TheAmerican Banjo clock — often just called a banjo clock — got its name from the shape of its case, which resembles a banjo musical instrument. Although the design was patented in 1802 by Simon Willard, a renowned clockmaker from Roxbury, Massachusetts, Willard referred to his invention as the “Improved Timepiece,” not a banjo clock.
Willard-style banjo clock, c. 1810
There are essentially three types of mechanical banjo clocks. Those with lever escapements are usually in the lower price range, spring-driven examples occupy the middle ground, and weight-driven banjo clocks represent the higher end, typically costing several hundred dollars in fair condition to several thousand for authenticated examples from reputable makers.
My wife spotted the clock on Facebook Marketplace one evening. It was listed for $100, and the photo immediately caught my attention. I had a feeling it might be a diamond in the rough. I suggested she offer the seller $75, and to our surprise, the offer was accepted almost immediately. It seemed they just wanted it gone.
Seller’s photo, looking a little tired
The clock was one of several items the seller had for sale as part of a community-wide garage sale. So, a day later, we made the drive out to the rural part of Nova Scotia where the sale was taking place. When we arrived, we found the house and asked to see the clock. It was sitting there among other odds and ends, looking a little tired but full of promise.
Broken dial glass, original hands
As we examined it, I asked a few questions about its history. The seller explained that it had belonged to an uncle in Wolfville, Nova Scotia—a man who had once owned a large collection of antique clocks and had been in the family for over 60 years. That was all I needed to hear.
Before we left, the seller asked if I might be interested in another clock—a 30-hour mantel clock. “It’s worth a lot more than that one,” they insisted, nodding toward the banjo clock. I smiled and politely declined. Something told me the first clock was the one worth keeping.
Classic time-only movement
And I was right. Once we brought it home, cleaned it up, and did a little research, we discovered that it was indeed something special—a genuine, 1830s timepiece, a true diamond in the rough.
The clock was surprisingly heavy when I first lifted it, which told me right away that the weight was still inside the case. That was a good sign. Had it been missing, finding a proper replacement would have been a challenge—and there’s really nothing quite like having the original weight that once powered the movement so many decades ago.
It turned out to be a Federal-style banjo clock, housed in a rich mahogany case that bore the quiet dignity of age. The movement was intact, a promising start. At the top sat a graceful acorn finial, original to the clock—a small but telling detail. At first glance, I thought the carved reverse-scrolled side arms and lower base section were missing, but later learned that not all examples from this period included those features; In fact, simplicity was a defining feature of early timepieces.
The original steel hands, shaped in that classic banjo style, were still in place, and the dial face, though worn, retained much of its character. The glass bezel was broken but easily replaced. The case itself was dusty and a bit tired, with a few areas needing veneer work, a split wood bezel, but nothing beyond repair. As I examined it more closely, I discovered the winding crank tucked neatly into the bottom of the case—perhaps placed there by its last owner many years ago.
After some careful study and comparison with other examples online, and with help from a few knowledgeable members of a clock forum, I came to believe that the clock was made in the late 1830s in Boston—possibly by John Sawin. Sawin had apprenticed under Simon Willard and later worked as a journeyman for Aaron Willard, two of the most respected clockmakers of their day and inventors of the original patent timepiece. Unfortunately, there are no identifying marks on the case or movement, leaving its true maker a mystery—but a fascinating one all the same.
Fully restored clock on display in my home
With that in mind, I decided to give the old clock the attention it deserved. What followed was a careful cleaning, a bit of veneer work, minor repairs, glass bezel replacement, and the satisfaction of seeing history come back to life.
Over the years, I’ve been happily posting two articles every week, and it’s been such a joy to share stories, discoveries, and little bits of horological history with you. But starting next week, I’m slowing the pace just a touch—one article a week, every Monday. Think of it as our standing Monday coffee together, with a side of clock talk.
You can still expect the same mix of topics that make this blog tick—clocks with a Canadian connection, my own adventures (and occasional misadventures!) in restoration and repair, the new treasures that wander into my collection, curious clocks with odd and wonderful histories, and of course, the thoughtful letters I receive from readers around the world. Those things will always be at the heart of what I share here.
A collection of clock movements in need of repair
You might think that posting less means I’m running out of ideas. In truth, it’s the exact opposite! The more I explore horology, the more I discover, the more I learn, and the more I want to share with you. Every clock has a personality and a story to tell, and there’s always another fascinating find or repair puzzle waiting just around the corner.
With nearly 1,000 articles already here (use the search feature), there’s plenty to explore while you’re waiting for the next post—whether it’s restoration tips, a deep dive into an unusual clock, or an interesting slice of history. This new schedule just gives each post a little more room to breathe, without changing the care and attention I put into them.
