see that the three protrusions on the hour wheel will align at the appropriate point in its journey with the three holes in the unlocking wheel. When that happens, the hour wheel lifts, and that action raises the disconnecting lever beneath, which in turn releases the alarm hammer to perform its task.
You can see how this works in the figure below taken from the 4006A technical document:
The calendar plate comes off next, exposing the somewhat intimidating complexity of the alarm and setting components but providing a clear view of the relationship between the disconnecting lever and the alarm hammer and alarm wheel.
The calendar plate on this watch was swimming in congealed oil but you can still appreciate what a nicely finished component it is, the decoration being entirely hidden from view when fitted:
The disassembly of the calendar side requires a measured, methodical approach, with plenty of opportunity to lose key components or forget the significance of triple-slotted screw heads. Here’s a shot with pretty much everything removed barring the setting components
Over to the top of the movement next, and we can get an idea of how the power is wound into the alarm mainspring and then subsequently transferred to the hammer. In the photo below we see how winding the crown transfers torque via the crown wheel to the intermediate winding wheel and on to the alarm ratchet wheel, thereby winding power into the alarm mainspring. When the disconnecting lever lifts, the alarm mainspring power is released via the alarm intermediate wheel to the alarm wheel beneath whose rotation causes the hammer to vibrate back and forth against the sounding spring.
With the mainplate stripped of most of the remaining components, we see how the alarm mainspring sits openly within the mainplate. We leave it in position for the cleaning step together with the balance cock and balance, minus its diashock setting.
Cleaning, lubrication and reassembly proceeded pretty much without incident, with no parts requiring replacement, nothing getting lost and generally no significant hiccups.
Diashock oiling and fitting to the balance, taking care not to mix it up with the mainplate Diashock setting which has a lower profile
and as you can see from the fourth shot above, she runs! I am inclined to skip through the remainder of the reassembly process as there is a danger of information overload with as complex a movement as this one, so lets pause only to look at both sides, almost complete; first the top side:
and the calendar side, minus the day wheel
On with the day wheel and film washer
followed by the alarm setting wheel and hands
The alignment of the alarm ring requires you to wind power into the alarm mainspring, pull the crown out to the first click and rotate anticlockwise until the alarm sounds with the hands set to 12 sharp. Check the clearance of the hour/minute hands (remembering that the hour wheel lifts when the alarm sounds) and then refit the alarm ring with the pointer aligned with the hour hand. Fix and check again. The tolerance is supposed to be plus/minus 5 minutes but I managed to get mine more or less spot on.
With the case cleaned and a fresh crystal fitted, we appear to be heading down the finishing straight
The movement requires a bit of jiggling to seat happily in the case
and with the automatic winding mechanism and rotor fitted, we can run it through its paces.
The only fly in the ointment at this point was a sticky alarm setting wheel. Out of the case, it rotated with no problems but in the case, it was sticking, eventually refusing to budge. The culprit turned out to be the crystal, a facsimile made by Sternkreuz but whose inner profile and/or steel tension ring was causing the alarm ring to foul. A temporary fix for the moment has been to remove the tension ring
while I wait for a couple of original Seiko 325T02ANS crystals to make their way to me from the USA. I am also waiting for a button gasket to arrive, also from the US but making do for the moment with a generic gasket from Cousins.
As usual then, let’s finish with a few shots of the completed watch:
and three more fitted to a Hirsch Duke
* As I have a 27 jewel 4006-7010 from 1970, I suspect this piece of intelligence may either refer to US market watches or simply to new models introduced from about 1969.
Excellent and fascinating as usual, a great read and a very nice collectable piece as a result.
What a wonderful web page! My watch is a 4006-7010T AD, rather than your 4006-7090T. It appears to be EXACTLY identical, except that it has 27 rather than 17 jewels. I dropped it recently and the day stopped advancing, the minute hand was getting caught on the face and the alarm setting ring wouldn’t turn (though the alarm would still go off). My local watch mender said he wouldn’t try to fix it since spare parts are hard to find so, sadly, I gave up on it. Then I found your webpage and was able to follow along as I disassembled my watch. Nothing was actually broken; the face and alarm setting ring had just become detached from the securing tabs when jolted. I was able to get things set to rights and now the watch works perfectly. Many thanks to you!
Glad to have been able to be of some help!
