In this post I’ll describe how I repaired this magnificent English James McCabe 8-day naval chronometer.
When I received the clock, although it was very dirty, it was quite well preserved. However, despite being in good condition, the chronometer had no spring detent and the escapement wheel and some teeth were worn. It also had no second hand or crystal.
First of all, I cleaned off the teeth of the escapement wheel, removing some bumps (remains of dry grease). The shank of the escapement wheel was soldered to the pinion with tin, so it was essential to start by correcting it. I centred and flattened the wheel without unsoldering it, as this was not necessary.
After this, it was the turn of the spring detent, which is the heart of a clock. Throughout my long career repairing chronometers, when I have had to make a spring detent, I have normally had a sample to measure and a model on which to base it. However, in this case, I had no sample, so I had no option but to take measurements directly on the clock.
To make a spring detent, I have always followed the steps outlined in the book “Watchmaking” by George Daniels (p. 165), which I think is the easiest and most effective way to make one.
The first thing I did was to measure from the centre of the screw hole to the centre of the jewel hole, which coincides with the centre of the spring detent stop screw. Then I measured the distance between the screw and the locking screw. With these three distances, we have a good basis. I measured the distance between the centre of the balance wheel jewel and the centre of the spring detent jewel, and subtracted the radius of the balance wheel plus 0.2 mm. Using these measurements, I calculated that the approximate length of the spring detent was 25 mm from tip to tip. After making adjustments, the final length was 24.2 mm.
One of the most common complications with a spring detent is definitely the position of the spring’s locking jewel. In this case, it was not at the geometrical centre of the part but more towards the inside, towards the escapement wheel, in a straight line with the cord. It was vital to solve this, even more than making the steel and gold leaves for the spring.
After finding the correct resting position for the spring, I finished it with a very fine file, sandpaper and jewels, as well as a lot of patience and delicate handling, until it was attached to the clock with its screw. Lastly, I checked that everything was in the correct position, without putting the sheet spring on as yet, because this is the last thing to do on the part.
After checking that everything was in place, I made a gold plate 0.03 mm thick, found a screw with a 0.5 mm thread to hold it on, took a synthetic ruby jewel 0.55 mm in diameter, and prepared the shellac for gluing it in. Everything was ready for the final process, making the sheet spring, tempering and polishing.
I tempered the spring detent in oil and finished it with diamond sandpaper, to give it a matte finish. I mounted the jewel and the gold plate, and checked the strength of the plates and the position and the flexibility of the finished part on the clock. I had to make some small alterations before everything was properly adjusted, to get a perfect, flawless, accurate escapement.
I disassembled the mechanism completely, cleaned it and buffed, corrected and blued the screws, polished the pivots and made a compensated second hand with a shank.
The final result was a balance wheel with an optimum amplitude, to keep the precision to between 290º and 300º. An 8-day marine chronometer in itself is a very pretty, spectacular and attractive mechanism, especially when it is of high quality and working well, as in this case. After setting the clock going, it was time to look at accuracy. On the fusee of this clock is a Kullberg system, which uses the law of the lever to have a more uniform, constant tension, as the mainspring barrel and the fusee go in different directions, forming the shape of an inverted S as the chain loops around the fusee, away from the barrel. Normally, in most clocks with a fusee, the chain runs linearly, but not in high quality, precision clocks.
I calculate that I can get a very high degree of accuracy, around +/- 1 or 2 seconds a week. To do this, I need to use a chronocomparator and a digital chronograph watch, a notebook, pencil and time. I will just have to wait to see the results.
I hope that you have enjoyed seeing how this spectacular chronometer works, as much or even more than I enjoyed repairing it, and how much I enjoy it every time I see it.