Arthur Pequegnat shelf and mantel clocks
Thank you for reading, for sharing your thoughts, and for being part of this little corner of the horology world. I’m excited for our new Monday routine—and I have a feeling you’ll enjoy the new pace just as much as I will.
Until next time—keep your clocks wound and your curiosity ticking.
We are quite accustomed to the accuracy of quartz clocks which lose or gain mere milli-seconds per week. Contrast that with an era when folks were content to accept that their mechanical clock would be a minute fast or slow through the week, and it was a common practice to make small adjustments over the run cycle of a clock.
Clocks generally fall into four categories: quartz, electro-mechanical, electric, and mechanical. In this discussion, we’ll focus on mechanical clocks—whether antique or vintage—as we explore the common reasons why a clock may run slow.
Our discussion on why a clock runs slow should be grounded in a few practical realities. A typical American spring-driven clock, even when properly serviced, may gain or lose a couple of minutes per week—this is considered normal. In contrast, weight-driven mechanical clocks that vary by only a few seconds per week are regarded as highly accurate, with astronomical regulator clocks being among the most precise of all.
That said, how many mechanical devices do you know run perfectly (relatively speaking) after 120 years?
A clock runs slowly for a variety of reasons.
Let’s consider several factors and examine each in detail.
Environmental factors
The pendulum is too low or too high
The pendulum is the incorrect weight
Incorrect suspension spring length
The suspension spring is not attached correctly
Lack of lubrication
Gummed-up lubrication (over-oiled)
Balance Wheel Requires an Adjustment
A weak mainspring
Changes or Alterations During Servicing
Clock cycle time variance
Slipping, Binding and Rubbing
One-weight wall clock with a large pendulum bob and rate adjustment on the bottom
Environmental Factors
Mechanical clocks are subject to a number of environmental factors that may cause them to gain or lose speed over the course of a year. These factors include heat, cold, and humidity. Increasing the ambient temperature of a clock will slow it down from the expansion and lengthening of the pendulum unless the pendulum is a compensating type using Mercury or dissimilar metal rods. Denser air also causes the pendulum to move more slowly.
Ogee clock showing replacement 1 oz rate adjusting pendulum bob
Moving a clock from sea level to a higher elevation will affect the speed of the clock.
It is common to make seasonal speed adjustments on a clock that does not have a compensating pendulum.
Pendulum bob with inset rate adjustment
Pendulum Too Low or Too High
The lower the pendulum bob, the slower the clock will run. Many pendulum clocks can be adjusted either by a set screw at the bottom of the pendulum or by an inset screw on the pendulum.
In the absence of an adjustment on the pendulum, there is a regulator on the clock face. Use the small end of a double-sided key and insert it into the dial face of the clock to speed up or slow down the clock.
Shortening the pendulum will speed up the clock. Anything that increases the length of the pendulum will reduce the rate of the pendulum and result in a clock that will run slower.
Parts of the clock related to the pendulum
Pendulum is the Incorrect Weight
If the pendulum is too heavy, it causes the centre of gravity to be too low, consequently, the clock will run slower. Having the correct weight pendulum for your clock ensures smooth running.
Incorrect Suspension Spring Length
Often, when a clock is repaired by a person who has limited knowledge of the effect a replacement spring will have on the running of a clock they will occasionally install an incorrect length or thickness of suspension spring. Choose the suspension spring that is correct for your clock.
Rate adjustment under the 12. Use the small end of the double-sided key to make the adjustment
Suspension Spring Not Attached Correctly, or Bent or Twisted Spring
The suspension spring is located at the top of the pendulum rod and serves as the flexible part that allows the pendulum to swing. It connects the top post to the pendulum leader. If it is not installed securely, the pendulum may not swing at all or could wobble, which robs the movement of its energy.
If the spring is bent or twisted, the imperfection will also drain its energy. This imperfection reduces the efficiency of energy transfer to the pendulum, causing it to lose amplitude and making the clock run slow or stop altogether.
Lack of Lubrication
Pivot holes that have dried up mean that there is no lubricating barrier between the pivots and their bearing holes, although the movement may otherwise be very clean.
Small drops of clock oil applied to the dry pivot holes will ensure the clock runs well and will have a long life. Without oil, the steel pivots will wear the brass pivot holes, resulting in wheels that will not mesh properly, eventually stopping the clock.
An approved clock oil must be used, although I have heard that synthetic oil works well too.
Note: a small drop of oil in each bushing hole is all that is required. As they say, Less Is More!
