The movement has the nice thin original Seth Thomas mainsprings.

Time Mainspring: 11/16 inch wide by 0.0165 inch thick.

Strike Mainspring: 11/16 inch wide by 0.0154 inch thick.

See more photos.

Repair job 5539. I polished the pivots, rebuilt the pinion on the third strike gear, and installed 10 bushings. The movement might not be original to this case (the mounting feet are bent way back in an attempt to move the movement forward in the case).

Most clock parts distributors offer a “standard” American loop end 8-day mainspring 3/4 inch wide by 0.018 inch thick and 96 inches long. These springs are too strong! They may cause severe wear to the clock’s mainwheel teeth. A thinner mainspring about 0.0165 inch thick should be used instead. (Regarding mainspring replacement, I believe that the original mainsprings should be kept in an American antique clock unless a spring is broken, damaged or too strong. Notice that I said “too strong” – the old mainsprings are almost never too weak!)

Several clock parts suppliers offer the thinner mainsprings for American clocks, typically 3/4 x 0.0165 x 96 inches. Clock repairers have noted inconsistencies in these springs – the thickness and strength can vary from batch to batch. I recently bought springs from 3 suppliers to test. For each spring, I measured its thickness, and its diameter when first removed from its retainer. On some springs I also measured the diameter after being fully wound and unwound 2 times.

The measurements are given below, with the mainsprings placed in categories based on my recommendations. At times, a supplier will ship mainsprings from a different source. The mainspring types pictured below show the appearance of the springs that I measured.

Recommended Standard Length Springs – 96 Inches Long

Timesavers 18790, Specified as 3/4 x 0.0165 x 96 Inches

Sample 1) 0.0165 inch thick, opens to 10″ (this spring has some rust on the outside and will be exchanged for another.)

Sample 2) 0.0162 inch thick, opens to 11.5″

R & M Imports 77.308, Specified as 3/4 x 0.0165 x 96 Inches

Sample 1) 0.0160″, opens to 11″, 9.5″

Sample 2) 0.0160″, opens to 11.5″

Sample 3) 0.0160″, opens to 11″

Sample 4) 0.0158″, opens to 13″

I have been using this spring from R & M for years. Typical batches from 4 – 5 years ago measured 0.0165 inch thick. There was a batch about 2 – 3 years ago that had brittle inner ends that broke easily. Current springs seem fine in this regard.

R & M Imports 83065, Specified as 3/4 x 0.0165 x 96 Inches

Sample 1) 0.0160″, opens to 11″, 9.5″

This sample is identical to the above spring (R & M 77.308) and costs over twice as much, so there is no reason to buy it!

Merritt’s Antiques P-1496, Specified as 3/4 x 0.0165 x 96 Inches

New springs purchased May 2010:

Sample 1) 0.0159″, opens to 12″, 11″

Sample 2) 0.0161″, opens to 12″, 10″

I first used this mainspring 4 – 5 or more years ago. Some batches have been significantly thinner than others. A batch that that I purchased in October 2008 is packaged the same as this current batch. The springs  have the following measurements:

Sample 1) 0.0165″, opens to 11″

Sample 2) 0.0165″, opens to 11″

Sample 3) 0.0158″, opens to 11″

Sample 4) 0.0165″, opens to 9″, 8″

Sample 5) 0.0165″, opens to 11″

Sample 6) 0.0156″, opens to 11″

An older batch with plastic inner wrap had thinner springs:

Sample 1) 0.0157″, opens to 11″

Sample 2) 0.0155″, opens to 11″

Another old batch (probably 4 – 5 years old) marked “Anchor” on the plastic inner wrap contained mainsprings that were very thin:

Sample 1) 0.0135″, opens to 10″

Sample 2) 0.0140″, opens to 12″

Around June 2007, Merritt’s was shipping a different mainspring under this part number. It was the same as the R & M 77.308 and the Timesavers 18790.

Recommended Long Springs

Timesavers 15959, Specified as 3/4 x 0.0165 x 120 Inches

Sample 1) 0.0160″, opens to 12″

Sample 2) 0.0168″, opens to 13″

120 inch long mainsprings are sometimes recommended for more uniform timekeeping over a one week running period. I have had excellent results with the 96 inch long springs, however, and recommend them for most movements.

Springs That Are Very Strong

Merritt’s Antiques MS309, Specified as 3/4 x 0.0165 x 96 Inches

Sample 1) 0.0158″, opens to 15″, 13″

This is the type of spring made in Germany that is very springy. The package has a yellow label saying “Beco Technic Germany”. They are tempered to a straw color instead of blue like many other springs. I don’t like to use these in average American antique clocks, as they provide more power than necessary.

