In this post I calculate the run time (1 day vs. 8 day) of a cuckoo movement I recently bought, and the Links Per Foot of the chain it uses. These two numbers tell me what parts to buy to make a clock case for this movement.
The economics of cuckoo clock repair are a bit odd: A new cuckoo clock movement costs about $70 USD; meanwhile, having a clock repairer fully clean a movement like this will cost at least $200. So, many clock repairers will swap out a dirty cuckoo movement for a new one rather than go to the expense – which they would have to pass on to the customer – of fully cleaning the dirty movement. Because of this swap, dirty but otherwise perfectly good cuckoo clock movements pile up in the clock repairer’s shop.
Often these dirty clock movements show up on eBay, listed as “cuckoo movement for parts or repair“. I imagine it’s a way for the clock repairer to recover a little of the cost of repair, and to clear out these old, dirty clock movements from their shop.
For the clock repair hobbyist, these dirty cuckoo movements provide a way to practice cuckoo clock repair on the cheap.
A short while ago I bought a rusty Regula 25 cuckoo clock movement on eBay for $26.95, with free shipping.
After the movement arrived, I disassembled it, gave the parts a good, gentle scrub with an SOS pad, a brass brush, and water, then tossed them into the ultrasonic cleaner for 20 minutes at 25 degrees C., rinsed them in water, then wiped them with paper towels soaked in isopropyl alcohol to drive the water out of the parts.
While I had the movement apart, it was easy to calculate the ideal pendulum period by counting gear teeth.
In this case, I also wanted to know how much the time train great wheel – the wheel that will hold the time chain – revolved per hour, because that will tell me whether it’s a 1-day movement, or an 8-day movement.
Lucky for me, the great wheel on this clock directly drives the minute wheel, and the minute wheel has the same number of teeth (36) as the great wheel’s gear that drives the minute wheel. So the ratio of movement of the great wheel to the minute wheel is 36:36, which is 1:1. So the great wheel rotates once per rotation of the minute wheel: that is, once per hour.
So the great wheel on this clock makes one rotation per hour.
Next I needed to calculate the Links Per Foot of the ideal chain that will wrap over the great wheel and hold the time train weight.
The cuckoo movement’s great wheel contains a sprocket that hooks the links of the chain.
As you can see in the photos above, the cogs (teeth) of the great wheel’s sprocket hold every other link in the chain. I counted the cogs in the great wheel’s sprocket – 7 cogs – so one rotation of the great wheel will move 7 * 2 (14) links of the chain.
Now I knew the links-per-hour of the chain: 14 links per one rotation of the great wheel, which makes one turn per hour.
Next I needed the length of those 14 links, which is just the circumference of the circle formed by the chain on the great wheel.
It’s a bit of a guess to measure the diameter of the path of the chain. That diameter will be somewhere between the outside and inside diameters of the sprocket. Here I sort of eye-balled where I thought the chain would rest, and measured that diameter as about 21 mm.
Circumference = Pi * Diameter, so the circumference of this sprocket chain is about 3.14159 * 21, or about 66 mm. Converting to feet – because chain is sold in Links Per Foot (LPF) – that’s roughly 0.216 feet.
I knew the circumference of the great wheel sprocket holds 14 links of chain, and now I knew roughly what that circumference was, so I could calculate the rough Links Per Foot of the ideal chain for this clock: 14 / 0.216, or roughly 65 Links Per Foot.
Looking at the chains offered by one clock parts supplier, Timesavers.com, I found they sell chains with 61 and 72 Links Per Foot, so 61 Links Per Foot is likely the right number. Working backward from 61 Links Per Foot, 14 links (the circumference of the great wheel sprocket) are about 0.229 feet, or about 70 mm. Dividing by Pi gives the actual diameter of the sprocket chain: 22.28 mm; only a little larger than my measured diameter of 21.13 mm.
…also, Timesavers describes their 61 Links Per Foot, 70 inch chain as “Used on Regula #25, 1-day cuckoo clocks”. Because this movement is labeled a Regula #25, we’re in business!
Finally, I wanted to double-check that this clock movement is for a 1-day clock. Cuckoo clocks are usually mounted about 6 feet from the floor. 6 feet * 61 Links Per Foot = 366 links from the cuckoo clock to the floor.
Remember that the great wheel of this clock moves 14 links per hour. Dividing the length of the chain by the links moved per hour gives 366 / 14, or about 26 hours of chain. So the clock is a 1-day clock, and not an 8-day clock.
