Rare Bulle-Clock with Garde-Temps movement
This clock is a very unique Bulle clock, possibly one of a kind. The clock has no serial number but I’m fairly certain that it was manufactured in 1927. This date has been determined with guidance from the various Bulle catalogues and price lists posted by Peter Smith; guidance from John Hubby from the NAWCC; and guidance and example of a similar clock from David Read. The exact date has been pin pointed by family accounts of my grandfather’s trip to Paris in the late 1920’s, when he bought the clock, and final determination by the ship’s manifest posted in the Ellis Island site which shows that my grandfather arrived in New York from Paris aboard the SS France on September 28, 1927.Although the exact clock is not referenced in any of the available Bulle catalogues or price lists, it seems to be a Garde-Temps movement fitted into a Model W case.
In the 1925 catalogue, the front page features the Garde-Temps Bulle clock. Fig 1. The clock face is exactly the same as my clock, although the hands shown are spade hands rather than the moon hands on my clock. While the photograph does not show the pendulum, the description states that it has a “new additional temperature compensating mechanism”.
David Read owned a Garde-Temps clock, Fig 2, and shared detailed photos showing the movement and pendulum, which match my clock exactly, including the unique “temperature compensating pendulum”. This confirms that the movement on my clock is in fact a Garde-Temps, although it is not housed in a wood case a shown in the 1925 catalogue. David’s Garde-Temps, which is housed in a wood case similar to the one shown in the 1925 catalogue, does have moon hands like my clock.
In the 1929 price list, Bulle lists all the models available: 22 models with some 89 variants. Model W is listed as a wall clock, with a bronze frame and 5 glass case and a 29.5 cm pendulum. The model W is the only model featuring a 5 glass, precisely like my clock. The Garde-Temps has a 29.5 cm pendulum so it makes sense that it would fit into the W case. Model Z is the Garde-Temps and interestingly it does not have a price, suggesting perhaps that it was either not ready for production or maybe an expensive model which was made to order. Perhaps as David Read suggested the accuracy fell short of expectations and the product was scrapped.
An interesting fact is that I remembered my father telling me that this clock ran on a wet cell which was placed on a small pedestal below the clock. I had some doubts if this was true or a figment of my imagination. However, the Bulle-Clock Practical Manual, translated from French by Robert Miles, states on page 32 that “…certain models have wet cells. These have a glass container with a hole fitted with a cork bung”. It makes sense that this clock indeed ran on a wet cell especially considering that the clock was destined to the Dominican Republic where in 1927 it may have been difficult to obtain dry cells. David Read corroborated this assumption by telling me that in the 1920’s dry cells were far from reliable and shared some photographs of wet cells like the one my clock probably had.
In conclusion, we know that the Garde-Temps was being advertised in 1925 and listed in the 1929 price list, albeit with no price, and my clock has been dated as manufactured in 1927. Evidently, only a few Garde-Temps ever made it to market. David Read tells me that he knows of only 4 existing Garde-Temps in their original regulator cases. He also knew of a Garde-Temps movement that was recovered but with no dial or hands and it was restored and housed in a decorative wood Bulle case. It seems likely that when my grandfather visited Paris in 1927 they opted to fit the prototype Garde-Temps movement into the Model W case. This would explain why the movement has no serial number and why there well may be no other Garde-Temps movement in a Model W, 5 glass case.
I have faint memories of seeing a photograph of the clock in my grandfather’s company conference room and I suspect that when the company moved headquarters in the late 1950’s the clock was damaged and stored in a cabinet in the new conference room. Later, in the late 1960’s, I remember my father attempting to fix the clock but the hour hand and gear were missing. I helped my father fix it and made a crude hour hand and adapted a gear with epoxy; very crude, but the clock ran, although it did not keep time well. The clock was hung on a wall mostly for decoration and eventually, was once again relegated to a cabinet and forgotten.
In February 2017 while on a fishing trip I met Joe Lyons who owns an antique shop in Vermont and specializes in restoring antique clocks. I told him about my clock and although he knew very little about Bulle clocks, he sparked my interest and guided me to some web sites where I started learning about the clock and identifying resources that would help me in the restoration process.
The clock was in generally good condition, with only a few broken pieces and missing the hour hand and gear. After having researched the clock and realizing how unique it is, I decided to keep all the original parts as long as they were serviceable even if they were not cosmetically perfect.
Clock and case were completely disassembled for servicing. Since I had no experience in disassembling a clock, I took detailed photographs and numbered each part and placed them in a grid so that I could trace back the assembly.
Case – Metal frame was fairly corroded and was cleaned with bronze cleaner, initially by hand and eventually with a slow speed Dremel polishing wheel. The glass was dirty and moldy. The two side panels each had broken corners. One of the corners was available and glued back in place. The other broken corner was missing. The glass was polished with cerium oxide which cleaned it pretty well although some deeper scratches were not removed. While installing the glass panels, one of the side panels broke under stress. Again it was glued in place. I am looking to find someone who can cut the two side panels which are beveled.
Back plate – was generally corroded and scratched. It was sanded and painted with mat black paint.
Bulle label – bronze letters were polished by hand to protect black background.