Merritt’s Antiques MS310, Specified as 3/4 x 0.0165 x 108 Inches

Sample 1) 0.0152″, opens to 21″, 18″

This spring is by Beco Technic Germany just as the above spring. It is intended for the Seth Thomas No. 89 movement. It may be stronger than necessary, and I prefer to use springs in the first or second categories above instead.

Springs That Are Very Weak

Merritt’s Antiques P-1956, Specified as 3/4 x 0.0165 x 96 Inches

Sample 1) 0.0165″, opens to 6″

Sample 2) 0.0165″, opens to 6″

Springs in this batch open out to only 6 inches when released from their clamp. They are not very springy and are too weak for most American antique clocks.

The springs in my first batch two years ago were excellent! They opened out to 11 – 12 inches when released from their clamps, and provided plenty of power. I used these springs in several clocks. But batches received last fall and later have been very weak. So I don’t routinely use these springs anymore.

Other Springs

Timesavers 29515, Specified as 3/4 x 0.016 x 108 Inches

Sample 1) 0.0153″, opens to 28″, 20.5″

This spring seems way too powerful. Also, the surface is not smooth and the spring has a rough action.

Conclusions and Recommendations

For American antique clocks, I keep original mainsprings in the clock unless there is a good reason to change them. For replacements, I prefer springs from the first category (Recommended Standard Length Springs – 96 Inches Long). Many typical Seth Thomas, Ingraham, Gilbert, Sessions and New Haven clocks can use these springs.

Ingraham mantel clocks with deadbeat (or almost-deadbeat) escapement can use a somewhat thinner spring for the time side of the movement.

Some Waterbury and Ansonia movements made from the 1880′s through the 1920′s originally had mainsprings 0.0145″ to 0.016″ thick. It is important to select the proper thinner spring for these clocks.

It is best to measure and label new mainsprings when you receive them. After a time you can build up a stock of normal, thicker and thinner springs, so you can choose the proper mainspring when a mainspring does need replacing.

I believe that routine mainspring replacement should be avoided in antique American clocks.

Thanks to Len Lataille for encouragement to measure these mainsprings.

What caused the spring to break into multiple pieces?

The clock had been cleaned without being dismantled (there was debris in the gears from incomplete cleaning). The cleaning fluid could not be rinsed out of the mainspring. The fluid attacked the steel, causing stress cracks to form, followed by breakage.

Repair job 5176. Back plate number 1278, no date. Uses 0.004 inch suspension spring. When the mainspring broke, it bent the front second wheel pivot. I inserted a new pivot, and polished several other pivots.

My newest batch of these mainsprings (received late in 2009) are weaker than the first batch I had. When removed from their clamp, they open out to only 5 1/2 inches (my first batch opened to about 11 inches).

These springs may be too weak for clocks with a recoil escapement. I still need to test them in a Seth Thomas No. 89 movement.

This particular example has thin, original mainsprings.

Time Mainspring: 3/4 x 0.0167 inch.

Strike mainspring: 3/4 x 0.0163 inch. (Unfortunately, I had to replace the strike mainspring because it had a rough action when nearly wound, typical of springs that may break soon. I used a new Merritt’s Antiques P-1956).

Many examples of the No. 89 movement have thicker springs, quite often around 0.0175 inch thick, sometimes thicker. After proper repair, this movement will run well with thin springs.

Here is a video of the escapement motion, first run down 6 1/2 days, then fully wound:

Repair Job. 4970.

The biggest weakness of this movement is that the mainsprings are attached to their barrels by tongues cut out of the barrel wall itself. When I received the movement for repair, the time barrel tongue was broken and had been unsuccessfully repaired. The tongue on the strike mainspring barrel broke as I was winding the spring in after cleaning

I made new steel mainspring hooks, drilled a hole in each barrel, and riveted them in. The slideshow below shows the previous repair on the time mainspring and barrel, one new hook I made and installed, and general views of the movement and dial.

The strike second gear had a gear tooth that about to break off. I inserted a new tooth. The pivots needed polishing, and I also installed several bushings in worn pivot holes.

Repair job 4961.

I want to thank Clockmakers Newsletter for their article on repairing barrel hooks.

W. Barnes Patented Nov. 27, 1866

The NAWCC Library & Research Center was able to give me the number of this patent: 59,943, issued to Wallace Barnes. Then I was able to find the patent on Google Patents.