Armed with these numbers, I bought the 61 Links Per Foot, 70 inch long chain, which turned out to be exactly right. Math works!
]]>What follows is a more detailed “how to” for calculating the pendulum period based on gear ratios.
Cleaning a clock properly involves disassembling the clock and cleaning each part. As part of the disassembly process, take lots of photos of how the movement originally fit together and the relative position of all the parts in the strike train – so you can put the clock together again.
The photo above shows the Ansonia clock’s movement just after I removed the front (top) plate that holds the parts together.
The time train is the set of wheels (gears) connected to the escape wheel. You can find the escape wheel – the upper-right wheel in the above photo – because it has distinctive, pointed teeth.
Note which wheels connect the escape wheel to the mainspring; those wheels make up the Time Train. In the photo above, I’ve labeled those wheels: The great wheel is the mainspring gear; the 2nd, 3rd, and 4th wheels are numbered based on how they relate to the great wheel. The minute wheel or minute post is the shaft that turns the minute hand.
I didn’t label the Intermediate Wheel – the gear just below the minute wheel in the above photo – because it turns the hour hand, and isn’t involved in the calculation of the pendulum period.
Now that you know which wheels are in the time train, note how the minute hand is – indirectly – connected to the escape wheel. How the minute hand is connected varies depending on the clock design. In some clocks the minute wheel is directly in the path from the mainspring to the escape wheel. In contrast, in this movement the minute wheel is not. That is, if you removed the minute hand, you could still turn the escape wheel by turning the great wheel.
So, in this movement the path from the escape wheel to the minute wheel is: escape wheel, 4th wheel, 3rd wheel, 2nd wheel, minute wheel.
Once you’ve cleaned all the wheels, make a table of the number of teeth (outer teeth) and pinions (inner teeth) for each wheel. Accurately counting those teeth can be tricky, especially for large wheels – it’s easy to miss one or two teeth. I like to take a photo of each wheel, looking straight down onto the gear. These photos simplify counting the teeth, and are also handy for finding bent, worn, or broken teeth.
The photo above shows the escape wheel – note the long, pointed teeth . You can easily count the teeth in this photo – 34 – and the number of pinions (the holes in the hub) – 7.
I took pictures of all the involved wheels, counted the teeth and pinions in each wheel, and made the following list:
Because the minute wheel is driven by the outer teeth of the 2nd wheel instead of its pinions, I wrote “60” – the number of outer teeth – for the 2nd wheel number of pinions.
Now that we have all the numbers, we can easily calculate the number of pendulum periods (a tick + a tock) required to turn the minute hand one full turn (one hour).
Each pendulum period allows the escape wheel to turn forward one tooth. When the pendulum swings one way, the “tick” releases one tooth of the escape wheel and catches a different tooth on one side of the verge (the pallet, which is connected to the pendulum crutch). When the pendulum then swings the other way, the “tock” catches the tooth next to the first tooth of the escape wheel on the other side of the verge, and the pendulum is back where it started. See Wikipedia’s Anchor Escapement page for details.
Because the escape wheel has 34 teeth, it takes 34 pendulum periods to advance the escape wheel one full turn.
So how many pendulum periods does it take to turn the 4th wheel one full turn? Because the escape wheel has 7 pinions, and the 4th wheel must move 42 teeth to make one turn, the answer is 34 * (42 / 7) =, or 204 pendulum periods.
Skimming over some of the math, an easy way to calculate the number of pendulum periods required to turn the minute wheel one full turn is to multiply the number of teeth of each wheel involved, then divide that by the multiplied number of pinions of each wheel involved:
pendulum periods per full minute wheel turn = 34 * 42 * 42 * 60 * 24 / (7 * 7 * 8 * 60), which = 3,672.
The minute hand takes one hour to turn, so the minute wheel takes the same time to make one full turn. One hour is 60 minutes of 60 seconds each, so we can calculate pendulum periods per second by dividing the periods per minute wheel turn by the number of seconds per hour: 3,672 / (60 * 60) = 1.02
We want seconds per pendulum period rather than pendulum periods per second, so to find that we just take 1 / 1.02 = 0.98039215686274509803921568627451. For clock-setting purposes, we only care about this time rounded to the microsecond (6 decimal places), so the ideal pendulum period for this clock movement is 0.980392 seconds.
