ClockInfo.com

I just completed the overhaul of an Ansonia iron cased mantel clock (my job no. 4422). It has the common 4 pillar 8 day time and strike movement of dimensions 5 x 3-1/4 inches. In this clock, a previous repairer had replaced the srike mainspring with one that is 3/4 inches wide and .018 inches thick. The time spring was an original that is .0153 inches thick.

When assembling the movement, I put the original thin mainspring in the striking side, and for the time mainspring used a spring 3/4 inches wide x .0138 inches thick by 120 inches long (no. 77.303 from R & M Imports). This spring gives a good pendulum motion of just over twice the escape arc. It will run the clock reliably without causing excessive wear.

I just overhauled an Ansonia oak shelf (kitchen) clock (my job no. 4357). It has the 4 pillar 8 day time and strike movement of dimensions 5 x 3-1/4 inches. The original mainsprings in these clocks are 3/4 inch wide and typically about .0155 inch thick. The previous repairer had installed a time mainspring .0185 inch thick! The loop end of the spring is labeled “USIBEL FRANCE”. I replaced this spring with a new spring that is 3/4 inch wide, .014 inch thick and 108 inches long. The pendulum motion is great (running arc slightly more than twice the escape arc).

I have seen (and discarded) these thick USIBEL FRANCE mainsprings many times before. They were sold by clock parts suppliers such as S. LaRose as “mainsprings for 8 day American Clocks” and sometimes described as the finest mainsprings, made in France. I guess they were fine for many repairers, as they are so strong that they will make a clock work without being correctly repaired! I have measured some of these springs to be as thick as .019 inches! Just by winding the clock, you can feel that they are way too strong. If used for too long, they will cause severe wear to the mainwheel teeth.

The force that a mainspring provides is proportional to the cube of the thickness: .0185 cubed divided by .0155 cubed is 1.7, so the USIBEL spring is providing 1.7 times the force of the original spring (actually it is even worse, as modern steel is better than steel made 100 years ago. So the new .0185 inch thick spring is probably at least twice as strong as necessary.

I used a no. 77.303 from R & M Imports that is specified as 3/4 x .017 x 120 inches. The springs I received have an actual thickness slightly less than .014 inch, and I shortened a spring to 108 inches and re-attached the loop end. This spring seemed so weak when uncoiled that I wondered if it would have enough power! But it works great for the time train, as mentioned in the first paragraph. NOTE TO MYSELF: Next time, leave the spring 120 inches long - there appears to be plenty of room.

Note: R & M has a no. 77.301 mainspring that is specified as 3/4 x .014 x 108 inches that may be German made, and seems stronger than the spring discussed above. I like the 77.303 the best.

It is interesting to compare the force provided by the USIBEL spring to the force provided by the .014 inch thick spring: .0185 cubed divided by .014 cubed is 2.3. This mean that the USIBEL spring provides approximately 2.3 times the force of the .014 inch thick spring. A clock running with the thick spring will wear itself out prematurely!

If you have an heirloom antique American clock that you want to pass down to future generations, make sure your clock repairer does not (or did not) install mainsprings that are too thick. Original mainsprings should be kept in the clock, unless there is a good reason for replacement (broken, too strong, rusted or damaged).

Merritt’s Antiques has the P1496 mainspring in the red, orange and yellow box again! Last week I received 12 of them. I took two springs, measured them to be .0158 inches thick, cleaned and lubricated them, and installed them in a Seth Thomas 89C movement in an Adamantine mantel clock (my job no. 4396). These springs are perfect for this type of clock. The pendulum motion is good but not excessive.

After a clock movement has been repaired, it should look like it has always been well taken care of, and not show obvious signs that it has been “repaired”. As part of this, the cleaning process should not be harsh. For typical American antique clocks, I use “Historic Timekeepers” cleaning fluid (available from Timesavers as part #17863 and #17864). I scrub brass parts such as plates and gears with a fine brass bristle brush (stock #16.310 from R & M Imports). This results in a uniformly clean finish that is not too shiny (some repairers polish the brass parts with metal polish, but this leaves the parts looking too shiny, and American clock parts never looked this way when new. It is customary to polish the back of the back plate of a round French movement, and Vienna regulator movements look great when polished).

Some repairers use an electric motor powered rotary brass brush to polish the brass parts. I DO NOT like the way this makes the movement look, and so I am opposed to this practice.

This note concerns a Revere A/C electric mantel clock with 60 RPM Telechron motor (yes, 60 RPM or 1 revolution per second) made in the late 1920’s. The movement is 5-3/8 inches tall and 5 inches wide and is labeled on the back:

1334428
1615664
PATENTS PENDING

REVERE CLOCK CO.
CINCINNATI O.

