- Kelly's World- A View into the mind of Uber Geek, Kelly Adams - https://www.kgadams.net -

Fixer clock #1 tear down and reassembly

I picked what I thought would be the least complicated clock to work on first. This circa 1913 Gilbert “gilt No. 115” clock has a simple time-only mechanism. Unfortunately for me, it is a small and “cheap” (mass produced) clock, meaning the thick brass and large pivots found in some of the other “fancier” clocks are replaced with pot metal and tiny parts that aren’t really made for easy repair.

What follows is sort of a journal of my experiences thus far in working on this clock. For anyone who doesn’t have at least a passing interest in clocks, it is probably advisable to skip reading the rest of this post. The short story: I successfully disassembled, cleaned, repaired the main problem, and re-assembled the clock. It still doesn’t work properly, and I’ve found at least one additional problem that I will have to fix later.

The folks on the NAWCC message boards helped me identify the clock [1], which gave me a bit of an idea what I was looking at. A mass produced novelty clock made by Gilbert Clocks [2] out of cast spelter [3] , which is sort of like a poor man’s bronze. Originally this was coated with gold or “gilt”, but any evidence of the gold is long gone. The internal mechanism is a balance wheel with what I was told is a club tooth escapement: again, the NAWCC folks were instrumental [4] in figuring this out.

It was also at this point that a small niggling doubt began to grow regarding making this my first clock to repair. One of the posters in the NAWCC forum opined that this type of clock is not necessarily a good first clock to repair. I decided to carry on and see how far I could get, although I’ve been looking for a “deal” on a weight driven clock to add to my collection ever since.

In examining the clock, I found my first problem. The mainspring sits in a “half barrel” permanently affixed (riveted) to the back plate. This means normal ways of removing and re-installing the spring by removing the barrel and using a spring winder are impossible. Further, the pin or “joggle” in this barrel (basically, a cut out bent into a hook) that the spring originally hooked on had broken off. This meant the spring couldn’t be tensioned, explaining why the clock wasn’t running

The broken joggle further complicated spring removal. Tensioning the spring and using a C-clip or similar to hold it for removal would be the normal alternative to a mainspring winder… in my case, I was going to have to remove the spring by hand- since it is still under some tension, this can be a bit dangerous. But its a small clock, with a small spring, so I bought some gloves and eye protection and proceeded.

I stripped the clock down, removing the external minute gear and faceplate, then the gears this step released, and finally the mainspring. Everything went without any difficulties and no injuries were sustained- yay! During this process, however, I really began to appreciate how small the pivots on the gears are. I don’t have a micrometer or caliper yet, but I’d estimate they are on the order of 0.07 to 0.1 mm. To put that in perspective, in the larger clocks I probably should be working on, pivots (the little axles that pass through the front and back plate and upon which the gears spin) are 0.5 to 1.5 mm in diameter: nearly ten times the size of what I’m working with in this clock.

One unanticipated issue: the balance wheel is attached to a little “hairspring”: absolutely miniscule in the case of this clock. The balance wheel basically fell out when I removed the faceplate of the clock just like the the other gears… except it was suspended by its attached hairspring, which was attached to the backplate. Given that springs of this size are nearly irreplaceable, and are generally considered very fragile, this was a bit panic-inducing. I managed to figure out in record time how to detach the tiny hairspring, and it seemed largely undamaged.

During this panic I missed taking pictures of the disassembly process: not, as it turned out, a critical problem as I had enough pictures of the still-assembled mechanism to figure things out later. But I can’t emphasize enough: you need either a near-flawless memory, an excellent understanding of how different clock mechanisms go together and a really good memory, or plenty of pictures of your mechanism. It may look simple, and there may only be seven or eight gears, but they can go together in a vast array of puzzling and ultimately wrong configurations.

Now I had a clock in pieces, all of which I carefully put in little box. I then pondered how best to clean the pieces. The main purpose with cleaning is to get all traces of the original lubricant removed, particularly on the bearing holes and pivots. This old lubricant is usually a cross between glue, molasses, and sandpaper: coagulated oil mixed with extremely fine bits of metallic grit. Modern clock repair generally involves use of an ultrasonic cleaner and special chemicals. Professionals without an ultrasonic cleaner or who prefer the manual approach will usually use similar chemicals and soaking/rinsing processes, plus lot of careful spot cleaning with brushes. Guys like me with neither ultrasonic cleaners or fancy chemicals will use a simple bath of water and dish soap, and a toothbrush.

I know this sounds like an ineffective method, but even some pros go this route. And I figured I’d give it a try. The process took me about an hour, and I doubt I did a truly adequate job. But the gross cleaning isn’t where you stop. Regardless which method and tools you use, the next step involves fine cleaning of the bearing/bushing and pivot surfaces. I used round toothpicks to clear out each bearing (i.e.: hole in the face and back plate). I then used a special polishing/burnishing file (which is one of the most expensive tools I’ve bought to date) to work on the pivots. This took another couple of hours of work. Note that the pros will usually have a powered lathe for this process, or at least a special “hand lathe”- I had nothing, and so I’m sure my polishing of the pivots was very poorly done.

During this stage I asked for and received more advice from the NAWCC forum denizens [5] (includes pictures of the mechanism). After removing the mainspring, they helped me confirm that it was “tired” (worn out) and needed replacing, so I ordered a replacement part. That took a while to arrive, and while I waited I ordered another clock and started the process of ordering an ultrasonic cleaner as well as one of those manual hand lathes I mentioned above. As I noted, I didn’t have either of these tools while working on this first clock, but I established that I want them. I also ordered various lubricants, tweezers, a jeweler’s saw and ultra-fine blades for cutting a new joggle in the mainspring barrel… it is never ending.

I began reassembly once the mainspring and jeweler’s saw arrived. The new mainspring arrived tightly tensioned and held with a wire, which gave me a method to reinstall it. After cutting the new joggle, I placed the mainspring in the half-barrel, donned my gloves and eye-protection, and carefully removed the wire. The new spring… sprung into place perfectly, with only one little problem- it did so with enough force to flatten the carefully crafted joggle I had made. With some patient manipulation of the spring and some fine pliers, I was able to get the spring hooked on the joggle successfully.

The rest of the reassembly was surprisingly easy. Note that I don’t have the hands or minute gear installed, but those sit outside the faceplate and my first objective was to check the running of the mechanism. Unfortunately, it didn’t run initially: however, with some careful adjustments to the balance wheel to get it “in beat”, I was able to get it running for nearly five hours. With the mechanism cleaned, I can see some excess “slop” in one of the bearing holes: that will require re-bushing to correct, yet another thing I don’t have tools for. I also think there may be some irregularities in the escapement wheel, but I need a lot more magnification than I currently have to see that: each tooth is about a tenth of a mm.

So, I turned a non-running clock into… a half running clock. Not a terribly obvious improvement, I’ll admit, but I’ve gained a great deal of knowledge. I’ll be setting Fixer #1 aside for the time being until I have bushing tools and better magnification, and will start on my second clock over the weekend.