Thanks Martin for the article. Very interesting. I wonder if the movements are interchangeable with cases? I have a 4006 -6061T (I think) and it has what looks like a specifically made bracelet which is proving hard to replace (there is a SQ deployant clasp I am keen to change but thought I’d see what other options there are). I have a photo but not sure I can post here. Thanks Andrew
Hi Andrew, there is at least some interchangeability between cases (I’ve seen plenty of franken-Bell-matics on Ebay to attest to that!) but you may need to check compatibility of the alarm setting ring. I’ve not attempted any such substitutions myself but have had problems with alarm rings fouling on the crystal and so probably the best thing to do is to give it a try.
david towers said:
Martin Just found your blog very interesting,think your dates a bit out as I bought a Bellmatic in September 1964 in Singapore ( Gold with black face No.4006-6031 Serial No 460158 used it for 25 years, a super watch and has been in a draw for 27 years alarm not working very well is it worth a refurbish?
Best wishes Dave
On the issue of production dates, every source I have consulted suggests a production lifetime extending from about 1967 to the late 1970’s/early 80’s. I’ve seen one or two dating the very earliest Japanese domestic Bell-matic (the 27 jewel 4006-7000) at mid-1966 but nothing at all suggesting they dated from earlier than this. I wonder if perhaps your watch is from June 1974? A watch from 1964 would use the term ‘Water Proof’ rather than ‘Water Resistance’ on the case back, the latter a term introduced from about 1970. I also believe that the horseshoe case backs (as shown in one of the photos above) were not used by Seiko until about 1966/7 and were phased out in 1975. But the most compelling evidence for me derives from the Seiko casing parts manuals from 1969 and 1973, both of which I have. In the 1969 guide, there is no mention of the 4006-6030/31 but it does appear in the 1973 guide which suggests its production started between those dates.
So a few things to check with your watch are:
1) How many jewels?
2) Does it have a horseshoe case back?
3) Does the case back read ‘Water Proof’ or ‘Water Resistance’.
If the answers to the three questions, are 17, yes and Water Resistance, then it must be 1974. If the answers are 27, no and ‘proof’ then we may need to rewrite the Bell-matic timeline!
As to whether your watch is worth reviving, that really depends on its condition and how much you’d like to see it working again. You probably would not be able to justify the cost in terms of adding value but these are great watches and it nice to see them ticking and used and not languishing in the back of a drawer 🙂
Very nice walkthrough. I hope you can help with an issue I have.
I have a 4006-7010 and I removed the movement out of the case to get to the crystal. The stem came out with no issue but the alarm button was a tougher to get out. I had to remove the sounding spring to get to the unlock bar and after I did so, I was able to remove the alarm button.
Now when I put the alarm button back in, it doesn’t click into anything – it just moves freely in and out. The stem goes in as it should, but when it is all the way in, I cannot wind the alarm. Did something get loose/break?
I frequent the SCWF site, so if you have a chance, you can post there. http://www.thewatchsite.com/21-japanese-watch-discussion-forum/129762-bellmatic-help-won-t-wind.html
It sounds like you have one of the Bell-matics with a longer alarm hammer which partly obscures the push bar. From what you describe, I suspect that the alarm setting lever, which is operated by the push bar acting against the setting lever spring, may have become stuck in its depressed state which then means the alarm button does not then click back into position. The ability to wind power into the alarm mainspring depends on the position of the alarm button and with that inoperative, you cannot wind the alarm.
The rust you describe in your Watchsite post may go some way to explaining why the push bar is sticky or has become detatched from the setting lever spring and so I think that the only way to properly resolve the problem will be to remove the dial and clean and reset the alarm setting lever mechanism.
Hope that helps
Stephen Chung said:
Thanks for the input, Martin. I think I may end up selling the watch and take a loss. I have yet to venture into the job of disassembling movements so even the simplest mechanical job is too much for me now. Do you recommend any resource to learn about watch matching? What kind of tools do you use? Keep up the good with your site.
Hello Martin! Love the walkthrough, but I’m a little uncertain to remove alot of parts from the movement and was hoping for your help. I have a 4006A with 27 jewels and it worked just fine, then all of the sudden the minutes hand came loose. I opened the case and with the help of your instructions removed the crowns and easily pushed the minute hand back on. Simple enough, I thought. The minute and is now working perfectly, but the alarm isn’t! All other functions are working perfectly, I can wind the alarm spring, the watch ticks fine, qiuckset date is working and so is date and day switching. When the hour hand meets with the set alarm nothing happends, no click or nothing. Some advice please?
The only thing that occurs to me, assuming you have correctly relocated the alarm setting button, is that perhaps somehow the unlocking wheel has become detached from the hour wheel during your manipulations. Can I ask how you refitted the hour hand without first removing the minute and seconds hand?