This movement relies on the second hand to reduce the endshake of the hour wheel. If the second hand is pushed all the way in, the rack tail will rest on the snail when the rack is released. With the second hand off, the hour wheel may move forward, and the rack tail can slide behind the snail, particularly at 1:00 and 2:00. So if you are testing the clock with the second hand off and hear it striking too many times, don’t be alarmed! Put the second hand on and all should be well.

I repaired a round French movement made ca. 1790 by Roque, Paris. It is a round time and strike movement about 3-1/2 inches in diameter. It is more delicate and looks more hand made than the typical round mass produced French movement from the mid-1800’s. After repair, the warning wheel kept getting out of synchronization with the locking wheel. Once or twice per day, the warning pinion would skip a tooth on the locking wheel, making the warning pin be in a different position when the strike train is as rest. After a while, the warning pin would end up adjacent to the warning lever, and the strike would fail to start. I re-synchronized the warning wheel, only to have the same thing happen again. The warning wheel was NOT loose on the hub, and the hub was tight on the arbor.

The arbors are very thin and delicate, and the fan fly tension seemed more than necessary. We thought that perhaps when the locking wheel is suddenly locked, the momentum of the fan fly caused enough flexing in the the locking or warning wheel arbor for a tooth to be skipped. My apprentice, John, suggested rebushing the pinion end of the warning wheel, as it was slightly loose.

First, I did the easy thing and reduced the tension of the fan fly tension spring, and the problem did not occur again until the 6th day (it had been occurring about every day). I looked closer and realized that the depthing between the locking wheel and warning pinion was very shallow, so I disassembled the movement and installed a bushing, at the same time moving the hole closer to the locking wheel.

I just overhauled a Seth Thomas 89C movement for an Adamantine mantel clock (my job no.4401). Both the time and strike mainwheels had significant tooth wear. The mainsprings had the ST logo and may be original, and the thicknesses were .071 for the time spring and .073 for the strike spring. To reduce future wear, I replaced both spring with the Merritt’s P1496 mainsprings, measuring .0155 inches thick. They provide noticeably less force, yet are plenty strong enough to run the clock, provided that the movement has the pivots polished and necessary bushings installed. See this post for information on these mainsprings.

Note 9-11-07: The strike train must be almost perfect to run reliably with this thin of a mainspring. My customer brought the clock back into the about shop two weeks after he had picked it up, saying that it did not always strike. I first established that the mainspring had enough torque by letting the spring down 10 turns from fully wound up (simulating a running time of 10 days).  The striking still ran. I wound it back up, and found that occasionally the strike would not go into warning, and that it was always when gears S3 and S4 were in a specific relationship. I let the mainsprings down, and removed gears S3 and S4. S3 had significant wear on the tooth tips, and there was slight wear on the pinion wires of S4. (With a stronger mainspring, this would not have caused any problem). I reversed the wires in S4 so that S3 would engage with the new portions. Then I removed gear S3 from its hub, turned it over and re-installed it, so that the unused side of the teeth would be employed.

I just repaired a Seth Thomas “Queen Anne” wall clock with time and strike movement. It is a large movement (about 7 1/2 inches tall), much larger and heavier than the standard American movement. In Tran Duy Ly’s book “Seth Thomas Clocks and Movements”, the movement is listed as #85, and the mainsprings are specified as 3/4 wide by 11 1/2 feet long by .018″ thick. On this clock, a previous repairer had messed up the escape wheel teeth. I trued them using the Webster escape wheel tooth straightener, then topped the teeth slightly in the lathe to bring them all to the same length, then carefully removed the burrs. I tested the movement, and the .018″ mainspring seemed to have WAY more force than needed for the time train. I tested a long thin mainspring intended for 31 day clocks: 3/4 x .015″ x 170″, and it had plenty of power for the time train, but was too weak for the striking (it could barely run the strike train until the mainspring was run down 8 turns, but striking was very slow). Thus, I kept the .018″ mainspring for the striking.

After pivot polishing and bushing, I cleaned and assembled the movement. The escape arc is .62 degrees according to the beat scale, and the running arc (not even fully wound up) is 2.5 degrees. Thus, this thin mainspring has plenty of power for the time train. The strike speed sounds normal with the 0.018″ thick mainspring.

Summary:

Seth Thomas Queen Anne Time and Strike Model:

Time mainspring: 3/4 x 0.015″ x 170″ (length is not critical)

Strike mainspring: 3/4 x .018″ x 130″ (length is not critical)