French lantern clock restoration

Being in the clock repair business means you are always restoring someone else’s clock. I have repaired and restored thousands of clocks in the last twenty years and done just a few for myself. I thought it was about time I did one that I particularly fancied. I have owned, repaired and sold a number of lantern clocks over the years but I wanted to undertake a major lantern clock restoration project of my own. I was motivated by the example of my mate Lindsay who owns several lantern clocks, all restored by him. His support and advice was much appreciated in the undertaking of this project.

I bought a French lantern clock (on eBay, where else!) and the photos below show it as received and after restoration.

The clock was engraved Mercier a Paris on the dial and he was my starting point to try and pin down the age of the clock. However Mercier proved to be a difficult bloke to nail down. My research uncovered a heap of clockmakers with that name from the late 1600s until the end of the 1700s but no clues as to which one my Mercier was. The name was clearly not going to help much in dating the clock. The lack of a first name or initial didn’t help either. Bearing in mind that Paris makers would have been at the forefront of their profession, it seems unlikely that they would have been signing their names to, or making, single handed clocks after about 1750, so it is probable that this clock predates 1750. In the absence of any further clues I guess that it is enough that it is some 200 plus years old.

Upon inspection of the clock it became clear that there was a great deal of work to do to get this little project up and running again. I really liked the extras that this clock originally had. It would have had a striking train, alarm and a half hour passing strike, which by the way was still there. Most of the strike train was missing and all of the alarm work except the alarm setting dial on the front. Also missing were the pallet arbor, pendulum, weights, backplate and doors. On the plus side it appeared to have the original dial and engraved centre, original single hand, original frets, the maker’s name, Mercier a Paris and, importantly, no one had previously tried to restore it.

The photos below show the condition of the clock as I received it. It was covered in dust and dirt and had clearly not run for a very long time. Although originally rope drive it had been converted to chain at some point in its life. The crucifix post wedges were all missing as were the strike warning and lifting levers.

Where to start with all this? I didn’t want to make an exhibition showpiece of the clock and restore it back to virtual new condition. I believe in old things looking gracefully old so I wasn’t about to polish and shine the clock so you’d need sunglasses to look at it. I wanted the clock to look old but OK and for it all to function properly. I also wanted to reinstall the alarm, but more of that later.

The Going Train

I pulled the clock all apart and washed off dirt and gunge so I could inspect the parts more closely. My aim was to get the clock working as a time only clock before tackling the strike train and then the alarm. The time train wheels were undamaged except for both the pivots on the contrate wheel which were badly worn and needed repivoting. I use carbide drills for this but always with my heart in my mouth in case the drill breaks for then you’ll never get the broken piece out. The other pivots were trued and polished. Bushes were inserted where needed. The chain wheel needed to be reconverted to rope and fortunately I had a couple of longcase rope wheels from which I could take the centres. I had to make a new rope wheel click as well.

Here is the contrate wheel after repivoting.

Here is the time chainwheel showing the modern clickwork and chain drive centre. This I removed and replaced with a click and spiked rope centre appropriate for the period.

Here is the same wheel showing the new click of the correct type to work against the wheel spokes.

Here is the going train re-installed in the clock.

With the time train wheels installed and running freely it was time to tackle the making of a new pallet arbor. The backcock was missing, so I fabricated one, silver soldering the components together. The tricky bit was to get the pallet pivot hole height matching the front cock and in line across the centre of the verge.

The pallet arbor was made from a steel chime rod, nice and straight, good steel and relatively easy to machine the pivots on either end to fit the new backcock. The pallets were soldered into slots, adjusted by filing to mesh properly with the crown wheel and then polished. The pendulum crutch completed the pallet arbor assembly and has a bend to allow it to line up with the backcock suspension and to provide clearance for the alarm assembly.

Now I needed a pendulum, so again, after looking at examples on the Net, I made this one. I think the pendulum is an important visual part of this clock so I spent a bit of time turning a design I liked on the lathe and also shaping the pendulum rod hook at the top. The bob has a wood insert which I tapped 6BA to suit the pendulum rod. This allows a friction fit and means you don’t need a locking nut and is true to the period. I made a guess at the length, expecting I’d probably have to make another rod when I got around to installing the motion work and thinking about timekeeping.

I made new cruciform plate top wedges so the plates could be fixed in place after the train wheels were in. I tested the running of the train using a spare 2 kilogram weight, getting it into beat and making sure it was reliable. No motion work on it at this stage.

Here is the wedge holding the cruciform plate in place. Note it is tapered and driven down to stop the plate from moving. Simple but clever and effective.

The pallet arbor and backcock in place.

The time train in place and wheels able to turn freely.

The Motion Work

With a single handed clock the motion work is dead simple. The hand arbor is driven by a pinion on the end of the time train wheel arbor. This pinion was missing so had to be made. The pendulum length for timekeeping is determined in part by the number of teeth on this pinion. I made and tried a seven tooth pinion which meant the pendulum bob was above the clock feet ie: the clock could sit on a flat bracket. However this hid the pendulum from view which I really didn’t fancy. I then made an eight tooth pinion which brought the pendulum down to just below the feet. This has meant making cutouts in the back of the purpose made bracket for the clock but means the pendulum is in view as the clock runs. Much better and is probably how it originally was.

Below is the eight leaf pinion pinned to the time main wheel arbor. The curved piece of steel is the bottom half of the return spring for the half hour strike hammer.

Here is the back of the hand arbor showing the wheel with which the pinion meshes and also showing the twelve points of the star wheel for setting off the strike train and the half hour passing strike.