Gummed-up Lubrication
Clock oil loses its viscosity over time. Aging oils also lose their adhesive properties. Oil becomes oxidized, and oxidized oil forms varnish-like residues or gums that can clog pivots and bushings, increasing friction.
When a clock runs slowly, the first instinct is to apply more oil. Old blackened or greenish oil in the pivot holes is a sure sign the clock has been over-oiled or the oil has degraded. The addition of fresh oil produces an almost immediate improvement in the running of the clock, but it will not be long-lasting. In a short time, the clock will begin to run slowly again as the new oil mixes with the dirt and grime in the old oil. When this occurs, the oil becomes an abrasive paste, resulting in exacerbated wear.
The only solution is servicing, which includes disassembly, cleaning of the movement, addressing wear issues, reassembly, and testing.
Balance Wheel Requires an Adjustment
Balance wheels can be adjusted to speed up or slow down a clock.
Regulation of the escapement is done by sliding the two small weights attached to the center of the balance wheel inward to make the clock run faster, and outward to slow it down.
To adjust, simply hold the wheel and push the small adjustment “finger.” Moving the finger to the right will speed up the clock, and moving it to the left will slow it down. Each movement of one dot typically changes the rate by about 10 seconds per day.
The movement usually has a directional indicator guide near the balance wheel, with an “S” for slow and an “F” for fast.
Balance wheel escapement
A Weak Mainspring
Often, the mainspring you will find in your antique clock is the original one(s). The steel used at the time the clock was made was generally of higher quality than the steel used today, with some exceptions. By their very nature, mainsprings become weak over time.
Weak mainsprings are called “set” mainsprings. If “set”, your clock will not run a full cycle, 8 days for eight-day clocks, a full 30 hours for one-day clocks, or whatever the designed cycle. When a spring-driven clock is brought in for a professional repair, the mainsprings are often replaced.
Most properly serviced clocks with their original mainsprings will complete their full cycle, even if the springs are not as powerful as they once were, because cleaning reduces resistance throughout the gear train.
If your clock requires a mainspring replacement, a correctly sized, quality American- or German-made mainspring should provide years of reliable service.
Changes or Alterations During Servicing
Changing or altering the mechanism, such as replacing a gear with an incorrect tooth count, may speed up or slow down a clock. Although movement parts may appear to be similar, manufacturers often made small changes, resulting in parts that may not be interchangeable with the exact movement over the years.
Clock Cycle Time Variance
American spring-driven eight-day clocks typically run slightly faster at the beginning of their cycle because they provide most of their power early on, then run more slowly throughout the week as the mainspring unwinds. A spring-driven clock that is one or two minutes fast at the start of the week may often be a minute or two slow by the end of its cycle. This is considered normal, and no adjustment is necessary.
Occasionally, you will see Geneva Works or stop works on a clock movement. These mechanical additions are designed to compensate for the variance of spring power by utilizing the mechanism to limit the mainspring’s unwinding, ensuring a consistent amount of power is delivered throughout the clock’s running cycle. I have also seen enough of these clocks with the Geneva stops removed. Why? Often, the repairer did not understand how to set them!
The power on a weight-driven clock, on the other hand, is constant, and the loss or gain in time at the beginning of the cycle will be the same at the end, assuming no wear issues are slowing it down.
Slipping, Binding, and Rubbing
If your clock is losing many minutes per day, something in the mechanism is slipping or binding. If your clock is losing minutes per day after all adjustments are made, it is likely that bushing wear or some other worn component, including a weak mainspring, is causing the problem.
Clock repairers often use a term called end shake. End shake allows freedom of lateral movement for each of the wheels between the movement plates. If the plates are tight and there is no end shake, too much resistance will slow a clock. It is why clock repairers always check for sufficient end-shake when servicing the wheels/gears on a movement.
If a wall clock is not mounted correctly, the pendulum might rub against the case. Make sure the clock is not only level and in beat but also properly aligned vertically against the wall. If a clock is not aligned vertically, the pendulum may not swing evenly, causing it to rub against not only the clock case but other parts as well. This uneven motion puts extra stress on a distorted suspension spring.
Final Thoughts
Your situation may be unique, and if it is not covered by this article, I recommend consulting a clock repair expert. If you have little experience and choose to work on an antique or vintage clock yourself, the mistakes you make could be irreversible. For beginners, there are many reputable books and online forums dedicated to clock repair that can provide guidance before attempting any work.
There is also a certain element of risk working with mechanical clocks, as the power contained in the mainsprings may cause serious injury if not handled properly. Always take proper precautions when handling mainsprings or winding mechanisms. If you are unsure or unfamiliar with safe procedures, it’s best to seek assistance from a professional clockmaker to avoid injury and damage to the clock. That said, the safest clocks to work on for beginners are weight-driven ogee clocks or time-only, spring-driven clocks.