We’ll use this number in two ways during clock repair and restoration:
About measuring the time the pendulum takes to swing, clock repairers often express the pendulum timing in BPH (Beats Per Hour), where a Beat is either a tick or a tock. We can calculate the ideal BPH for this clock by taking the calculated pendulum periods per second, multiplying by two, then multiplying by 60 * 60. We multiply by 2 because there are two beats (one tick plus one tock) per pendulum period; we multiply by 60 * 60 because an hour is 60 minutes of 60 seconds each. So the ideal BPH for this clock is 1.02 * 2 * 60 * 60 = 7,344 BPH.
By the way, another number that’s handy for regulating the clock is the ideal number of pendulum periods per week. You’ll use this number to calculate the minutes of time gained or lost each week based on the current, measured pendulum period.
The number of seconds per week is 60 seconds per minute times 60 minutes per hour times 24 hours per day times 7 days per week: 60 * 60 * 24*7 = 604800 seconds per week.
Divide that number by the seconds per ideal pendulum period to get the number of ideal pendulum periods per week: 604800 / 0.980392 = 616896.0 ideal pendulum periods per week.
To see how this number helps in regulating the clock, let’s assume we’ve measured the current clock pendulum period as 0.980497 seconds. Sounds pretty close to ideal, no? Let’s subtract it from the ideal period from it to see how much time (seconds) is gained or lost per pendulum period: 0.980392 – 0.980497 = -0.000105. That says this clock is currently losing about 100 microseconds per pendulum period. Still sounds pretty good.
Now we take that difference per ideal pendulum period and multiply it by the number of those pendulum periods per week: -0.000105 * 616896.0 = -64.77408, which is a little over a minute per week. Actually, that’s pretty good.
Notice that because this clock’s ideal pendulum period is about 1 second, a difference per period of 100 microseconds, or 0.0001 seconds means gaining or losing about 1 minute per week.
So we have a handy rule of thumb: For an about 1 second pendulum period, each 0.0001 second difference from the ideal pendulum period will produce about 1 minute of error per week.
So for example, if you measure the pendulum’s current period at 0.980,792 seconds, you can easily see that the clock will lose about 4 minutes per week (0.980392 – 0.980792 = -0.0004 seconds; -0.0004 / 0.0001= -4).
Practically speaking, if your 20th-century consumer mechanical clock gains or loses only 1 minute per week, it’s doing pretty well. So you should call it good if you can adjust your clock’s pendulum’s period to within 0.0001 seconds of the ideal you calculated.
]]>It seems the cut arch was still attached with the original, 100-year-old glue . I found several posts that claimed you can often separate such glued wood using a heat gun and putty knives, which seems incredible. I was skeptical until I watched this furniture restorer’s video.
I assembled my tools: a heat gun from Sparkfun, a couple of thin putty knives, and a steak knife (I didn’t use the steak knife).
The various posts I’d read provided a few tips:
I went to work, and was soon surprised at how quickly the glue began to soften.
Everything went well as I loosened first one side, then the other. I was just coming to the center of the arch when the whole thing suddenly popped loose!
I should have moved more slowly, or used a thinner putty knife. Checking out the two previously-glued-together pieces, I saw that a thin layer of wood had torn free from about 1/3 of the arch. Not enough to notice, but it could have been a lot worse. A thinner blade would have put less pressure on that wood, perhaps removing the piece without damage.
With a little sanding to remove the remaining glue and paint, the original Ansonia clock case is now ready for new gingerbread – all I have to do is design and build it!
]]>
In my previous post I had set the pendulum length, regulated (adjusted the speed of) the clock, and set it running for a 30-day test, to see whether its springs still run the clock for its full 31 days. If they don’t I’ll need to replace the springs.
I had to stop the test because I noticed the strike train great wheel gear had come loose from its arbor, risking the click letting the spring loose, which could damage the wheels – or worse, the hand of the person winding the clock when the click let loose.
So now I’ve disassembled the clock until I find out how to fix the problem.
Take a look at the above image.
When you wind this wheel by turning a key you place on the end of the arbor, the arbor and ratchet turn – indicated by the smaller arrow – while the larger gear remains in place. That gear is held in place either by the escapement, in the case of the time train, or the lock, in the case of the striking or chime train. As you turn the winding key, the ratchet, which is fastened tightly to the arbor (center post) turns, clockwise in this image. As the ratchet turns clockwise, the click raises to let each tooth of the ratchet pass.
