Those of you who read the recent account of the restoration of an early Citizen Cosmotron X8 electric watch may remember that I completed that project with at least one loose end requiring attention. The principal fly in the ointment concerned a reluctance of the movement to run with a standard 344 cell: the only way I could persuade it to run correctly was to attach a blob of solder to the negative terminal of the battery and add an insulating disk to the battery compartment to prevent contact between the negative terminal of the cell and the main plate. I knew at the time that this was a stop-gap fix that had failed to address an as yet undiscovered underlying problem but it allowed me to get the watch running while I regrouped.
It is now clear to me that I had been drawn to an incorrect deduction that the problem related somehow to the architecture of modern button cell batteries compared to those produced in the 1960s. Of course, this neglected to explain why at least some examples of these watches are still in use but the relative sparsity of information on vintage Citizen wristwatches provided a fertile patch in which my folly could take root.
In order to rework this conundrum, we need to take another look at the marriage of mechanical and electrical that forms the basis of the stop seconds function: the completion of the electrical circuit when the crown is pushed in is achieved by the lever action of an elbow at the opposite end of the yoke against an insulated trapezoid block attached to the end of a lever stop.
As a result of this action, the metal arm of the lever stop makes contact with pin protruding from the base of the negative terminal that sits beneath the battery and the circuit is completed.
The reluctance of the electrons to flow was due to the underside of the terminal shown in the photo above not making contact with the lever stop when the battery was fitted. The battery terminal needs to depress the tab in the centre downwards, whereby it comes into contact with the metal arm of the lever stop. The only way I could get it to complete the circuit was to add some depth to the negative side of the battery by soldering a blob of silver solder which resulted in the tab being forced downwards sufficiently to make contact with the arm beneath.
Clearly, this was not any sort of proper fix, not least because I suspected that the heat from the soldering process may have compromised the integrity of the battery but as a short-term solution, it worked.
Since publishing that original post, I have been rooting around for potential donors to give up parts that I may be able to use to address whatever issue it was that was affecting my watch. After a couple of false starts involving incompatible evolutions of the 08 series, I secured a Cosmotron X8 pocket watch whose movement looked like it would serve up what I needed.
The path to redemption starts with extraction of the troublesome movement and removal of the dial and hands. The first port of call in tracking the misalignment was the lever stop.
That kink is not supposed to be there. The arm is supposed to be straight. The inside of the kink marks the point at which the terminal pin on the battery connector spring makes contact, but the contact point will now be further away than it should be if the arm is adjusted in such a way as to ensure that the insulating trapezoid is able to make contact with the yoke extension. In a belt and braces approach I ended up substituting the yoke and setting lever spring…
… followed by the lever stop and battery connector spring.
For the watch to function correctly, the dial needs to be refitted because it provides a back stop for the lever stop (the rear of the dial is insulated with an acetate disk). With that done, we can turn the movement over and check the alignment of the lever stop with the yoke.
That looks much better. The battery connector spring from the donor movement fits more snugly and its connecting terminals are in pristine condition.
The final step is to select a battery not requiring home-made insulating disks. Fortunately, the 387S (a 394 with a plastic spacer), designed for use with Bulova Accutron, does the trick.
It only takes the slightest tickle of contact between the securing strap and the contact point at the base of the balance for the movement to spring into life.
The timegrapher reports a very clean timing curve with an amplitude that suggests that the movement is receiving as good a current from the smaller battery as it was with the larger 344.
In having had to remove the hands as part of this process, I took the opportunity to remove the tarnishing that had remained on the minute hand. The two hands are now evenly matched and I think I can now declare this one well and truly finished and free from loose ends.
Adventures in Amateur Watch Fettling 
I did think that the temporary fix was not like you Martin so I’m glad you have revisited it!
Yes, that project turned into a bit of a marathon even before encountering the problem with the battery. By that point, I needed a break from it but also wanted to get the post out so I could move onto something else. I always knew I would return to it and in the intervening period between the original post and this update, I bought three donor watches from which I hoped to be able to farm parts. Only one of those turned out to be compatible!
I have just bought a Cosmotron, from the next generation. It was bought as a non-runner (just needs a battery!). I opened it up to find a Renata 344 and the screws at either end of the battery contact with their heads snapped in half. Not being able to change the battery would be a reasonable enough reason to me for it to slip into non running status.
The screws were removed easily enough and when I added the new battery it instantly sprung to life, before reinstalling the battery contact. Pushing down on the battery made the balance pick up pace and, having got one replacement side screw into the battery contact, it took right off. I was feeling pretty pleased with myself.
Now to get the next screw in. I pushed down on the contact, in order to push the end into place and wrestle a screw in. At this point the movement stopped and wouldn’t start again. After reading your article I am guessing something might be shorting out though the lower battery contact looks a lot different to yours (it’s a star shaped spring in the centre of the battery well). The movement is actually a bit loose in the case, so I wonder if something might be shorting on the dial side as a part of the movement touches the case. I am still not sure about the correct battery. This was considered, through review of other recent articles, to be a Renata 386 which is actually the ‘high current’ version which I figured wouldn’t harm the watch. Now I am not so sure because even if the overall dimensions are the same the size of the positive ‘nub’ may be smaller, or bigger in a detrimental way.
That does sound a little like the problem I was having with the negative terminal of the battery shorting to the body of the mainplate which is why I fashioned that insulating disk. The 387S deals with that issue so perhaps you might give that a try.