Back to Notes From The Bench index    back to home page

   "And ever has it been known that love knows not its own depth until the hour of separation."    
    ...Kahlil Gibran The Prophet  (from the first chapter: "The Coming of The Ship") ¹

Clock Repair Archive - -

An Essay on Clock Repair

Time can be an elusive concept. Sometimes an hour can be very short, or very long depending on your perspective; the above quote reminds us of this fact. Clocks measure time giving us a comforting way to have some control over our lives by being able to count on something predictable and this may be what makes them very interesting to many people. If you are a person who does not necessarily want to fix clocks, but you are interested in them, please read on. If you are having a clock fixed, as much as possible, be patient with the person doing the work. If something goes wrong with the timekeeping, they will need time to check its operation. Clock repair is as much an art as it is a science. If you rush them you will only hurt yourself and your timepiece because it is not possible to change the passage of time.

To be a clock repair artist requires many years of experience. One must be humble enough to learn from their mistakes and admit it to themselves and at the same time, still maintain a level of confidence far above the mediocre. Others must feel the confidence of the repair person, or they will not trust their treasure to them for one precious moment, and the repair person will have nothing to repair.

Because there are thousands of different types of mechanisms, an apprentice will spend about the first 10 years learning how to repair, and the rest of their career getting better and faster at it. A good repair person will need to get to the point where they can take a mechanism apart and know exactly where each part goes without having to think about it. Possessing a photographic memory along with all the other skills needed to do clock repair will allow an apprentice to learn quickly. Short term memory, however is not as important as long term memory. For example, it might be 2 years between the times a certain type of clock is in for repair. The repair person's long term memory had better be good or they will not have a chance of fixing it without writing down detailed notes about all the work. Remembering concepts is just as important as remembering parts placement. If the mechanism can be seen in the mind and every assembly detail remembered exactly, but the problems cannot be recalled, the repair person will fail. Clock repair is very demanding work. The reward must be the feeling of accomplishment and pride at having restored a timepiece. the money earned must be a secondary concern and nothing more.

A person who repairs clocks must be able to feel a sense of fulfillment and achievement as they become proficient at the craft. Self discipline is of the utmost importance, because without it they will fail for sure. They must be a perfectionist; but not to the point of obsession. This craft is art and technology combined in a very unique way. These artists work with their hands, and the majority of their skill is in the ability to "see" and "feel" how things work. Modern tools such as a drill press, micrometer, calculator, propane torch, and many assorted hand tools are used, along with tools the repair person will make as necessity dictates. The main tool that a clock repair artist uses is the watchmaker's lathe; which really should be referred to as the "clockmakers lathe" but this is not customary terminology. This lathe is a small precision machine designed for working on watch and clock sized parts and it is often used for making tools.

