It is not that common these days that I find myself confounded by a fault affecting a watch movement. Well, actually that’s not quite true. I should qualify that. It is not that common that I find myself confounded by a fault affecting a mechanical watch movement. Quartz movements on the other hand can sometimes present mysteries to which only the electrons in their circuits are privy. Quartz movements have defeated me twice before (with the second defeat drawn out to encompass the content of this present entry). I suspect that if I had the will, I could probably apply myself to resolving such conundrums but cannot conjure up sufficient motivation to do so because somehow my interest in the workings of the wrist watch align squarely with a fascination at how all those tiny parts conspire to bring about coordinated action with such precision, powered only by the potential energy bound up in the tightly wound coils of a spring.

The first of my two defeats wasn’t a defeat really; just a failure to muster the will to diagnose and attempt the revival of a completely dead ETA quartz movement when the cost of the parts required to achieve a fix would likely exceed the cost of a complete replacement movement.   The second was much worse because the fault was sufficiently intermittent that I thought mistakenly that a service would resolve the issue. It did too until, as fate would have it, I sold the watch and the glitch resurfaced. That watch was the Hellas ‘Heuer’ clone whose ‘revival’ is documented here.

The symptoms reported by its new owner were that the movement exhibited a reluctance to restart when the crown was pushed back into position following time or date adjustment and, if it did restart, it might then stop some time later. The glitch could be rectified by tapping or partly unscrewing the crown, following which the watch would run reliably. The fault proved not to be sufficiently vexing for its new owner to persuade him to return it and so he has lived with this operational quirk since. However, recently we have been discussing this watch again and decided to try to resolve the problem once and for all, but the solution I have proposed circumvents the actual fault by substituting the original ETA 536.121 no jewel quartz movement with a 25 jewel automatic ETA 2836-2.

Such a proposal might seem uneconomic given the modest cost of the original watch but thrifty instincts cultivated by 20 years living in North Yorkshire saw the acquisition of a non-functioning ’80’s TAG-inspired Swiss Army abomination for a song, inside of which lurked an ETA 2836-2 hopefully itching to be put to some good use.

Not pretty but I didn’t buy it for its face but for its heart.

The success of such a transplant hinges on a number of requirements: 1) that the caliber diameter and stem height of the two movements are identical. 2) that the dial feet positions are compatible 3) that all three hand hole sizes are identical between the two movements, 4) that the crown positions on the donor and receptor watches are the same and 5) that there is sufficient case back clearance to accommodate the rotor.

As it happens, the 536.121 shares much of its DNA with the 2836-2. Both are 11.5 ligne (25.6mm) diameter calibres; both have a stem height of 2.25mm (the 2824-2 has a stem height of 1.8mm and would therefore be incompatible with the case); both have the dial feet holes located in the same positions; and importantly, both accept hands with hole sizes of 1.5×0.9×0.25 mm (hour/minute/seconds). The crown positions of both cases are at 3 and so the date disks should be interchangeable. The final requirement of rotor clearance will have to wait until I have the watch in hand.

So, the order of play is to see if the 2836 can be revived without too much additional expenditure and if so, then the Hellas wings its way back from Portugal for the transplant. We start then calendar side, having extracted the 2836 from the Swiss Army carcass.

Initial impressions are wholly consistent with the bargain bucket pricing of the donor. Both calendar disks are tarnished with more deep-seated corrosion in the centre of the day disk. Sometimes, exterior tarnishing is just that and need not necessarily betray more serious issues beneath. In this case, removing the calendar disks yields some mixed messages. The calendar parts are grubby, yes, but otherwise in decent condition but the Novodiac shock spring has gone AWOL, allowing the balance cap jewel to wonder and the shock setting to work free.

Stripping away some of the fittings and the mood lifts slightly. It all looks neat and tidy and largely familiar territory based on the experience of a couple of years back with the original 536.121 quartz movement. A few jewels in evidence this time though.

Over to the upper side of the movement, and impressions are again dominated by a combination of the light patina of neglect offset by some properly grotty bits.

The balance wheel looks particularly afflicted but the good news is that the hair spring looks uniform and unmolested.

The inside of the case back indicates maybe one service in its lifetime. I can’t quite make out the date but the fact that three of the four wheels in the automatic device framework remained firmly glued into place when detached from the main body of the movement suggests that it has been quite some time.

Continuing the deconstruction, let’s pause to look at the neat click and click spring arrangement beneath the crown wheel

before removing the train wheel bridge and admiring the arrangement of the train wheels and barrel. The second wheel proved somewhat troublesome, having become bound into its hole by a combination of congealed oil and rust.

The discolouration around the barrel arbor in the photo above is a common enough sight in Seiko movements that use molybdenum barrel greases but in a Swiss movement, suggests something unpleasant within.

