The Seiko 6105 is arguably the most glamourous of the Seiko 150m mid-tier divers’ watches, the 6309 the most ubiquitous but for a bit of understated elegance, a bit of pukka ’60’s style, we have to turn to the first proper Seiko dive watch, the 6217-8000.  Produced from 1965 it came fitted with the robust 17 jewel 6217A automatic movement with date complication and a lovely domed acrylic crystal.  At about 38mm in diameter it is small by today’s standards but all the better for it in my book.  For a bit of flamboyance, we have to make do with the over-sized crown fitted to watches emerging towards the back end of 1965 and beyond.  The problem with this watch, particularly in recent years, is price.  Routinely these days they go for north of $1000 US in decent nick, and casing parts are thin on the ground and so you really have to seek out watches that have not been too abused.  Up until the middle of 2011, I’d tried all sorts of strategies to snag one cheaply, coming close once or twice, but often lacking the courage at the last minute, wanting to hold out that bit longer for something just that bit better – but of course without actually having to pay the going rate. By way of a distraction, I started picking up the odd spare part here – a couple of excellent reproduction bezel inserts, a replacement crystal, a set of original hands taken from a 6119 dress watch and a couple of new old stock crowns. I even bought a bezel from South America that looked like it could substitute for the original if I ended up snagging a bezel-less example. But then finally, I stumbled upon a UK Ebay auction for a SEIKO AUTOMATIC WRISTWATCH 17 JEWELS, Divers style watch in working order. Nothing else in the description, no model number, serial number. Nothing. But some half decent photos:

In the end I did I did have to fight off a bit of competition from another eagle-eyed fanatic who had also spotted it, but I won out in the end, securing it for a sum small enough to satisfy the Yorkshireman in me.  So, let’s see what we have. The case looked excellent. The bezel a bit slack and in need of a new retaining spring. The dolphin case back worn and having suffered some slips from case openers over the years but ok. The dial looked very good although clearly with some characteristic black gunge around the outer edges of the lume. The hands not so good, having been poorly relumed in the past with some ghastly green gunk. But remember, I have those hands from the 6119. The movement ran but had clearly been worked on by someone not quite yet having achieved the status of artisan. So, all in all, I was pretty pleased and it needed some attention – which suited me just fine. The first job was to get the dial and hands sorted – a relume to the dial in C3 to brighten it up and the hands to match. I removed the dial and hands and sent them over the pond to one of the acknowledged lume(inaries) together one of the excellent reproduction inserts sourced from the UK (in fact these are the only proper aftermarket inserts ever to have been produced for this watch and they are perfect).  Here’s the detached dial and hands ready for the journey across the Atlantic:
I thought that while I was waiting I would get on with a movement service and sort out the case with a new bezel spring and crystal.  As we see from the following, the rear (top) of the movement has its share of battle scars,
and the eagle-eyed among you will notice it is also sporting an incorrect winding weight – claiming 24 jewels when 17 better reflects the actual jewel count of the 6217A. My guess is that this rotor may be from a 6205 dress watch – more of that later. You can also see that the movement is screwed to a movement ring, without which the movement would fall straight through the aperture on the crystal side of the case. With the auto winding mechanism removed we get a better view of the actual movement:
Plenty of wear from the rotor but looking clean enough. Stripping it further down, things still look ok – at this stage just the pallet cock, fork and staff and barrel to go before turning the movement over to strip the calendar parts.
But at this point I was struggling to remove the barrel: it appeared to be binding and required some determined but gentle effort to release it. And we can now see why it had stuck:

The lower barrel arbor hole is badly worn, with the plating worn though and a 180 degree gouge raising some sharp edges which were snagging on the barrel arbor. I ran this one past Richard Askham (look him up) who is always happy to provide sound advice and he suggested that perhaps the gouge was deliberately inflicted by a watchmaker to close the hole a little when the side shake became excessive due to wear. This theory seemed plausible, especially considering the condition of the underside of the centre wheel bridge which had either been worn from barrel side shake or had had material filed or sanded away to create enough space to accommodate the side shake.
The best solution appeared to be to smooth away the sharp edges, lubricate the barrel arbor generously when rebuilding the movement and hope for the best. Some sanding and polishing and it looks like this, the nickel plating now worn through.
Turning the movement over, we get a look at the calendar mechanism with hour wheel centre, held in position using a film washer sitting between it and the rear of the dial. The date wheel on this movement is immaculate, and something of a relief because these are often discoloured and difficult to get hold of from watch materials houses.
Removing the calendar plate reveals some interesting differences compared to the later 6105A and B movements.
In particular the date jumper (the bit over to the right nestling into the groove between the 21 and 22 on the date wheel) comprises two parts with the spring a separate component – making reassembly a considerably fiddlier affair than in the 6105 and 6309 movements for which the spring is an integral part of the jumper. The rest of the disassembly proceeded without incident and with the parts cleaned, the reassembly begins. Back to the top of the movement, we can pause to note that I am using a centre wheel bridge from a 6206 in place of the original damaged component – hence the gold colour:
The barrel slotted home happily and rotates smoothly without snags. Here you can see the rear of the barrel arbor bushing on the calender side of the movement

With the 3rd and 4th wheels together with the barrel and train wheel bridge fitted, we can pop in the balance and see if she runs. And bingo, off she goes with a couple of turns wound into the barrel:
At this point I let the movement run overnight and continued the following evening, test fitting the dial

and hands

to check all’s well. Time to clean and reassemble the case. Here are the main components, cleaned and laid out:
The bezel spring to the right seemed rather flimsy which was most likely the reason for the slack bezel and so I substituted the spring that came fitted to that South American bezel I’d bought months previously.  On with the crystal, a lovely 1:1 reproduction sourced from the USA, complete with new tension ring. The crystal snapped beautifully into position with the aid of a crystal press
And we can now refit the movement, with seconds hand in place, auto-winding mechanism also now refitted
Then the bezel with new spring, again snapping on with a confident click, all slack banished and then in with the new insert, a perfect fit too requiring no sizing at all
And we have what appears to be a watch again. That evening though, when resetting the time, the crown became extremely stiff, the clutch wheel slipping horribly. Off again with the hands and dial, I re-seated the cannon pinion and hour wheel and in the process of dissembling the quick-set mechanism discovered the clutch wheel had a couple of teeth missing. By chance the following morning the postman delivered a 6205 dress watch and that evening I swapped over some of the quickset parts and replaced the clutch wheel. Everything back together, the setting mechanism now smooth as silk and I think we are there:

Post script:  Well, I couldn’t get on with the white lume on this one.  It was a poor choice, not in keeping with the original and so I asked a UK-based relumer to re-relume it in vintage yellow (green).  It turned out to be a massive pain for him to remove the superluminova (he likened it to concrete) but he won out in the end and this is now what I’ve ended up with:

I think, a great improvement.