When you stop winding, the click falls back between the teeth on the ratchet, and the arbor and ratchet try to unwind, counter-clockwise in this image. The click keeps the arbor and ratchet from unwinding, and transmits all the force of the wound mainspring through the arbor and ratchet to the larger gear. That gear in turn transmits that force to the next gear in the train, which…. and so on.
So you can see that while winding, the arbor and ratchet turn and the larger gear doesn’t; while unwinding the arbor, ratchet, and larger gear all move together, as if welded into one piece.
I never thought about that winding and unwinding, so I never realized that the larger gear is able to turn on the arbor. That gear is held down onto the ratchet by a “Tension Washer” – the grey disk in the back of the wheel in the image above. If the tension washer wasn’t there, the larger gear could tilt relative to the arbor, creating a gap between that gear and the ratchet. At worst, if that gap becomes large enough the click will fall into it instead of stopping against the ratchet, and the arbor and ratchet will spin out of control until the mainspring winds completely down.
After my Korean clock had been running for 11 days with no problems, I looked down into the works and saw a terrible gap in the strike great wheel, between that wheel’s ratchet and gear. It was about 1/2 the height of the click. If, as the mainspring unwound, the gear and mainspring tilted more, the click might let loose.
So I had to stop the 30-day-run test, disassemble the clock, and see what the problem is.
After comparing the strike train tension washer to the time train tension washer – which works fine, with no gap – I saw the problem: the strike train washer is bent and flattened so that it isn’t held in place by the ring on the arbor.
So I’ve ordered a clock-maker’s anvil, a set of stakes (rods with holes in their centers), and a riveting hammer. I already have a small ball peen hammer I use for 3D printing. Using these tools, I should be able to hammer the strike train tension washer back into shape. Then I should be able to reassemble the clock and start the 30-day run test again.
Until then my clock is in a set of jars: one for the tiny parts, one for the time train parts, and one for the striking train parts.
Update: Woohoo! Using the anvil, a ball peen hammer, a riveting hammer, and my thumbs, I was able to gently form the strike train great wheel tension washer back into shape. Then using a pair of snub-nose pliers, I – again gently – snapped it back into place. Score! Both great wheels are now snugly held onto their respective arbors by their respective tension washers.
I reassembled the clock, fully wound both of its mainsprings, then set it running on its 31-day run test at about 5 pm December 5th. If all goes well, it should still be going and striking on the evening of January 5th: Twelfth Night.
]]>
Since my last post I’ve reassembled and oiled the clock. I was planning to post a video of the reassembly, but since then I’ve found quite a few videos of that sometimes time-consuming process. Being inexperienced, it took me about 2 hours to reassemble this clock’s movement.
Now that the movement is oiled and assembled, the next step was to cut the pendulum to length. The pendulum on this clock consists of a suspension rod and a bob (weight) that hangs on the end of the rod. The top of the rod is fastened to a short strip of spring.
You may recall in an earlier post I calculated the desired period of the clock, 1.03434 seconds per period, based on the number of teeth on the gears in the time train. In that post I also calculated the theoretical length of the pendulum, 265.6 mm, based on a formula for calculating pendulum’s period given its length.
The difference between the theoretical and real length of the pendulum is very difficult to calculate. Luckily, there is a much more practical way to cut the pendulum suspension rod to the right length: to calculate the difference between a longer pendulum and the desired one:
Following this method, I mounted the clock movement on a test stand, bent the end of one of the suspension rods I ordered, and set the clock running with this over-long pendulum.
To measure the period of this pendulum, I used a Zoom H4n audio recorder, set to record at 44K samples per second, and used Audacity, an open source audio editing application, to measure the time between ticks.
Using Audacity, I measured the over-long pendulum period at 1.128 Seconds per period. Using the formula I explained in the earlier post: L = ( T / (2 * pi)) ^ 2 * 9.8, I calculated the theoretical length of the over-long pendulum to be 315.8 mm.
The difference between the current theoretical length of 315.8 mm and the desired theoretical length of 265.6 mm is 315.8 – 265.6, = 50.2 mm, so I needed to shorten the pendulum suspension rod by about 50 mm. Just in case I’d miscalculated, I simply bent the pendulum 50 mm higher than the current bend, leaving the extra length attached for the moment.
I then measured the period of this new pendulum, again using Audacity, and saw a period of 1.030 seconds per period. That number was close enough to the desired period of 1.03434 seconds. To double-check the length, I compared the new height of the pendulum bob to the height of the window in the clock’s case: they were close enough; math works!
