Bell-matic Research Thread

[dapellegrini]

I suppose it was inevitable, but I don't know if I will have the energy to take these quite as far as I took the world timers.

As some of you probably saw, I have received a couple Bell-matics, with many more in the mail.  Of the first two, I already have a taste of the reckless mess of parts these can be.  At this point my ambition is to sort out a list (with photos) of the valid watches, by model number.  Based on image data, information already available online, and feedback here, I hope to be able to build a list of valid watch combinations - including case (ss or sgp) with matching dial, hands and bezel combinations + bracelets (when possible).   

First a big thank you to the following write-up - this really got things rolling:

www.beyondthedial.com/post/seiko-bell-matic-the-complete-guide/

To kick things off, I've spent the last few days starting a collection of sales images.  This list is likely to grow quickly as I find time.  So far I have 350-some-odd.  I dropped these in a folder and wrote a quite bit of code to organize the data - this is early days - not optimized (lots of large images etc) - but here it is:

seikoworldtime.com/scanner/bellmatics

I don't know what any of this means yet.  At this point, I think my approach is going to be to work through each model in isolation - and try and pull out the original combinations - based on what looks like most common - first into a Google Doc (here: docs.google.com/document/d/1yHcU-A--EWiivGl8HU3W3cT0lxOweii5bJxBLn8aZvc/edit?usp=sharing) - perhaps eventually a website or some articles.

Discoveries after spending 10 minutes with the data:

• I have already uncovered more sub-model numbers than documented in that (excellent) article - in BLUE in the google doc for now
• It would appear that models ending in 1 and 2 are successive, replacing the prior - for instance the 4006-7000 seems to have been replaced by the 4006-7001 and then the 4006-7002 - so I don't think those are regional variants. 
• Models ending in 4 or 6 appear to be SGP - so the successive thing seems to be for the 0's to 1's to 2's - stopping there.
• Models ending in 9 are the US Market versions

PM me if you would like to help - with edit rights on the Google Doc - which would just mean in your spare time you might take a submodel, and start updating the document with the supposed accurate combinations.

[dapellegrini]

Creating a model index in this post became too large a task... Instead there is a publicly available Google Doc here:

docs.google.com/document/d/1yHcU-A--EWiivGl8HU3W3cT0lxOweii5bJxBLn8aZvc/edit?usp=sharing

As of this writing it is over 130 pages, with a table of contents that will quickly get you to any of the 13 different major models in this series!

[inboost]

Looking forward to whatever this amounts to dapellegrini. Your methodical and analytical approach to such research really makes it enjoyable to learn along side of you on such things. The 4006-6070 was really my focus as I just loved the Arabic version dial. I've only seen the two models of the Arabic dial (4006-611L T). There's the white one you have and a navy blue one with white dial text. I'll email you a picture set I have of that to add to your photo collection.

[Rod]

The main issue with BellMatics is that you can swap dials, rotating rings between models so there are lots of franken Bells out there. The two base models are the 4006 (day, date) and the 4005 (date only), the 4006 comes in 17 and 27 Jewel varieties the 4005 27 jewel only.

There is a rarer 4006 that has only 21 Jewels that was made for the American market. Be very careful buying these as they are often rebuilt frankens.

In rarity the Nov 1966 dolphin case back 4006 is extremely rare, I know of only 5 and have been collecting serial numbers and info on these for 10 years now. Mine is the second oldest I have found. These have the Diashock 27 jewel on the dial and the dolphin case back and the notched case for the alarm button.

Then the 1966 Dec - Jan 1967 Diashock dial without the dolphin case back

The 21 jewel 4006-7019 would have to come next, 7 digit serial number (pre 1968) and very few seen, but they do come up for sale.

The 4006-7010 (generally from June-july 1967) Business Bell would be number four. Hard to find in good condition outside of Japan, strangely the all have an English only day wheel!

Lastly any of the 4005 models would come next.

These 4006/4005 movements are considered to be one of Seiko's greatest movements and are super reliable and keep excellent time.

[dapellegrini]

@rod - thanks for that context.  

I've just shy of 400 images cataloged - and just based on the caseback data - regardless of what case or dial they are married to, a few things are starting to become apparent:

1 - each model seems to follow the last so far - so the 6040 started production before the 6050, etc



2 - based on what is listed for sale, nothing clever with production numbers, there is not an obvious most common - it seems a pretty good mix - so far the 6070 and then 6000 and 6050 seem the hardest to come by




3 - with some degree of certainty, it looks like model numbers ending in 1 and 2 were successive updates from the 0's - for instance:













And that trend continues through the rest of them ...

[inboost]

Beautiful work dapellegrini - seriously - This could hang on a wall in a watch workshop!

