What is the difference between a design flaw and a design vulnerability? Simple. The latter is referenced in some detail in a separate appendix to the technical manual whereas the former remains entirely unacknowledged. I’ve referred repeatedly to the problems that can affect the day-date corrector rocker in the 56 series movements and yet ironically, for both the 5606 Lord Matic and the 5626 King Seiko we’ve met so far, this has not been a problem. As we’ll see shortly, the second of our two automatic high-beat King Seikos similarly rides our wave of good fortune in having a fully functioning corrector rocker. I seem to have been flattering only to deceive on that front. So why am I wittering on about design flaws and vulnerabilities? Well, counter to what the external appearances may have lead us to expect, the movement fitted to the 5626-7111 – introduced in the preceding post – refused to display any signs of life and as we shall see its failure to run was the result of one of three technical advisories to make an appearance in the technical manual appendix. As seems frequently to be the case, I am getting ahead of myself so let’s roll back and review the patient in its state as received:
What a lovely thing it is too. Removing the movement allows us to confirm just how perfect the dial is.
Barely a mark on it, nor any signs of general degradation. The hands retain their full complement of go-faster stripe but the minute hand is slightly flattened in the region of the hole where a previous watchmaker has over-exerted him-(or her-)self in refitting the hand set.
To avoid too much in the way of repetition, I’ll gloss over the details of the strip down but will pause at the outset to take note of differences in appearance of this B-variant of the 5626 compared to the A-variant fitted to the earlier 5626-7000 featured in the previous post.
The most conspicuous differences are the un-plated brass pallet bridge and barrel lid, both of which make for a distinctive and attractive contrast to the rhodium plating elsewhere. You may also notice that the intermediate date wheel and date finger are both made of plastic where in the 5626A they were both metal. You might also notice that the day-date corrector wheel rocker has a plastic wheel rather than the metal in the earlier movement. Closer inspection reveals a complete absence of cracks and consequently a fully functioning part.
One curiosity I encountered on further dismantling the movement was that, in spite of the fact that it was not and had not been running, the mainspring was more or less fully wound. After a bit of investigation, the culprit responsible for the movement’s reluctance to convert mechanical potential energy into kinetic motion in the wheel train was identified.
One of the pallet stones was wonky and appeared set too far back in the fork. Interestingly, just such a scenario is anticipated in the appendix to the technical manual:
Unfortunately, my attempts to reset the stone resulted in its end splintering and so at this point I found myself in need of a replacement part. I have a small pile of spare Lord Matic 5606 movements but the 5606 runs at a lower beat rate and consequently the pallet stones and escape wheels are incompatible with the higher beat 5626.
You should be able to make out in the photo above that the 5626 pallet stones (left) are significantly narrower than those serving the lower beat 5606 (right), a consequence of the greater number of teeth on the escape wheel of the 5626 and consequent smaller spacings between each of those teeth. Needless to say, 5626 pallet forks are like hen’s teeth and I do not have any spare high-beat 56-series movements knocking about from which to harvest a spare. In the interests of completing the servicing of this movement, I undertake to borrow temporarily the pallet fork from the 5626-7000 until I’ve sourced a replacement part. I encounter no further hiccoughs until opening the barrel.
Nothing too much to cause concern other than to notice the striations in lubricant evident on the barrel floor. When I removed the mainspring, it became clear that the spring had gone out of flat close to the arbor and was rubbing against the barrel. Of course, it goes without saying that 5626 mainsprings are no longer available nor are there any suitable replacements from the Generale Ressorts catalogue. So I will probably just have to make do with the original and see how the watch performs once it’s back together.
With everything cleaned, reassembly follows its now familiar course. I’ll pause though to illustrate the steps required to refit that slightly out of whack mainspring into its barrel: 1) identify a suitable winder and drum and thread the mainspring onto the winder post, oriented such that it curve follows an anti-clockwise path when observed from the post side; 2) wind the spring into the drum by turning the handle clockwise and having threaded the bridle, release the power by winding back half a turn or so. Gently lever out the handle, whilst winding anticlockwise; 3) with the handle removed, give the coils a quick clean with Rodico and then; 4) press into the pre-lubricated barrel.
The final step is to lubricate and fit the arbor, the upper coils of the mainspring and then fit the barrel lid.
The rest of the reassembly proceeded smoothly and in no time the movement is up and running and displaying figures more or less identical to the 5626A covered earlier (225 degrees amplitude, zero beat error, noiseless timing curve). On with the first reverser idler, its bolt and screw and then the oscillating weight.
The calendar parts next, ready for the dial and hands.
The plastic dial spacer on this watch is of a different design to that on the top-loading 5626A and appears rather similar to those used on the 61 and 63 series movements. Unfortunately, this one was cracked and I had to source a spare from my Lord Matic stash.