When I look at the movement when I set the alarm setting button en the out position I see the alarm hammer get loose so to say. The quick set date with the alarm button is working as well. It was the minutes hand that came loose, I refitted it by carefully pushing it back with plastic tweezers. My amateur guess is, by judging from the instructions of how it works, that somehow the protrusions in the hour wheel don’t aling with the unlocking wheel, thus not activating the alarm. Just as you say, the unlocking wheel is loose or detached. Strange, since I didn’t unscrew anything but the rotor and the alarm sounding spring.
Perhaps the unlocking wheel became unseated as a consequence of whatever it was that caused the hour hand to become detached. Did you check to see if the alarm was still working after the hour hand had become detached but before you opened up the watch?
I solved it! By looking at a pdf of the movement and how to service it (http://www.thewatchsite.com/files/Casing%20Guide/16.%204006A.pdf) I saw that the hour hand is supposed to jump a little bit when it hits the alarm time. I figured that I maybe pushed the minutes hand a bit too far and that it was obstructing the hour hand to make the little jump it needed to trigger the alarm. So what I did was to push up the minutes hand just enough so that the hour hand can do the little jump and set off the alarm. Thanks for the input anyway, and keep up the good work with the blog!
I should probably have thought of that myself – I even mentioned this in the blog entry when describing the refitting of the hands. Well done on a bit of top notch detective work and problem-solving! Glad you got it sorted.
Yes, the alarm was working when the minutes hand was off. It was after I had opened it and reassembled it in the case the alarm stopped working.
Hi Martin, Great advice and fantastic photos here, thanks so much for the effort to post these. I have become quite fond of the Bellmatics and bought two online, a 17 jewel model with great gold dial (4006 – 6040 & 4006 6041). One of these actually works pretty well – the only problem is a slight sticking when pushing in the alarm button which i believe is used to set the date. The other works but the alarm bezel is stuck so think this is rather a more serious problem.
I hoped i could ask you two quick questions:
1. would some lubricant on the alarm button probably sort this out or do you think it would need a full clean and oil? (appreciate you cant see the watch but any advice would be helpful)
2. The alarm bezel fix – if this is fouling on the crystal as you mentioned above and i need a full service any idea what sort of cost it might be? (not keen on attempting this myself as i would probably just break it but dont want to spend a huge amount on it).
Thanks again for any thoughts and great job again – really enjoy the posts and photos. Andrew
I would hope that a clean around the button tube and recess plus a fresh, greased gasket would sort the alarm button issue. It may be that an incorrect crystal is the source of the jammed alarm bezel but if not then it would need, at least minor dismantling dial side to investigate. It should not need more than that assuming the movement is otherwise running smoothly.
Hello, I stumbled across your site a few months back and have read a ton of the past articles. I just picked up a 17J Bell Matic and its in great over all condition, everything works great and its very clean. The only issue is that there seems to be a very very small amount of play between the movement and the case. It causes it to knock just a little bit when the wearing the watch. I can live with the knocking however I wanted to check to see if it was an issue that would cause damage or if there was anything I could do about it.
Thanks for sharing your work with us.
Hi Nick, the movement in a Bell-Matic is held in place by a casing spring which sits around the outer edge of the movement and is compressed by the case back. These springs have a habit of corroding and breaking and I suspect that your issue is probably the result of a broken or missing casing spring. The good news is that these springs are being re-manufactured by third parties and you can find them for sale on Ebay quite cheaply (search for Bellmatic Casing Spring).
As to potential damage, I would imagine limited to possible wear to the edges alarm ring.
Fantastic. Thank you for replying so quickly. There was definitely no casing spring in there when I got it, and I had no idea that there should be. Looking at your pictures again I can see the spring now. Thanks for tip. I have one on the way now.
MARK NOVACK said:
hi martin, what a superb bellmatic disassembly series!!! i have 2 bellmatics & this really did the trick for help when i took mine apart. i have a questions (probably know what the answer will be): both bellmatics worked perfectly except for the day & date not moving. after taking mine apart, i can see that the tiny little pin/post that sticks up to move the day & date wheels is off of the wheel on both movements. is there any chance in heck somebedy sells replacement wheels with this tiny post on it?? also, will this wheel from another common seiko day date work with a bellmatic? i have several scrap seiko day date movements. would love to know. my bellmatics are pristine if i can replace this tiny broken pin/post. thanks so much for any observations, best, mark
Hi Mark, if I understand you correctly, you are saying that the post on the date and date driving wheel on both of your watches is missing. I am surprised that the same problem exists with both as I would not have thought the pin to be especially vulnerable (unless perhaps at some point in the past the wheel has been mounted upside down) but if they are missing, then they are missing!
So, to answer your question, the day and date driving wheel part number is 867805. Both Cousins and Jules Borel report it as obsolete or discontinued but I see that a watch parts seller in Sweden called Mehima SIngh has stock. You might want to send him an email to ask for pricing.