[dapellegrini]

Rod - I've started a Google Doc here - if you have time to peruse the models you are familiar with, I would be grateful for any feedback.  If you have the time and interest, I am also looking for a couple people to promote to editors, to help round it out and make sure it is accurate.

docs.google.com/document/d/1yHcU-A--EWiivGl8HU3W3cT0lxOweii5bJxBLn8aZvc/edit?usp=sharing

I currently have 11 cases / major models, with a total of at least 62 different dials across the range.  I think I am close to having the entire range encircled - but lots of detail to confirm and sort through.

[dapellegrini]

And, I've just rolled in the 4005 (date-only) data into the Google doc, using 39 recent sales with adequate image information to start a dataset.  Of those 39, they all line up June 1968 through May 1969 - for what currently looks like a 12 month run.  

Here's an updated models run chart - the 4005 is the 4th one down.  Definitely the shortest run of the bunch.

[endy]

These data visualisations are good stuff.

[DanS]

Not sure if this is the right place to ask, but I have been wondering this for several years. I haven't been able to find anything definitive. Does anyone know if Bell-matics were sold in the PX during the Vietnam war? I ask because my late father-in-law wore one as a combat helicopter pilot, and I once saw a photo of it. Unfortunately, I haven't been able to track it down to inspect it in person and take good photos.

[dapellegrini]

DanS - interesting question.  I personally have no idea, but it seems plausible, as many other Seiko models were bought through the exchange back then (and still today).  I'm curious, do you know which (major) model he had?  

[dapellegrini]

Another point of curiosity I have been pondering... Why do some of the dial codes here increment from ending in 0 to ending in 1 - e.g. the 6060T dial became the 6061T dial at some point. Here is a summary of the data points I have collected so far:



At first I thought it made sense that the dial codes would have incremented to end in "1" at the same time the case backs incremented to end in "1", but I don't think that is accurate. Instead my current working theory is that the dial codes incremented to end in "1" at the same time that the case back design went from the horseshoe style to the standard/plain style.

There are 4 models that never got a horseshoe case back (later production) and those also do not appear to ever get a dial code incrementing from 0 to 1. These are the 6050, 6060, 6070 and 6080 models.

It appears that the 7010, 6010, 6020, 6030 and 6040 all have dials that do increment from ending in 0 to ending in 1, and pouring over the data, this does seem to align fairly strongly to the same time the case back design changes on each of these.

That leaves the 7000, 7020 and 6000 - all of which never received the switch to the plain case back... and also never saw a dial code increment from 0 to 1.

So... as my theory goes, if you have a model that was produced with both horseshoe and plain style case backs, then your dial code should end in "1" to match a plain style case back and "0" to match a horseshoe

[dapellegrini]

Another working theory... this one to do with the last digit in the model numbers - often thought to be a region designation. With a solid data set of case backs I think it is highly likely that

• Models ending in 0 were replaced with a 1, and sometimes a 2
• Models ending in 6 were replaced with a 4 - also these appear to be all BASE METAL, Gold tone watches
• Models ending in 9, were replaced with a 7, and in one case an 8
• Importantly, I can find no evidence that these models were EVER produced at the same time - instead 1's always come after 0's, 7's always after 9's

So here's my current thinking on this. Models ending in 0, 1 or 2 are our World/JDM models. Models ending in 9, 8 and 7 are USA/North American models. And Models ending in 6 and 4 are some kind of special fake gold/budget run.

I have not yet found a plausible reason why these last digits got incremented, though it may have been an across the board (for the 4006's) decision at different trigger dates. It is clear that one such date is Nov 1977 for the then running 6040, 6060, 6070 and 6080 - which all ticked over to end in 1's in that month.

Here is what I think happened... NOTE: I am not 100% each model switch at exactly the same year/month (yet), but the data does line up to suggest it:

October 1969

• 4006-7000 becomes the 4006-7001
• 4006-7010 becomes the 4006-7011
• 4006-7020 becomes the 4006-7021
• 4006-6000 becomes the 4006-6001
• Also ...
• 4006-7029 becomes the 4006-7028

May 1971

• 4006-7001 becomes the 4006-7002
• 4006-7011 becomes the 4006-7012
• 4006-6001 becomes the 4006-6002
• Also ...
• 4006-6020 becomes the 4006-6021
• 4006-6030 becomes the 4006-6031
• Also ...
• 4006-7006 becomes the 4006-7004
• 4006-6016 becomes the 4006-6014
• And finally ...
• 4006-6029 becomes the 4006-6027
• 4006-6039 becomes the 4006-6037

November 1977

• 4006-6040 becomes the 4006-6041
• 4006-6060 becomes the 4006-6061
• 4006-6070 becomes the 4006-6071
• 4006-6080 becomes the 4006-6081

[inboost]

Excellent data reduction efforts here dapellegrini ! I wish I could add a data point but my bell is such a mutt we will never have any confidence in what it was.