The precise positioning of the dial spacer needs a little thought but eventually I figure it out.
And at last, we can refit the dial and hands.
I fully wound the movement and left it to run for a day or two. Initially all seemed fine, but on checking the movement the following day, I noticed that it had somehow lost about 10 minutes. The timing machine displayed a timing curve as flat and true as a doldrums horizon. What on earth? Could the hands have fowled at some point and then freed themselves? Perhaps the day or date jumpers are too strongly sprung and are slowing the movement as it approaches midnight? The hand clearances are fine and so I decided to test the second theory by setting the watch to 10.30pm and putting it on the timing machine again. Absolutely no effect at all – the movement continued to run happily at more or less 0 s/d, with undiminished amplitude. However, as I watched the seconds hand tick around, I noticed that the minute hand was not moving. The watch was clearly running but the minute and hour hands were stationary. At this point a little light bulb went on above my head and the identity of the likely chief suspect in this rather tortuous whodunit stepped forward: one off-centre cannon pinion/driving wheel. Turning to the technical manual appendix, we find the instruction ‘When only the second hand is in motion and the hour hand and minute hand stop during calendar shifting, check procedure 4’. Procedure 4 is illustrated below:
I confess to having noted a comparative lack of resistance when setting the hands which must clearly have been due to the loosened caulking of the cannon pinion but it did not occur to me at the time that this lack of grip might actually cause the driving wheels to come to a halt under the additional resistance of the in-progress calendar change.
Removing the faulty part requires the autowinder mechanism and barrel bridge to be removed, having first removed the balance, pallet cock and pallet. The sweep second wheel must come out too which of course requires that the seconds hand is removed. The third wheel can stay in place for removal of the driving wheel but in practice has to come out when fitting the replacement.
Fortunately, that pile of 5606’s I have can offer up one or two potential candidates to replace the faulty part. Here are two such:
I cleaned and sparingly lubricated both, selected one and reversed the process, rebuilding the movement back up to the point where I could get it running again and test the replacement part. Unfortunately, the replacement wheel had also lost its grip (I know how it feels at this point) and so, once more unto the breach dear friends, once more. Out it comes, and in goes candidate number two.
The reason the third wheel has had to come out before refitting the driving wheel is that the pinion at the bottom of the driving wheel has to mesh with the sprung minute wheel on the calendar side and this requires some maneuvering (with the dial and hands still in place). The arrow in the photo below shows where the pinion will sit against the teeth of the minute wheel.
A second partial strip down and rebuild complete, and we are ready to install the balance and give it a try again.
It was with some relief that this time I find the movement happy to power through the calendar change without the slightest flutter of discontent and so we are finally ready to declare this movement fit for purpose once more. I still need to source a replacement pallet fork but I have a couple of leads on that front and hope that before this little series of entries is complete, I’ll have sorted that problem.
An intermission now follows to allow me to catch my breath. We’ll pick the story up next time when I turn my attention to the question how to deal with V-type bonded crystals.
James Gleason said:
Hi, can you tell me the precise diameter of the dial?
Not off the top of my head but I may be able to make a measurement from one of my other 56 series projects. I’ll update this comment when I get a chance to do that.
Nicholas Mork said:
Just tried to service mine, however, same problem with the large driving wheel.. and cannot fine a deplasement online. Is the part 220560 a common problem with these movements?
Yes! The fact that Seiko choose to include an entry in their technical guide on the matter suggests that it is endemic. As you will have read in the blog entry, I was able to source a replacement by farming my small collection of spare 5606 movements. Good luck!
Nicholas Mork said:
Thanks for a quick reply Martin. I will start to look for 5606 movements. Did you manage to find a new pallet fork for your movement?
I did yes. Good luck with your search.
Do you think putting the cannon pinions in the ultrasonic cleaner is what caused the calking to loosen?
Is this issue limited to only the offset cannon pinion?
I’m reluctant having my 5625-7110 serviced due to this issue, or at least advising my local watchmaker about this issue and suggest he use a less intrusive method to clean the cannon pinions.
As I understand it, the cannon pinion is secured by friction and the loosening must just occur with time and wear. I don’t see how the ultrasonic could play a role other than to shift congealed lubricant. And the latter is not what is supposed to be supplying the friction.
I wondered since “procedure 4” specifically states “…due to loosened caulking…”
I do have a nice 5625-7121 here, but the rotor is heavily corroded. Do you know by chance where I could source one?
Have you tried any of the watch materials houses? If not then eBay or Yahoo Japan for old stock parts or used movements/parts.
Pingback: King Seiko 5626-7041 Officially Certified Chronometer from May 1974 | Adventures in Amateur Watch Fettling