Knowing why your clock runs slowly is the first step in diagnosing the problem. Addressing the issue is the next step. Beyond that, periodic maintenance and servicing with quality parts is the key to a long life for your clock.
Back in 2017, I took on the restoration of a Junghans Crispi wall clock—a project that began with enthusiasm but has since tested my patience more than once. While the case restoration went smoothly, the movement has been another story. Despite professional servicing, bushing work, and later my own repair attempts, the strike side has never been entirely reliable.
Junghans Crispi wall clock, circa 1899
Over the years, the clock ran well for a while, then faltered, especially in the strike train. A broken lever spring, a bent arbor, erratic striking, and a few close calls with delicate repairs have all played a part. In late 2024, the clock began stalling just before the warning phase, prompting me to extract the movement from its case to determine the cause of the stoppage.
Now, in Spring 2025, I’ve returned to the movement with a fresh perspective. The time side still runs fine, but the strike side remains unreliable. Based on my experience with similar movements, I suspect a combination of issues: a weak strike mainspring, one or more bent pivots, and poor alignment between the paddle and star wheel.
Junghans movement with the rear plate removed
In this post, I’ll walk through the next phase of troubleshooting—cleaning the movement, inspecting the spring, inspecting pivots, and adjusting the star wheel—to see if this clock, stubborn as it is, can finally be brought into reliable service.
Junghans movement showing hammer and suspension components
The first part of this two-part series can be found here.
Minor issues, But Nothing Too Troubling
After having disassembled the movement, cleaning it thoroughly, and inspecting for pivot and bushing wear as well as the possibility of a broken lever spring and a bent pivot, I am now satisfied that there is nothing mechanically wrong with the movement.
I assembled it and ran the time side for several days. The time train ran smoothly. I then wound the strike side and discovered two issues. One, the stop wheel was not in the correct position. For warning to take place, it should rotate about half a turn, which will set up the strike. The locking pin should be in the 10 to 12 o’clock position. The photo below shows it in the 2 o’clock position.
The locking pin is at the 2 o’clock position when it should be between the 10 and 12 o’clock positions
To make this adjustment, there is no need to take the movement apart; simply loosen the movement nut nearest the fan as well as slightly loosen the nut on the time side. Of course, if in the process of testing the strike side, it is important to double-check that both mainsprings are let down before opening up the plate. Once the fly and the locking wheel are free, they can be positioned correctly. The locking point is also determined by the cam, and the lever must be in its slot while the adjustments are taking place.
The Strike Hammer Tail
The other issue is the strike hammer tail.
Once reassembled, I also discovered that the hammer tail was resting on one of the star points. At rest, the strike paddle must sit between two star points. When the hammer tail rests on a star point, it impedes the action of the strike train. The hammer tail must be clear of the lifting star at the end of each strike sequence.
Hammer tail is resting on one of the stars of the star wheel, the lever spring on the hammer arbour is shown
I believe I have found the problem with the strike side. However, there are two more adjustments that are needed before I put the movement on a test stand. One, the hammer lever spring is not providing any tension. I know that if I risk bending it, it will snap, but I will take that chance. I have repaired lever springs in the past, and it should be an easy fix if I break it. The other issue is a loose click on the strike side, which will require completely disassembling the movement to access the rivet from the inside. Once these two items are attended to, I can test the movement.
It’s still a work in progress, but I feel I’m very close to achieving a successful resolution to issues regarding the movement.
After working on the dial of a Pequegnat Jewel mantel clock, I was very pleased with the results of the repaint. While the detailing may not withstand close scrutiny, the clock looks excellent from a comfortable distance.
When I brought the clock home, it appeared that a previous owner, frustrated by the loss of some or most of the paint on the dial, chose not to source a replacement but instead stripped the paint entirely, leaving only the numerals.
Photo taken on the day the clock was purchased
Someone had gone to great lengths to meticulously scrape away all the paint, but the result is far from appealing.
Something Still Troubled Me
I decided to address the missing paint on the dial by repainting it and touching up the numerals, and the results met my expectations.
After repainting the dial and retouching the numerals
It continued to bother me that the glass surround had a brassy gold appearance, while the dial bezel exhibited a bronze-like tone.
Although brass is an alloy of copper and zinc, with variations in composition potentially explaining this difference, I suspect that’s not the case here.
Instead, I believe the glass surround was originally painted gold, as small flakes of paint came off while I was cleaning it with a toothpick and cloth. Should I leave it as-is or match the outer bezel to the inner one? I’d have to think about it—a two-toned dial looked odd.