The Striking Train

The only wheel left in the strike train was the chain wheel which I dismantled and riveted in the spiked rope centre. A new click spring was made up and also riveted in place.

The three missing wheels, hoop wheel, warning wheel and fly came from a thirty hour longcase movement which examination had showed could be a good fit. There was not a lot of wear in the pinions or the pivots so that made the rest of the job a bit easier. The arbors were all too long and had to be shortened on the lathe. Of course the pivot holes had to be repositioned on the plates. A depthing tool was used to determine the centres and after new pivot holes were drilled and the pivots cleaned and polished, all the wheels meshed and fitted perfectly with a little backlash between the teeth and end play at the pivot ends. The old pivot holes were riveted with brass rod and filed and polished back to not show. Finger pressure was enough to make the train run smoothly and easily.

Warning and Lifting Levers

These two levers were missing so I looked at the old longcase movement to see if the levers on it could be utilised. They were of course too short and meant for operation from the other side of the movement anyway but they were square bar like the originals would have been and were about the right dimensions. I got them out, cut off all protrusions and welded on two cms to make the each bar longer, then filed both ends to match the existing square bar. After turning pivots on the ends to fit between the front and back cruciform plates I made up and silver soldered on lifting and warning levers to allow the strike train to function. I copied exactly the original levers from a French lantern clock from the same period. I used all old iron as it blended in with the other parts much better than new iron. This component looks like a trident in shape and the position, bend and length of each of the three prongs had to be determined by trial and error. Note the warning lever has a return bend on its end because of the rotational direction of the warning wheel. Getting the strike train and levers installed into the plates correctly was a test of patience as everything had to fit exactly right and of course with cruciform plates everything wobbled and fell out just when I thought I had it done. However I eventually got it put together and it worked fine with a 2 kilogram weight.

The strike train needed a six tooth pinion made to drive the count wheel so I made this one by first gashing the teeth on the lathe and then hand filing to fit, just like they did some two hundred years ago. This pinion fits on the rear of the chain wheel arbor and meshes with the wheel teeth on the inside of the count wheel. You can see the fixed arbor for the count wheel above the pinion.

The Alarm Train

On this lantern clock the alarm train would have been between the rear cruciform plate and an iron backplate which was pinned to the top and bottom clock plates. Existing holes in the plates for the arbor pivot and for the rope to pass through show this had to be the case.

My first step was to make a backplate which I cut and shaped from a quite ancient piece of flat iron about 1.3mm thick. Here it is pinned in place at the top. It has lugs on the bottom to fit slots in the lower plate.

I lined up the existing alarm arbor pivot hole on the cruciform plate with a spot on the new backplate to fix the other arbor pivot hole. Then I used a 45mm hole saw to cut out the opening for the alarm wheel and riveted in an iron bridge with arbor pivot hole in place. The bridge has a dish in it to allow for the alarm pallet arbor to fit past. I made up a new alarm rope wheel with spikes and click. My mate Lindsay kindly made a 40mm alarm crown wheel, put in a stop pin and crossed it out, so all I had to do was fit it to my new arbor and rope wheel and install it.

The alarm hammer and pallets are on the same arbor so I made this one to fit, copying from examples I had seen and also riveted a new bottom cock on to the backplate. The arbor is inserted from the top, through a slotted hole in the top plate which allows the pallets to pass. Some fiddling around with height, hammer size and pallet angles was necessary to allow the hammer to rattle back and forth hitting the bell properly. The alarm setoff and stop lever had to be made as well, being tripped from the back of the hand arbor. Again some adjusting had to be made so the alarm tripped at the time set by the alarm setting wheel on the dial plate.

The last parts of the alarm to be made were the tapered steady pins poking up through the top plate which hold the alarm hammer heads away from the bell at rest and prevent secondary annoying vibration when the strike operates. These were hand ground to a taper on a diamond lap. These pins have to flex as the alarm operates and are riveted in from below. On the top right of the backplate you can see the pivot hole for the alarm setoff and stop lever. There is a corresponding hole for the other end on the front of the clock at the top left of the dial.

The Clock Case

The frets all appeared to be right for the clock and the period. The front fret was in good order and this naïve pattern is seen on quite a few other French lantern clocks. The excellent book Franse Lantaarn Klokken by Ton Bollen shows several examples. One side fret had been broken so I made a piece up and silver soldered it in.

Both side doors had to be made as well. These have a cutout in the top near the hinge to allow the door to be inserted without dismantling the case. This is how they were originally made.

I silvered and lacquered the dial but not so brightly that it would look new. You can see the other end of the alarm letoff and stop lever pivot at the number XI on the dial.

Lindsay and I made new lead weights and counterweights for the three trains and bound on rings to the ropes. The main weights were 2 kilograms.

The lantern clock needed a wall bracket. This was not a hook and spike clock as there were no holes at the rear of the top plate where a hook would have been attached. No spikes either but then on French lantern clocks the spikes were attached to the backplate, and on English clocks to the back feet.

I planned a design I liked from examples I had seen and made the bracket from oak. I wanted the open front design to be able to see the clock and pendulum better and it’s also easier to set the clock up. Note the rear cutouts to allow the pendulum to swing freely. The two arms on which the clock feet rest have countersunk holes for the feet so the whole setup is very stable.

That completed the restoration of the lantern clock. The point about a project like this is that both the workmanship and the research have to be done properly to get a good result. Lindsay and I spent a lot of time collecting information, analysing its worth, discarding much and planning each step so that the clock could be restored to its original appearance and function. The end result has been well worth it.