A clock is, in fact, a mechanical device with integral parts that fit together in a predetermined way much like our modern concept of mathematics with its roots in the mechanistic thinking from the time of Sir Francis Bacon, Sir Issac Newton, and Rene Descartes; however an interesting set of circumstances comes into play as a clock runs. The natural world of chaotic events that cannot be fully explained by this mechanistic way of thinking affect the clock as time passes and it runs: bushings wear, springs become fatigued, oil dries out, pivots wear, gear teeth wear, hammer pads wear out and these events do not happen with simplistic predictable regularity because many of them are heavily influenced by human activity which is often quite chaotic. Forces of nature that change air temperature and moisture content in very complex ways also effect the condition of a mechanical timepiece. So what is, in fact, a very well organized mechanical device becomes chaotic in nature as it runs and is influenced by the complex array of non-mechanistic events taking place around it - sort of like trees in a windstorm. Consider the question: "If a tree falls in a forest, does it make a sound"?. If you are one of those who say immediately, why yes of course, because you can hear it... and we all know you certainly can hear a tree falling in a forest, right? Well you'd better think twice about that one. First, how large is the forest, and how far away are you? Have you ever been in a forest in a fierce windstorm? Do you know how much noise there is in a forest in a windstorm? How does a tree fall in a forest? Does someone cut it down? Does it just fall when it is dead? Does it get ripped up by the roots? Have you ever seen and heard a tree fall in a forest in a windstorm? Determining if a tree makes noise when it falls is likely to be the last thing on your mind. Seeing and getting to safety is most likely going to be what you will want to do. Trees often don't hit the ground hard when they fall, because there are other trees around that catch them, and there is so much noise from the windstorm, that there is no way you are going to hear a tree fall unless you are standing right below it because the sound from snapping branches will not be much different from the sound of trees falling. Am I digressing? I think not; there is a connection between trees falling and clock repair. Most people who have not spent much time in forests may be able to grasp this concept more fully than complex mechanical interactions that take place within a clock. The "theoretical tree" simply falls to the ground; a logical and predictable event. The "real" tree does not always fall like the one in theory but it is easy to visualize a forest. In much the same way, in theory and design, a clock is ordered, mathematical and has predictability, however most people are not clock repair technicians and don't know what happens when a clock is effected by the actions of motion and friction which together, are in turn ordered by chaotic events. The overused expression about a tree falling suggests a preconcieved anthropocentric notion of predictability regarding things of nature. This is true sometimes. There will be a sunrise and sunset due to the rotation of the earth, there will be spring summer and fall due to the regular periodic tilting of the planet; and due to the force of gravity, trees will fall; but not with the kind of predictibility that can be found in man-made machinery. So there is some predictiblily in nature and some chaos. The same is true when describing the nature of clock repair: in theory and design clocks are predictable and mathematically perfect; but as they run and wear out they become chaotic. Even the slightest imperfections are amplified as the clock runs. A microscopic etching in a bushing or pivot will increase in size very slowly over the months and years. A spec of dust in a sleeve bearing will slowly wear away the brass inside, sometimes in a random fashion, often in a regular pattern. Changes in temperature and humidity are often random and will influence the clock mechanism accordingly. When a tree "falls" it does just that; but it is not that simple because trees exist in diverse environments where things become chaotic. If you can accept and understand what is being said here regarding chaotic events, mechanistic predictability and their relationship to each other, then you have one of the main qualities that a clock repair person must absolutely have, and that is the ability to think creatively and logically at the same time. The point is, that what seems totally logical and normal does not apply in clocks and clock repair the way we are all used to. One plus one equals two, but not right now, not until the trip cam gets to the 90 degree position and releases the strike gear train.

Clock repair must never be confused with activities like appliance repair, or automobile repair. Clocks measure time, this is their function and to test their function therefore, takes time if it is to be done with the utmost precision and care. The meaning of intervals of time, the action of levers and the nature of motion are all familiar to the clock repair person; and they have the ability to use inductive, deductive and abductive reasoning to analyze mechanical devices and to decipher things other people tell them about problems with clocks. The inductive and deductive reasoning are used in the basic repair process which is part of making decisions based on observations and / or a set of rules. The abductive reasoning is the problem solving as it relates to a clock that will not function correctly after the owner has had it for a while. For example, a repair person will frequently be confronted by problems that appear to defy logic and they must solve these without fail every time. There will be some data that is fact, some that is not, and possible conclusions that may or may not be accurate. This is the most difficult part and requires the most experience; and furthermore, this skill can only be acquired from experience. This is where most people fail because this is the most important part of the work.

A good clock repair artist will be honest with customers. They will, by necessity, always seek more knowledge by reading books and making careful observations as they work. A strong knowledge of physics, chemistry and mathematics is necessary to be successful at clock repair. For example, formulas from trigonometry and geometry are used to figure angles and distances ( lift angles and gear depthing ) and vectors to figure forces ( how much power is available from a mainspring at the center of the arbor ). Rational equations are used to figure gear ratios and power ratios; and to fully understand about the actions of a pendulum ( harmonic motion ) requires the application of knowledge acquired in the study of physics. A knowledge about acids and bases and their effect on various metals is necessary to understand fully what happens to metals as they heat and cool, including the effect of minor temperature variations. Along with all this a successful clock repair artist must be a strong visual learner with excellent cognitive reasoning ability - a very unusual combination.