I hope that helps
Charles and Martha Sides said:
I am a pocket watch collector and know little about collectible wrist watches. I do wear a Ball 25j automatic and Omega 24j Constellation. My high school buddy and I have reconnected after being adrift in careers and grandkids for 50 yrs. He has a Bell-Matic he bought on RR from his tour in Viet Nam. He asked if I could get my po cket watch guy to look at fixing it, but he is getting up in years and gave up on it. I would like to get it fixed for my old friend. Could you possibly take this repair on? I know my friend would really appreciate it, as would I.
I am not a professional watchmaker, but rather a hobbyist and the projects featured here are conducted at a geological rate compared to the rapid turnaround achieved by folk to do this for a living. I barely have time plod my way through one project a month and simply do not have the capacity to do more. Having said that, in the UK I would happily recommend either Richard Askham or Duncan Hewitt as people with lots of vintage Seiko experience who I am sure would be able to take on your friends Bell-matic. In the USA, you could give Spencer Klein a try. All three names if fed into Google with Seiko attached, should throw up contact details. Good luck!
David Shutt said:
Hello sir, my name is David Shutt. I have a Bell-Matic 17J, 4006-6011 sn: 101228. I bought this watch in Vietnam, Oct 1971. I just discovered your site and it is very interesting. Alas, it was too late. I sent mine to the Seiko Service Center for repair. I’m getting up in years and couldn’t have fixed it anyway, but I will refer to you from now on.
Pingback: The Seiko Bell-Matic 4006-7010 27 Jewels – Ding-a-ling-a-ling!!! | musingsofawatchaddict
Amit Jones said:
Would you know if the day and date indicator disks for this movement, are significantly different from other seiko movements? For example, would it be possible to use the day/date disks from a 7s26c or 4r36 in a bellmatic?
Hi Amit, no, the day/date disks tend to be specific to particular base movements and certainly, the Bell-matic disks are unique to 4005 and 4006.
Amit Jones said:
Thanks Martin, that’s good to know. I’ve got a beaten up Bell-matic and was hoping to swap the dial and discs for parts from a 7S26.
I’d always thought that something like a day/date disc wasn’t worth the effort of a redesign between different calibers, and I do still wonder if by chance there are any other models which would have compatible discs.
Are the 4006 discs plastic as well?
There are thousands of parts that might not seem worth redesigning between calibres but to do so would be to compromise the whole. The 4006 disks are metal, not plastic.
just got a 2nd hand bellmatic in the mail,
everything wsa beautiful and functioning until i pushed in the alarm crown to quickset the date, the crown stayed stuck in.
other functions are still functioning, time, alarm works and can pull out to arm alarm, and day/weekday wheel will advance despite not able to quickset.
i read online that sometimes it could be a dirty case/button, however I checked and its not the case
Do you have any ideas? I hope it will it be a quick fix *fingers crossed*
If it is not a sticky pusher gasket, and general purpose jiggling does not resolve the issue, then the only other options I can think of are either that the date corrector spring has become dislodged from the date corrector or that the latter has somehow become stuck. The solution to either requires the dial to come off I’m afraid.
Great breakdown of the Bellmatic.. really helps me on my journey to becoming a budding watch repairer.
One question I have a Bellmatic whereby the date changes as it should but the day won’t advance unless I press the quickset?
The quickset should only operate the date not day so I’m not sure why pressing the button would advance the day. Unless that action somehow unsticks a partially changed day that has stuck. Whatever the cause, it sounds like the dial needs to come off for further investigation.
David Rodríguez said:
After reading this a couple of times and actually finding a Bell-matic to restore i am gonna try the service. Thanks for all the great documentation you’ve done here.
Wanted to ask you (regular question but haven’t found a real answer) I have the tech sheet and I al still confused with tbe oils. What are the moebius equivalents for the Seiko S-4 and the Synt-a-lube. I am confused because it seems that they use them sometimes regardless. I thought the S-4 could be replaced by D-5 or HP1300 and rhe other by 9010. But just wanted a confirmation since they seem to use both on the keyless work and gear train …. Etc.
Anyway thanks again! Love to check you website.
I substitute 9104 (HP1300) for S-4 and usually 9010 for Synt-a-lube. Occasionally I’ll use 9020 for the latter. For pallet stones, either 941/2 or 9145/2. Sometimes 9501 for setting lever parts. And either 8201 or 8200 for the mainspring.
David Rodríguez said:
Thanks a lot Martin.
Exactly what I was thinking to do. It is just a bit confusing rhat in some places they just say that you can use one or the other. Let you know how it goes! (Fortunately I have a spare donor just in case ) 🙂