A Decision Was Made
Since the dial and surround had already been altered, I decided to take it a step further and repaint the inner bezel as well. With the right shade of acrylic paint on hand, I went ahead and tackled the task.
Some might argue that I went too far with the repainted dial and bezel, believing that any changes made to the clock over time are part of its history. However, I don’t share that perspective. To me, a poor repair from the past justifies taking steps to set things right. Leaving a damaged dial on a clock case that is otherwise in near-perfect condition would be a shame.
From a different angle
I have no regrets about my decision—if anything, I’ve now contributed to its history.
By correcting such mistakes, I believe I honour the intent and skill of the original maker. Restoring a clock to a condition closer to its authentic appearance helps preserve its aesthetic value for future generations.
I am currently working on a mantel clock from the Hamburg American Clock Company. This is Part II in a two-part series. For Part I go here.
The Hamburg American Clock Company is otherwise known as Hamburg Amerikanische Uhrenfabrik or HAU and in many parts of the world, it is simply known as HAC.
HAC mantel clock circa 1926
In Part I, I described the steps taken so far which are disassembly, inspection, cleaning of the parts, and polishing of all pivots. The next steps are bushing work, reassembly, oiling (the mainsprings were oiled in a previous step), and finally, testing.
I believe this is the No. 36 movement by HAC.
HAC clock movement #36
The wheels that spin the fastest typically show the most bushing wear, but wear can also appear in other areas, such as the main wheels, which endure significant torque. For instance, I observed punch marks on the cap side of each mainspring barrel. While punching around any bushing hole is generally poor practice, it appears this was done at some point in the clock’s history and can’t be reversed. Nonetheless, the repair seems to have held up well.
Bushing Work
I have seen similar movements with much more wear than this one, but I felt that some bushing work was necessary to extend the clock’s lifespan.
Bergeon Bushing Machine
Six Bergeon brass bushings were installed in total. For the time side (T), bushings were placed at T1 (front and back) and T3 (back). For the strike side (S), bushings were installed at S3 (back), S2 (front), and S4 (front). The numbers indicate the wheel location in the train beginning from the bottom. A couple of pivot holes were borderline, but I decided to leave them as they are.
Reassembly
It pays to be patient and methodical during reassembly. Rushing can lead to mistakes, such as missing steps or improperly aligning parts, which could cause malfunctions or even damage later. Taking the time to carefully double-check the placement of parts, and ensure everything is in its correct place will help ensure the reassembly is smooth and successful.
An older HAC movement, without the barrel cutouts
Once the wheels are in place the lever and there is only one multi-arm ever that resides between the plates. Position it such that the long arm rests in one of the deep slots of the count wheel while a short arm is placed in the cut out of the cam wheel (as above).
Positioning the pivots is a delicate process, and any forceful handling can result in a bent or broken pivot. In this, as in all cases when aligning pivots, a pivot locator is your best tool.
Pivot locator
My usual practice is to position all the wheels and the lever(s) on the plate with the movement posts, first. Once the wheels are correctly located, I lower the front plate onto the back plate. I secure two nuts to the mainspring end which prevents the movement plates from coming apart as I position the upper wheels in the train.
The pin-wheel, which is uppermost and one wheel removed from the fly on the strike side, is placed in approximately the 12 o’clock position, the warning position. At the same time the hammer paddle is positioned between two of the points on the star wheel. More than once I have had to go back in to make minor adjustments when the paddle hangs up on one of the star points. It looks like this one is also a candidate for that very adjustment.
A Timetrax machine is used to adjust the beat and the rate of the movement.
The movement has been placed on a test stand and the time side Is in beat and running as it should according to my Timetrax machine. Curiously, the Timetrax manual lists only one HAU/HAC movement, a three-train one but this one is likely similar to Junghans bracket clock movements which runs at 164 bpm. Junghans is mentioned for comparison because, by 1930, the company had successfully taken full control of HAC.
Testing
The testing phase is crucial in the process of servicing a clock, often taking the most time because it ensures that the clock is functioning properly before it is reunited with its case.
During this phase, the clock is put through various operational checks to simulate real-time performance, which helps in identifying any issues that may not have been apparent during the initial servicing or cleaning process. Issues such over-tightened or loose components can present themselves as well as problems with the escapement mechanism.
The testing phase serves as an important diagnostic step to identify what further repairs or adjustments may be needed.
This concludes the servicing of an 8-day HAU/HAC movement. If you have any comments or feedback on any of the steps in the process, please feel free to share!
Antique and Vintage Mechanical Clocks 
You must be logged in to post a comment.