Clock repair involves becoming intimately familiar with the way gear trains work in clocks. The repair artist must know their operation by feel and by logic, and they must know both ways or they will fail. There must be a knowledge of magnitudes; for example, what it means when one force is 10 times stronger than another. Another example of this can be illustrated by looking at motion. Most people can run about 10 miles per hour. Ten times faster than that is what? Now take that times 5. That is 500 miles per hour. Now go backwards from that. It is important to know the nature of the power that causes a pendulum to swing in most clocks in order to fully understand what makes them run and stop. After the escape wheel, there still must be power delivered to the verge, and that usually has a long wire pushing the pendulum. A further loss of power.

Clocks work in a different reality than most mechanical devices. Most clocks have five gears and the turning force is applied from the main power to the escape wheel through a gear reduction system working in reverse - instead of small gear driving large gear to reduce the speed and increase the power as in an automobile, the power starts out with a large gear. This gear typically has 65 teeth or so. That gear is either attached to a mainspring or a weight cable. That large gear then drives a much smaller diameter pinion gear of five or 10 teeth (typically the difference in diameter is a factor of between 4 to 8; the pinion is 1/8th to 1/4th the diameter of the larger gear supplying power to it). This pattern of larger gear driving smaller gear continues up the gear train all the way to the escape wheel. Some clocks have six gears in this system and some have less. Now if you have studied math and physics at the high school level you will know without even testing, that when power is applied to the large gear, the escape wheel will spin at an incredibly high rpm when the mainspring is wound; or when the cable is wound up with a weight pulling on it. You may even realize that the escape wheel's speed will be limited by the friction in the gear train and not necessarily by the power of the spring or the amount of the weight. If the friction were reduced enough the escape wheel would likely break apart from excessive speed. This does not happen in real life, because the friction of the gears and the bushings keep the speed of the wheel down. The escape wheel will have such little power that it can almost be stopped from moving by the pressure of the weight of a feather. By the time the power transfer gets through the gear train from the powerful mainspring to the escape wheel there is such a small amount of power (torque, or twisting force) left that it is all the mechanism can do to just turn the escape wheel. It is the reduction of torque that allows the clock to maintain potential energy while using very small amounts of kinetic energy and thus run and keep time for (typically) 8 days. The part that causes problems is the friction. This is an unknown; a variable that depends on the condition of the clock, the oil, and the amount of wear in the bushings and the amount of wear on the gear teeth. The person repairing clocks knows this because they deal with it very often and it is critical information. An experienced repair person can tell by feel whether or not there is enough power to run the clock they are dealing with. The ability to intuitively solve for this unknown cause of friction will determine success as a clock repair artist. The person repairing clocks has a complete understanding of this motion, if they did not they would not be working in the trade.

A person who repairs clocks must have the ability to concentrate on sounds and listening. They must listen to the gears, listen to the sound of the mainspring winding, and the sound of the ratchet clicking as the clock mechanism is wound. After a person has been startled by a defective ratchet, they will become a very good listener or they will loose the ends of their fingers. A ratchet has a certain sound when it is working correctly. A solid sound. Not a tight sound, and not a loose sound; a solid sound. An experienced clock repair artist knows to look at the ratchet as it works, watching the rivet or the mounting screw that is holding it to the plate (some clocks have a rivet holding the ratchet and other clocks have a screw). If the rivet or the screw moves when the clock mechanism is wound they are fixed: not made too tight, and not made too loose. The repair person listens and observes carefully. They listen to and watch every thing on the clock they are repairing. They listen to the customer because they know if a customer complains, there is always a good reason. If there is a complaint about the sound, a good ear for music will be needed to hear if it is out of tune. There is the need to listen for odd vibrations that make the sound disagreeable. The repair person's job is to find out what is wrong. They are the cat in the window. They miss nothing. They hear everything. They can see and hear what others do not.