4006-6070 MArch 1977 non-horseshoe caseback with 4006 - 611L T dial

[dapellegrini]

The 4006-6070 has its own peculiarities...  The dial for the standard edition has a dial code ending in 6090T and is the only dial in the entire range that does not specify the jewel count - theory being this one was sold in Japan and the lower 17 jewels being called out what have hurt sales. 

It appears that the 4006-6080 was then released, getting the next dial code increment, ending in 6100T - then back to the 4006-6070 where we get the updated version with the arabic dial later in the production run - and this one's dial increments to 611LT - I guess they ran out of track at this point...

[dapellegrini]

Ok, next on my list - sort out when the case backs went from PROOF to RESIST to RESISTANT...

My notes:

1) As I am sure is consistent with other models, only the US versions ever saw a RESIST case back variant. For the 4006's that means the the 7029, 7028 and the 6029. These are the only models in production during this window.

2) There are two models with 21 jewels - both very rare, the 7019 and the 7029. It appears that the 7019's all had PROOF case backs - but the 7029 survived into the transition with some later examples getting the RESIST case back - and these 7029 RESISTS might be some of the rarest of all...

3) I believe these are the only models that see a change in this notation:

• 4006-7001, PROOF to RESISTANT - Between July and October 1970
• 4006-7011, PROOF to RESISTANT - Either in December 1970 or January 1971

• 4006-7029, PROOF to RESIST - at latest by April 1969
• 4006-7028, RESIST to RESISTANT - November 1970
• 4006-6001, PROOF to RESISTANT - at latest February 1971
• 4006-6010, PROOF to RESISTANT - Between August and October 1970
• 4006-6020, PROOF to RESISTANT - at latest February 1971
• 4006-6029, RESIST to RESISTANT - either February or March 1971
• 4006-6030, PROOF to RESISTANT - Between October and December 1970

The above timing uses the last known and first known case back dates for the transition. In bold is the earliest date for the transition, based on photo data

As you can see timing is a little mixed per model, but PROOF to RESISTANT (and RESIST to RESISTANT) seemed to happen in Q4 of 1970, whereas PROOF to RESIST changed as early as April 1969.

Of note, the change from PROOF to RESIST or RESISTANT does not line up with case code increments or dial code increments (eg 4006-7000 changing to 4006-7001 or dial code 6020T going to 6021T).

Also, I was a little surprised to see a RESIST acase back as early as April of 1969. I wonder if there are any (from any other models) earlier than that? Here it is:

[dapellegrini]

And on to the next curiosity. The models ending in 6 and 4...

These are found in the 4006-700X series and the 4006-601X series, but nowhere else. What's more, it appears ALL of the models ending in 6 or 4 are gold-toned - but there are other gold-toned watches for both of these series that do not end in 6 or 4... WTF?

While pouring over photos tonight I realized that the 6's and 4's are all marked BASE METAL, whereas the normal versions in gold tone are marked SGP. I only own one SGP version, a 4006-6010 with original bracelet, and I was surprised when I received it to see that bracelet marked GOLD FILLED 18KT.

So now I am thinking perhaps these 6 and 4 versions were budget fake gold cases... whereas the main line models in gold-tone were actually gold plated/filled

I rolled through the rest of the 60XX models, most have SGP versions, none have BASE METAL, gold-tone versions...

[inboost]

I often wondered about the various types of gold colored cases and bracelets. Many of the bracelets are still silver on the inside while gold plated on the outside surface (perhaps better sweat resistance?). The actual 18k and 'gold filled' watches always give me pause, as it seems sellers really like to push values up for these. Honestly it seems to me the actual gold content isn't measurable or recoverable from the watch and the value of the watch really shouldn't rise due to the actual precious metal content.

[dapellegrini]

I agree - and often point out to sellers that most gold colored Seikos are NOT gold watches...  especially the digitals from the 80's and 90's.

And for those following along with the data collection - I was able to find 14 Dolphins (Nov-1966) and 2 Horseshoe Diashocks (Dec-1966) - added to the catalog here:

seikoworldtime.com/scanner/bellmatics#4006-700X

That brings me to 567 total bells to-date... so about half way there on the data front!  Enough to start making some educated guess though - as in the previous few posts...

[yankeeflyguy]

Sept 6, 2022 5:17:21 GMT -8 DanS said:

Not sure if this is the right place to ask, but I have been wondering this for several years. I haven't been able to find anything definitive. Does anyone know if Bell-matics were sold in the PX during the Vietnam war? I ask because my late father-in-law wore one as a combat helicopter pilot, and I once saw a photo of it. Unfortunately, I haven't been able to track it down to inspect it in person and take good photos.

Hello everyone! First post here. I did a search on a watch I recently inherited and it brought me here. Then I saw the above post and figured I’d mention my watch. I inherited this from my uncle who purchased it in new in Vietnam while he was serving as a aviation supply clerk (helicopters). From what I’ve found out on Facebook it’s a 1968 Seiko Bell-magic 27J JDM. If I can figure out how to post photos I’ll do it below.