As a last step in regulating the clock movement, I repeatedly measured the pendulum period (using Audacity) and adjusted the bob’s height using the Rating Nut at the bottom of the bob, until the pendulum had a period of 1.0343 seconds per period – very close to the right value.
By the way, in a nearly-1-second-per-period pendulum, an error of 1 ms per period translates into an error of 60 * 60 * 24 ms per day, or 86.4 seconds per day, which is about 1 and 1/2 minute per day, which is terribly inaccurate! In contrast, an error of 0.1 ms per period translates to 8.6 seconds per day, or about 1 minute per week, which is about the best you can expect from a pendulum clock like this one.
Happy with the pendulum period, I set up the test stand in a back bedroom where it can run undisturbed for a month, wound the strike and time mainsprings fully, set the time, and started the clock.
Because it’s a 30-day (31-day) clock, a month from now the clock should still be running and striking. If it isn’t, the clock probably needs new mainsprings. We’ll see in a month.
Spoiler: it didn’t make it. In my next post, I discover there was a dangerously-loose part in the clock.
]]>
I found it in an antique mall in Hillsboro (Oregon). It’s a mantel clock – a Kitchen Clock – similar to my family Seth Thomas clock that is my eventual target for repair(once I know what I’m doing). It also fit the bill of a) not working well, and b) not worth a lot.
The seller was asking $45 for it because it’s an antique Ansonia clock in reasonable working order. I offered $20 and bought it for $25 because the case and pendulum have been heavily altered over the century of the clock’s life.
…not to diminish craft or Crafting. Crafts are artistic expression. It’s just that in this case, I’d like to see more of the original clock that’s hiding under all this paint and alteration.
The first thing I noticed was that the case had been painted… multiple times. The whole thing had been painted gold, and the base showed traces of white underneath the gold. I knew I was going to have to remove all that paint to get to the original wood underneath.
I later learned from Wes at The Clockmaker’s Gallery in Eugene, Oregon that people often paint kitchen clocks that have broken gingerbreading (see below), because any new wood they add wouldn’t match the type and grain of the old.
The inside back had been papered with Christmas wrap.
You can also see in the above picture that the wire running from the alarm to the movement is missing, and the pendulum bob has ugly blobs of solder on the top. I later found this bob is likely not original – the original bob was likely a lovely silver-colored one, with concentric rings and an ornate floral top. At any rate, I knew I’d need to remove that paper.
The face had been painted in a floral motif. Unfortunately, it’s almost impossible to remove such paint without ruining the underlying paper face. Fortunately, I kind of like how some past owner painted this face.
I also noticed the gingerbreading – the elaborately cut wooden plate – from the top of the clock was missing. It seems to have been cut off at some point, possibly in the 1920s or ’30s, when clean, geometric lines had come into fashion and the ornate Victorian styles were out. It also seems part of the gingerbreading may have broken off – I may never know. At any rate, the top arch of the clock is all that remains of the lovely original woodworking. I’ll probably scrollsaw a replacement.
Compare the clock as I found it…
…to a nearly original condition one that appeared in a discussion on an NAWCC message board.
Identifying and dating a clock can be tricky, but in this case, there are several strong hints.
First, the movement and face of the clock are labeled “Ansonia Clock Co.”
Second, what looks like the original paper label remains, in poor condition, on the back of the clock. A little work in a photo editor brought out the model name.
So I’m confident that what I have is an Ansonia “Derby” model, 8-day, striking clock with alarm. One helpful poster to that thread above showed that the Ansonia Derby was first manufactured in 1894. So this clock is no older than 1894. I don’t know when Ansonia discontinued this model, but Ansonia went out of business about 1929.In my next post about this clock, I remove the damaged remnant of the original gingerbread, in preparation for attaching new gingerbread.
]]>
Now that I’ve searched for antique clocks here and there, I’ve found that a number of clocks for sale As-Is have no pendulum bob. Looking at eBay listings, I suspect it’s because the seller can sell the clock in parts (bob, case, works) for a higher price than the whole clock.
Pendulum parts can also be broken and tossed out: the delicate spring at the top of the pendulum suspension rod is easily broken. I suspect that’s what happened to my clock.
Parts can be lost in the process of donating an old clock to a thrift store: the person getting rid of the clock may not take care in packaging the pendulum parts (and the key) with the clock.