When considering an old clock the study of mathematics must be kept in mind, along with the realization that most of the math in use today has its roots back thousands of years. A repair person will know that 150 years ago the math needed to design a good clock was available; and they will not be so full of 21st century arrogance so as to assume they can improve on the design of an old clock. An experienced mechanical engineer humble enough to recognize the contributions of persons in the past might be able to improve the design of a clock; but consider this: a clock may work, but how long will it work? What will happen to its function over, 15 years? 20years? 50years? 100years? There are at the very least, tens of thousands of clocks over a century old still running and keeping good time, requiring only the occasional attention of a skilled repair person. Why? Think of the gears. Brass, and steel. Pinions are steel, larger gears are brass, hard brass. Pinions typically have between 5 and 15 teeth. Large gears have 55 to 75 teeth, so there are more brass gear teeth with less individual active wear processes on their working surfaces than on the steel pinions. The steel must, therefore be somewhat harder than the brass or it will wear out too fast. Now, where are the forces applied to the gear teeth, and what are their relative magnitudes? How are the forces applied to the bushing surfaces? All of this happens in a slow moving gear train. There is virtually no heat involved in the wear process in a slow moving gear train, so the wear process is completely different than most other mechanical devices. To more completely understand this, consider would happen if the percentages of zinc or copper in the brass of a clock's bushings and gears were to be changed slightly ? Would it wear better over a period of 120 years? You can't just run the gear train at some predetermined super-high rpm for many months straight to get the linear mathematical equivalent of 120 years for the same effect. The results must be observed in real time. What if the clock is in a cabin near the ocean? The moisture in the air will cause the clock to wear much faster than if it were in a dry warm climate. How can it be possible to know this? Repair persons see clocks that come from houses near the ocean and they see how these timepieces wear. It becomes obvious how they wear after they are repaired. They see how other clocks in other places, the same types of clocks, wear after they repair them. There is a difference.

The amount and type of dust in the environment around a clock will also determine how it wears. When a gear and a pivot move slowly, the whole process of mechanical stress on moving parts is different than it is in a fast moving apparatus. Therefore the only way to know for sure is to let the clock run for 120 years and then inspect it.Because events that take place in our environment are by nature chaotic, all the design ability in the world cannot determine precisely how a clock mechanism will wear over time, only time can determine that. After seeing what happens to various types of clocks over long time periods most repair persons with 20 or 30 years of experience can tell you things you can't learn in books or design with computers. The condition of a clock mechanism tells a story that a good repair artist can read like most people read a book. One could almost say that each tick of an individual escape wheel tooth meeting a pallet face is recorded in the mechanism along with any changes made to it over the years.

The design problems of mechanical clocks have all been solved over time by an enormous amount of empirical research. What other man made devices do we have from the past that are still working and relevant after 100 years? The formula for the manufacture of brass used in clock gears and parts is a well kept secret that is held by a select few, and it is the key to the functioning of most mechanical clocks. We can only hope that it is not lost forever when they are gone.

This essay is derived from approximately 25 years of experience in the actual repair of clocks as an occupation. The author worked nearly a quarter century from the early 1970's to the late 1990's as a repair person restoring various types of contemporary and antique mechanical clocks including, but not limited to; spring drive mantle clocks, grandfather clocks, wall clocks, ships bell clocks,cuckoo clocks, and french carriage clocks. This also involved extended interaction with customers as a representative and co-workers as an instructor.

This journal is published under the GNU free documentation license. There are many attacks on the GNU licenses. All of them are based on false assumptions, fear, greed, doubt, and just plain stupidity. If you read the history of math, science, the printing press, and such things involving communication and study you will know that sharing information is the way the human race has managed to survive into the modern age. It is how we discovered how to grow food and help others grow food. Whether you like it or not this is the best way of learning devised to date. It is the new way of thinking. (It is new to technology in this century, but something tells me it has been around a while). More and more people are discovering this to be true. You can either accept it or be left in the dust. The choice is yours. I suggest you read over the license carefully before you copy this journal.

Notes From The Bench index( from clock repair journal )

back to home page

¹ The quote is taken from Kahlil Gibran's book   The Prophet  originally published in 1923; Ninety-second printing, December 1973. Alfred Knopf, New York.

Copyright (c) 2002 David Tarsi. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.1 or any later version published by the Free Software Foundation; with the Invariant Sections being no invariant sections, with the Front-Cover Texts being no Front-Cover Texts, and with the Back-Cover Texts being no Back-Cover Texts. A copy of the license is included in the section entitled "GNU Free Documentation License".