Wall stabilizers – one of the screws was broken and the other was bent. Since the broken screw is shorter than the other, I am using the longer screw to anchor the clock horizontally by drilling a hole in the wooden mounting surface.
Movement – completely disassembled and cleaned with ultrasonic cleaner and tumbler with corn husk media.
Contact spring and isochronous spring – Although in serviceable condition, were replaced.
Suspension – , which had been made from a silk ribbon by me years ago, was replaced with an original one. The groove on the fork arbor where the contact spring rides was broken during the ultrasonic cleaning. A small groove was made on the remaining shaft stub and although not perfect, keeps the contact spring in place and is working well.
Electrical connection to pendulum, which had a crude wire, was replaced by a balance spring per photos from the Garde-Temps.
Dial was cleaned but left pretty much as is, showing its age, for fear of damaging it more. The bronze rim was hand polished.
Second hand needle was very loose on its arbor. I made a jig to ensure perpendicularity and glued it back in place. Second hand painted mat black.
Hour gear was procured from a donor Model A Bulle clock but the sleeve was a little bigger than the arbor. I made a bushing by wrapping bronze foil on a drill bit of the same diameter as the arbor (it took about 6 turns); gluing the foil with Crazy glue as I wrapped it around the bit; inserting the bit with the wrapped foil into the gear sleeve and then removing the bit, leaving the copper foil inside the sleeve as a bushing. The fit is perfect and given the slow movement of the hour hand, I expect the foil will not deteriorate and last a long time.
I have searched many sites on the Internet and have not found an original moon hand to fit this dial. I hope to find one someday. In the meantime, I did find a set of moon hands from a large clock and have adapted the minute hand into an hour hand so that the proportions to my original minute hand are adequate.
Magnet – After the clock was assembled I had a hard time making it work. Although I adjusted the pin and fork and the isochronous spring as best I could, it still wouldn’t work so I started suspecting the magnet. I tried Peter Smith’s test for the magnet and the indication was that it was in fact weak. I also found an app for the iPhone that mimics a gauss meter (measures magnetic strength). Both of these tests suggested a weak magnet so after much deliberation I decided to re-magnetize the magnet. I found guidance on how to do this on the web. In short, I wound the magnet with copper wire, and applied a car battery current. In order to protect the battery and avoid fusing the contacts and causing a fire or explosion, I used a 15 amp fuse in line, which would blow immediately, but would give the magnet a brief jolt. Apparently it worked as the iPhone gauss meter showed an improvement in the magnetic force.
Coil – After the re-magnetization, the clock would still not run so I began suspecting the coil. I measured the resistance and it was about 1200 ohms, well above the recommended minimum of 1100 ohms. I decided to do a test by placing a compass on top of the coil from the donor clock (smaller clock and coil). When I applied current, the compass needle would deflect about 45 degrees. I then did the same test on my coil and the deflection was only about 10 degrees. This convinced me that the problem was the coil. I decided to rewind the coil (which I had never done) and ordered a small winding machine. However, I was very concerned that I would take apart my coil and not be able to rewind it, or that it still wouldn’t work. Therefore I decided to make a new coil and do my magnet test.
With the new coil finally made, I first measured the resistance and it was about 1280 ohms. (About 4,200 turns of 41 AWG magnet wire). I was sure it would work, but when I tried my compass test, the deflection was the same as the original coil. I was frustrated, but at least I made a coil!
So I decided to put the original coil back and concentrated on calibration. Sure enough, I soon learned how critical the depth of the pin into the fork is as well as the importance of proper tension on the isochronous spring. I also learned how critical it is to ensure that there are no unwarranted friction points in the movement. For example, the worm gear has an adjustment which initially was not set correctly and the worm gear exerted too much pressure on the minute hand gear. This prevented the clock from running. I also discovered that I could adjust the pressure of the back track spring and this was also an added friction point.
After several attempts, I finally got the clock to run without stopping!
Cosmetically, the black enameled casing on the coil had a few scratches and I debated whether or not to paint it, finally deciding to leave it as it is for authenticity.
Battery – During the initial runs, when I couldn’t get the clock to run properly, I decided to try higher voltages. I found that by using three batteries in series (4.5 Volt) I could get the clock to run somewhat. However, David Read had advised that a higher voltage would simply be masking the real problems and that it would be almost impossible to get the clock to keep accurate time. I got a regulated battery and after many hours adjusting the pin and fork and isochronous spring and trying different voltages, I eventually was able to run the clock with 1.5 Volt. The regulated battery sold by Carlton Clocks is too large and given the glass case it is quite unsightly. I decided to make a regulated battery with a slim black case. It is fitted with 6 lithium AA batteries and regulated to 1.5 Volt. I suspect that battery pack will last a very long time.
Calibration – After several days of calibrating the clock with the rating nut, I have been able to keep time with reasonable accuracy. In the last four days the clock gained 20 seconds. I am hopeful that as I continue to calibrate, I will be able to keep time to within a minute a week or less.
Acknowledgement. A very special tanks to David Read who without his guidance I would have never been able to finish this project.
Santo Domingo, D.R.
May 29, 2017