For all these reasons, it’s a good idea to find a supplier of replacement pendulum parts. I’m happy with Timesavers.com; others like Merritts, UsedClockParts.com, ClockWorks, or others.
I was delighted to find Timesavers carries a suspension rod that sounded like a simple replacement: “Korean 31-Day Suspension Rod”. But I was puzzled by the length: it’s listed as 12 inches long, and my calculation of pendulum length was a lot shorter. Eventually I realized that the shipped suspension rod is designed to be too long so that I can cut it and add a hook at just the right length. I bought 2 rods just in case I mess up the first one. I’ll be cutting that rod once I have the clock running again.
The pendulum bob was a little trickier to choose. There are a bewildering number of bobs to choose from. I narrowed it down to a smooth, brass bob with a “rating nut” (fine height/speed adjustment), and a loop at the top. Then I simply picked a bob that looked the right size: “2-1/4″ Brass Adjustable Bob”. Once I put the works back in the case, we’ll find whether I made the right choice.
Incidentally, I also bought a new winding key to replace the missing one: “#7 Extra Large Wing Single End Key-3.8mm”.
In my next post about this clock I cut the pendulum to the correct length, and adjust the clock to the correct speed.
]]>
It turns out that figuring out which gears are part of the Chime Train and which are part of the Going Train was a pretty straightforward process of identifying the gears unique to each Train, consulting the photos I took before disassembling the clock, and a little bit of trial and error.
TL;DR – I have a video showing the successful spinning of the going train gears – whee!
A few parts of the Going (time) train were easy to identify:
From my photos I could see that the Going (time) Great wheel (mainspring gear, upper left in the photo above, turned to see from behind) wound counterclockwise, compared to the one for the chiming train. The Escape wheel (lower right) is easy to pick out because of its pointed teeth, that the Escapement ticks and tocks its way through. The Center wheel – that the minute hand attaches to – (upper right) has a a very long Arbor (shaft). In this clock, the hour gear (lower left) has the unique Snail (shell shaped gear) that controls the number of chimes that happen for each hour. The hour gear also has a hollow arbor that fits over the center wheel’s arbor.
Similarly, I could pick out a few gears that were clearly part of the Chime train:
The chime great wheel (left in the photo above, seen from behind) is wound clockwise. In this clock, the Fan – the governor that controls the speed of the chiming (far right) has a small pair of weights attached to springs. Another chiming gear has a pin sticking out of it (lower center). I’m not sure exactly what that pin does yet, but I know the pin has to do with running the chimes. [Update: it’s the “Warning” pin, that starts the chiming] Finally, I recognized the gear with the round holes in it (upper center) as part of the chime train in the photos I took before disassembling the clock.
Once I’d separated the easily-identified gears, there were only a few gears remaining. I’m surprised how few gears a clock like this has.
None of these gears were particularly easy to identify in the photos I had taken – oops, so I had to do a bit of trial and error by trying to assemble the gears into the back Plate of the clock and seeing what fit and seemed to mesh into which gears.
With only three unknown gears, it didn’t take long to find which gears fit where.
You may have noticed in the earlier pictures that a little stub is the only part of the pendulum this clock still has. I suspect the pendulum suspension spring finally broke off as the clock aged. …or maybe someone cut the pendulum off to sell as a separate part. I’m not sure. At any rate, this clock’s pendulum is missing.
I’ve found a replacement suspension rod, and I suppose I can buy a Bob (pendulum weight) to fit it, but before I do that I want to know the length of the ideal pendulum: the distance from the pivoting top of the pendulum to the center of gravity of the total of the suspension rod and bob.
Actually, I went though all the calculations below before I found replacement parts, when I thought I’d have to make my own pendulum. Now they’re here mostly to show that math does work.
The first step to calculating the pendulum length is to calculate the pendulum period – the time for a tick plus a tock – and to do that I needed to count the teeth in all the gears of the going train. Here are the results:
Great Wheel = no pinion leaves (pins), 84 teeth
2nd Wheel (that meshes with the Great Wheel) = 8 leaves, 60 teeth
Center Wheel (the minute hand wheel) = 12 leaves, 68 teeth
4th Wheel = 7 leaves, 66 teeth
Escape Wheel = 7 leaves, 38 teeth.
Additionally, the Center Wheel is the start of a small gear train that runs the hour hand:
Center Wheel = 15 leaves
Funky little wheel (I don’t know the name), driven by the center wheel = 40 teeth (would be pinion leaves), 10 teeth
Hour Wheel = 45 teeth.