Here are the mainspring ends:

The patent describes how the mainspring can be hardened, tempered and blued, then the loop end attached. The patent implies that this is the first use of a separate piece for the loop end, riveted on. Previously, the end of the spring was rolled to form the end.

Here is the text of the patent

IMPROVED METHOD OF TEMPERING CLOCK SPRINGS,
WALLACE BARNES, OF BRISTOL, CONNECTICUT.
Letters Patent No. 59,943, dated November 27, 1866.

SPECIFICATIONS

TO ALL WHOM IT MAY CONCERN:

Be it known that I, Wallace Barnes, of Bristol, county of Hartford, and State of Connecticut, have invented certain new and useful improvements in the mode or process of manufacturing Clock Springs; and I do hereby declare that the same is described and represented in the following specification and drawings, so as to enable others skilled in the art to produce or make the same therefrom.

The nature of this improvement will be understood from the specification and drawings.

The object desired to be attained thereby, is to produce a spririg from a poorer quality of metal, and, at the same time, render a spring of given dimensions more effective for use, and more merchantable as an article of trade. And to effect this object, the springs are first hardened in the common way. Then, after the hardening process, they are placed between metal plates having their surfaces (one or both) corrugated or perforated, so as to allow the tempering fluid to flow freely between the coils of the spring and the plates, and immersed in molten lead or its equivalent, for the purpose of drawing the temper, or, in other words, for the purpose of producing the required temper to the spring, and also to produce a true, even position of the edge of the spring relative to all parts of the coil. The advantage of this operation, or result thereof, is to produce a greater amount of effective and prolonged action of a spring of equal dimensions, and to avoid the otherwise great amount of friction or obstruction to the mechanism of a clock, &c., when compared with those now in use. Then the spring is polished and blued in the ordinary way. After the above is accomplished, a clasp, a, having a sleeve formation, c, is rivetted to the outer end of the spring, d, instead of coiling the end thereof, as in the old way. The advantage gained by the use of this process will be apparent. The spring can be finished complete its entire length, hardened, tempered, polished, and blued, before the clasp is attached thereto. The clasp may be made of any desirable metal, and secured on each side of the end of the spring by rivets between the two parts of the clasp, thus protecting the end of the spring where it is liable to break or give way, and thereby provide a more perfect sleeve, by mean’s of which it (the spring) is held more perfectly in its desired place. It also produces a more merchantable article. I believe I have thus shown the nature and process whereby I produce this improvement.

What I claim, therefore, and desire to secure by Letters Patent, is—

The mode or process in the manufacture of springs for clocks, &c., substantially as described.

WALLACE BARNES. [L. S.] ,

Witnesses:
E. M. BLiss,
Jeremy W. Bliss.

This beautiful mahogany Adamantine clock came into my shop recently with a broken time mainspring. Nothing too exciting about that, but I think this clock is so nice looking that I decided to show it here. I had previously repaired it 11 years ago, overhauling the movement and polishing the case and bezel.

I replaced the broken time mainspring with a Merritt’s P-1956 3/4 by .0165 by 96 inches (the original spring was 0.0172 inch thick). The pendulum takes a superb motion (almost too good) and so an even thinner mainspring would work. During my previous overhaul, I replaced the way too strong strike mainspring (an incorrect replacement) with a spring about 0.016 inch thick.

Repair job 4981.

The clock had its original mainsprings, both of which were marked “W. Barnes Patented Nov. 27, 1866″. The strike spring was quite thick (0.02 inch). Both springs were stiff and very hard to remove from the mainwheels. I installed new Merritt’s P-1956 mainsprings that are 3/4 by .0165 by 96 inches. They provide a good escapement action and striking speed. Gilbert clocks of this type take a large pendulum swing with a small amount of supplementary arc (overswing) due to the design of the recoil escapement. I slanted the back of the mainspring arbor hooks to make the mainsprings easier to remove for future cleaning.

The video below shows the escapement action with clock fully wound and then run down 8 1/2 days, then shows the movement and then the entire clock.

Repair job 4957.

The mainsprings appear to be original (they match in color and style and look old).

Time mainspring: 3/4 inch wide by 0.0172 inch thick. This spring provides sufficient power, as the escapement motion is excellent.

Strike mainspring: 3/4 inch wide by 0.0142 inches thick. This spring is very thin, yet the striking speed is fine.

I retained both original mainsprings in this clock. They were in good condition, operated smoothly, and provided enough operating power, so there was no reason to change them. A good old mainspring is no more likely to break than a new one. (Some repair shops install new mainsprings in every clock they repair, I disagree with this practice.)

The case is 33 inches tall, and the minute track diameter is 11 inches.

Repair job 4955.