The arbor is badly corroded, so much so that it is unsalvageable and I’ll need to source a replacement. The barrel itself doesn’t look too bad and should clean up satisfactorily. The final step before the cleaning stage is to refit the balance and then remove the shock jewel.

Talking of shock jewels, you may have noticed that this movement is fitted with Novodiac settings rather than the more familiar Incabloc. The former is rather similar to the Diashock settings used in Seiko movements but the springs are considerably fiddlier to refit without the correct tool (as I discover later).

The considerable quantity of dirt on this movement necessitated both ultrasonic pre-cleaning followed by extended agitation in my watch cleaning machine. I then also took care to manually clean all of the pivots with pegwood.  With the parts looking considerably smarter than their previous state, I could set about reassembling the movement starting with the mainspring and barrel.

I’d gone through the motions of cleaning the arbor with the other parts but post-clean it is clear that it needs to go.

Following the order suggested by the ETA technical guide, we continue with the keyless works and setting lever.

Now, what about that exploded Novodiac setting? Without the actual spring to hand, I had to exercise a bit of guess work in sourcing a replacement because the springs themselves come in four different sizes but in the end the choice was reasonably clear and I opted for a 231 spring at 1.7mm diameter. The basic design of anti-shock settings is pretty standard and so with the cap jewel pre-oiled and captured by the setting itself, it is placed into the bearing with the new spring following.

Unlike the equivalent Seiko Diashock settings, the spring cannot be finessed into its setting prong by prong. You need instead to push all three corners simultaneously downwards and then rotate the spring into position. Without the correct tool, I had to improvise by using a suitably sized stake from my staking set. Easy as pie.

We are ready now to fit the train wheels, having replaced the corroded second wheel with a new part.

In goes the escape, intermediate, third and second wheels in that order.

Next, the train wheel bridge and barrel, having carefully teased each of the arbors into their settings.

Notice too the hacking lever in position waiting for the barrel bridge, crown wheel and click and the ratchet wheel.

The pallet fork and its bridge have also been fitted and lubricated in the photo above. The final step on this side the movement (for the moment) is to fit the balance and its Novodiac setting, lubricate, wind in some power and see how she runs.

Rather well, as it happens. A quick adjustment to beat error and timing and the movement is running true and displaying hugely impressive amplitude (up to 310 degrees at full wind). I don’t for a minute imagine that its performance overall could not be improved upon but given that I’ve only had to replace three parts, I am very satisfied.

The technical guide suggests refitting the autowinding framework next and so that is what I do, even though it runs counter to the way I normally work, generally anticipating that the appearance of any flies in the ointment later on will require its removal again.

The four wheels are located first into the framework and secured in place by a jeweled bridge.

The guide recommends scant lubrication of the reversing wheels through the spokes (the two brownish coloured wheels in the photo above) with Moebius 9010.

With the framework refitted, it is time to tackle the calendar side. We begin by refitting the date unlocking yoke, its spring and the retaining plate.

This turns out to be an unexpectedly fiddly business, particularly locating the retaining plate. So too, is the process of fitting the date driving wheel, the main obstacle being the insistence of the now sprung date unlocking yoke to make this as fraught a process as possible. Eventually, I get it to sit still and can then set about fitting most of the other wheels on this side of the movement.

It should be plain sailing from here but inevitably this has all gone too smoothly. In fitting the minute train bridge that keeps the cannon pinion, minute and hour wheels in place, the screw shears off. I turn the air blue and contemplate the meaning of life for a moment or two.

Resignedly, I have to move into reverse gear, dismantling some of the calendar side again before removing the autowinding framework (doh!), the balance, crown wheel and barrel bridge in order to gain access to the upper side of the main plate, hoping to be able to manipulate the broken screw thread from beneath. Fortunately, my staking set came to the rescue once more and I was able to gain purchase on the thread and screw it out of its hole without any great difficulty.

Disaster averted. Refitting all the parts once more and we approach the finish line because by this point the Hellas has arrived from Portugal. The substitution requires me to gain entry

and perform the extraction.

Having removed the dial and hands from the 536.121 and we can pause to compare the outgoing movement (left, below) with the incoming (right) and note the extent to which the design of the calendar side is essentially identical.

Before fitting the dial support ring, dial and hands, I decide to remove the now redundant day corrector (indicated earlier). The anticipated compatibility of the dial and hands proves to be well-founded.

All that remains is to refit the movement to the case, and size the freshly-fitted 2836 stem.

I am hoping that this will banish any further behaviour problems with this particular watch but on that of course, time will tell. We finish with a couple of closing shots, neither of which betray the change of heart beneath.

Save