From all those numbers, we can calculate the gear ratio – the number of turns a second wheel makes for each turn of a first wheel – based on the number of teeth and leaves in the two gears. For background information, see Working out Gear Ratios at TechnologyStudent.com.
For example, suppose we want to know how many turns of the center wheel are required to turn the hour wheel one turn. In other words, how many hours (full turns of the minute hand) are in 12 hours (a full turn of the hour hand)?
The funky little wheel turns 45 / 10 turns for every full turn of the hour wheel. Similarly, the center wheel turns 40 / 15 turns for every full turn of the funky little wheel. So the number of turns of the center wheel for every one turn of the hour wheel = 45 / 10 * 40 / 15, which comes out to 12. That is, the minute hand turns 12 revolutions (12 hours) for every one revolution of the hour hand (12 hours).
So, our gear ratio math seems to work.
Now we can calculate the time per tick-tock of the clock:
The escape wheel has 38 teeth, and the escapement advances one tooth per tick-tock pair, so it takes 38 pendulum periods to turn the escape wheel one full turn.
One turn of the 4th wheel, which drives the escape wheel, takes 66 / 7 turns of the escape wheel. One turn of the center wheel takes 68 / 7 turns of the 4th wheel.
So the pendulum periods per full turn of the center wheel is 38 * 66 / 7 * 68 / 7, which is about 3480.49
The turn of the center wheel takes an hour, which is 60 minutes of 60 seconds each, so to convert our pendulum periods per center wheel turn into per second, we take 3480.49 / (60 * 60), which is about 0.966803 pendulum periods per second.
But we want seconds per period, so we calculate 1 / 0.966803, which is about 1.03434 seconds per pendulum period. I was puzzled why the clock designers picked this number instead of a round 1 second, until I found a hint on this thread about gear trains: “With an ‘integer ratio’, the same pairs of teeth (gear/pinion) always mesh on each revolution. With a non-integer ratio, each pass puts a different pair of teeth in mesh. (Some fractional ratios are also called a ‘hunting ratio’ because a given tooth ‘hunts’ [walks around] the other gear.)”
So it seems clock designers prefer non-whole-number gear ratios to even out the wear of the gears’ teeth. Who’d have guessed?
The basic physics formula for an ideal (friction-less) pendulum’s period, given its length is: T = 2 * pi * sqrt(L / G), where T = the time of the pendulum’s period, in seconds; pi = 3.14159…; sqrt() represents the square root; L = the length of the pendulum, in meters because I like metric; and G = the acceleration due to gravity, which is about 9.8 meters / second / second.
That’s fine, but we want to know the ideal pendulum’s length given its period. That is, we need to solve the equation above for L, given T. Algebra comes to the rescue and produces: L = (T / (2 * Pi))^2 * G.
Given a period, T, of 1.03434 seconds, Pi of 3.14159…, and G of 9.8 meters per second squared produces a result of about 0.265579 meters. Millimeters are a little more convenient, so we multiply by 1,000 to get an ideal pendulum length of about 265.6 mm.
To roughly double check the math, I lined up a meter stick with roughly the suspension point of the pendulum and held my thumb by the 257 mm point on the stick: it roughly matches the middle of the window where this clock’s bob should be. So the math is in the neighborhood of correct!
Here’s a head-scratcher: the replacement suspension rod (pendulum rod) for this clock is 12 inches long, which is 304.8 mm rather than 265.6 mm. Why?
The difference in length is because the suspension rod has weight, which makes the center of gravity of the rod + bob higher than the center of the bob. I’m assuming that the suspension rod is heavy enough, compared to the proper bob, that the center of gravity of the whole rod + bob is about 265.6 mm. I’ll know once I get the replacement parts in and get the clock running again.
Update: Now that I have the replacement suspension rod, I see that it’s deliberately as long as you could possibly need, and I’ll need to cut it to length and bend a hook in the end to hold the bob. To cut it to length, it will be useful to have the 265.6 mm estimate.
Later Update: I’ve learned that the theoretical pivot point of the pendulum is somewhere in the middle of the suspension spring, depending on the thickness of the spring, because that spring bends as the pendulum swings.
This clock is advertised as a 30-day clock. That is, you need wind it only about once per month. How many turns of the winding key is that? That is, how many full revolutions of the going train great wheel happen in 30 days?
The center wheel (minute hand) rotates once per hour. The 2nd wheel turns 12 / 60 turns per turn of the center wheel. The great wheel turns 8 / 84 turns per turn of the 2nd wheel.
So for every hour, the great wheel turns 12 / 60 * 8 / 84, which is about 0.0190476 turns per hour. There are 12 hours per day and 30 days per winding, so we multiply by 12 * 30, which is about 6.89 – let’s say 7 – turns per 30 days.
So every 30 days, you should need to turn the winding key 7 full turns. I’ll be able to verify this number once I get the clock running.
In my next post, I buy new pendulum parts for this clock, to replace the missing ones.
]]>
After a few hours of scrubbing parts with SOS pads and toothbrushes, then rinsing in water – twice to get all the soap scum off – then rinsing in alcohol (wearing gloves this time), the clock parts are all “clean”.
I’m happy with the results of this first try: the clock is a lot cleaner than it was.
At the same time, I noticed a few things that an ultrasonic cleaner would likely clear up.
In the photo above, you can see gunk remaining on the pinion (the little gear) of the fly (the speed control for the chiming). Because of the delicate weights in this part, I was reluctant to scrub it.
Also, in the photo at the beginning of this post, you can see all the tiny parts that aren’t cleaned much at all. Again, because they were delicate, and easy to lose, I only gave them a little swish in alcohol instead of the full cleaning.
Another class of poor cleaning is a lack of disassembly: in the part above, I don’t yet know how to remove the spring clip, so I cleaned the part without removing the clip. Dirt and rusting water will remain under the clip regardless of how I clean the assembled parts. You can even see some rust on the gear, possibly from a previous, similarly-incomplete cleaning.
In my next post, I figure out which gears are part of the time train vs. the chime train, and calculate the ideal length of the pendulum.
]]>
The unattributed VHS video I mentioned earlier shows how to repair a clock on a budget. Instead of using an ultrasonic cleaner, that person does it Old School: he uses an SOS pad (soaped steel wool) and a toothbrush to do the cleaning, then rinses the parts in water, then does a final rinse in alcohol and dries the parts. The steel wool helps remove rust and grit; the soap cuts the grease and oil; the water rinse removes the soap; the alcohol rinse drives the remaining water out; the final dry removes any remaining soap scum or alcohol.
In the picture above you can get an idea of how dirty this clock was: all the holes were gummed up with old oil and dirt. The wheels were also very greasy, and there was a huge amount of dirty oil on the mainsprings. All that dirt can keep the clock from running like it should, and can cause the parts to wear prematurely.
As a first step – unlike in the budget cleaning video – I used toothpicks to scrape out old oil, dust, and who knows what else from the pivot holes and the “oil sinks”: the little indentation around the pivot hole. Then I wiped as much old oil off as I could, using paper towels. Then I was ready for the real cleaning.
Like the video suggested, I used an SOS pad (soapy steel wool), a toothbrush, and water to break down the old oil and rust.
After the rinsing in water, the alcohol rinse, and the drying, I was pretty disappointed in the results. As you can see below, there isn’t much visible difference between the clean and dirty parts.
You can also see a lot of remaining oil and gunk (dust, ground brass, and old oil) in the photo below. The steel wool also scraped up the brass quite a bit, which you wouldn’t want to do to a good clock.
So now my plan is to “clean” this clock with the SOS pad, etc., and to buy an ultrasonic cleaner for the next clock. I see I can get a good, “mini” ultrasonic cleaner for just over $100, that should be big enough for the mantel style of clocks I’m interested in repairing.
I mentioned above that a dirty clock can wear parts prematurely. Another part of a clock overhaul – which I won’t be doing anytime soon – is “bushing” worn pivot holes. As you can see in the photo below, the circled hole is nowhere near round; over the years, the pivot in this hole pressed against one side of the hole, causing the hole to wear in that direction. Out-of-round holes like this keep the clock’s parts from meshing correctly, which can make the clock stop before it should, or even not run at all.
Bushing a pivot hole involves drilling the hole out so it’s round and centered where it should be again, then pressing in a brass fitting that has a hole the right size for the pivot (the end of the wheel that fits in the hole).
Unfortunately, a decent bushing tool – which is a precision machine – can run $1,000, which is well beyond my Hobby Exploration budget.
In my next post, I finish cleaning the clock and have a look at the remaining dirt.
]]>