Job Number 180070

This one is a grade 10 Elgin, 18 size, 11 jewels, made about 1893. 

This lever is part of the setting mechanism. On the left of the part is a post that carries a gear underneath. The post is threaded in, but seems to get striped somehow. Sometimes they are loose, and sometimes missing.

This one must not have been staying in place as some prior repairer has places a series of divots with a punch, all around the part.

See more "creative repairs" here.

The serial number prefix, used on stamps on all the major parts, is a small 't'.

New Arrival

New York Standard, 5th model, 18 size, 7 jewels...

Job Number 180068

This one is an American Waltham "No. 620", 1908 model, 16 size, 15 jewels, made about 1913.

The teens into the '20s are sort of a golden age of American watches in my opinion. The technologies used in their manufacture had developed nicely. Everything works, everything fits. And yet there is still a great deal of decorative finish. This era produced many fine watches, at reasonable prices for consumers.

Excellent workmanship, everything is in great shape, almost no tool marks.

There is a small issue with winding/setting. In setting mode the shipper spring that pushes the clutch (upper left) toward the middle to engage setting is a little weak with age (old steel). The clutch has a tendency to "skip" against the idler gear (under the round cap), that in turn moves the minute wheel, which in turn forces the cannon pinion (center) to turn, slipping on the center shaft. The minute hand rides on the cannon pinion, so this is how the watch sets.

The first thing one might do about the skipping is to loosen the cannon pinion so the watch sets easier. In this case though the cannon pinion is in excellent shape. It's grip on the center shaft is perfect. I don't want to ruin that. If a cannon pinion is too loose it will slip all the time, when the watch is running, and the hands will not move even as the watch ticks away.

In the case, it turns out, the issue is greatly reduced. The stem depth has been previously adjusted just right so the arbor maximizes the push of the clutch in winding mode. I'm just going to leave well enough alone.

Another option would be to re-tension/harden the shipper spring with heat. It's a judgement call, but my general philosophy with these antique watches is to do as little as possible. Leave no footprints, as it were.

New Arrival

You see a lot of hunter movements in open face watch cases. With the stem usually at 3:00 people sometimes call these side-winders.

This watch is an example of something less common. It's an open face movement in a hunter case. This combination would generally end up with the stem at 12:00.

Job Number 180066

Here's a grade 317 Elgin, 18 size, 15 jewels, made about 1910. 

Oddly, this watch was missing the little pusher pin that goes inside the winding arbor. It's not a hard part to replace, but it is a strange part to be missing.

On closer examination, I noticed a lot of wear marks on the vibrating arm that the post pushes (into winding mode). The vibrating arm likely wasn't as in the way as normal and the post may have simply pushed out and fallen away at some point.

Since the watch did not run (broken roller jewel) and was fully would already, one would not have noticed that it didn't go into winding mode properly. Just a guess...

Here's the balance assembly with the broken roller jewel, sometimes called a roller pin. I was slowed down a little cleaning this in preparation for shellacing a new jewel. It turns out a prior reparer had used glue to fix the roller jewel. Unexpected... Some acetone took it right off though.

There's also a lot of weight on this balance. Someone really wanted to slow it down. I've seen worse.

1) A lot of weight on the balance.
2) Regulator all the way to slow.

Sure enough, on the machine the watch still ran fast by 2-3 minutes per 24 hours. You can see in this animation though that the balance was pretty out of true. The beat is way off too, although you can't really see that, I could sure hear it.

Improving the balance and beat also improved the rate significantly. I didn't take any weight off, but the regulator is now 3/4s way over to the fast side. We'll see how it does. The hairspring is good, no damage, but the age of the steel is likely a significant factor in the rate.

This is one of those watches with a lot of wear and tear all throughout. It should work pretty well but I won't be looking for great accuracy. It's very best days are behind it.

Job Number 180054, Hands Replaced

The watch owner sent two replacement hands that they preferred to have on this watch so I swapped them out today. Looks good! These are an uncommon style.

Job Number 180064, Testing

An inlay dial, fancy gilt hands and a hunter case complete the picture.

I had a hard time finding a replacement case screw for this one. The case is an oddball in several ways. The case screws are extra long and thick, but with a small head.

Also the bow is held with a long screw going into the neck of the case. I have never seen that before.

Job Number 180064

This one started out on an auspicious note, with a broken off case screws. The watch had a broken balance staff, so I suspect the impact that broke the staff also popped the screw head off.

Fortunately the body of the screw came out easily.

Many older Elgin watches have nice details that in later years were dropped from entry and midrange products, for example finish work. This watch has polish and grain added to even a simple part like the click.

This grade is one of those with a male winding arbor. The post goes into a square hole in the stem of the case. Cases like this are extremely rare.

This movement is Elgin's grade 4, 16 size, 15 jewels, made about 1884.

It's hard to see but there are watchmaker's marks on the base plate. These are usually found on the case, especially inside the back. But I have seen them a few times on parts of the watch.

These marks were made by people that serviced the watch. They have no standardized meanings, although these are pretty obviously a date and initials.

The watch needed a balance staff. It's not unusual to have to alter a replacement staff slightly to work, even with absolutely "correct" factory parts. It's usually the pivots though. On this one the lower body of the staff was way too large for the roller table - way too large. I had to cut it down quite a bit. I'm not sure what the deal was with that, but it worked fine in the end.

The old staff, cut off, and the replacement...

Installing the roller table...

Ticking...

A Man of Time

Job Number 180055 - Mystery Gain

This watch went together really well. On the machine, it rates near flat line and almost no beat error. I had high hopes. But in practical tests, it reads really fast. The watch gains, reliably, 22 minutes per 24 hours.

What is the problem?

Watches are machines, they behave in completely deterministic ways. The key elements of a watch are a power source (mainspring), the gear train, an governor mechanism (escapement and balance), and an output (the hands). The gears are not arbitrary, they have to have specific ratios to each other in motion. For example, there are 60 seconds in a minute. So the 4th wheel, that carries the seconds hand, has to go around 60 times exactly for each single revolution of the center wheel, which carries the minute hand.

The exact speed of these wheels, what we think of as the watch's accuracy, is a function of the escapement. But the wheels must turn at the appropriate rate relative to each other for minutes and seconds to be even meaningful. Gear ratios dictate this relationship.

Sometimes, when a watch is running fine, but reading incorrectly, the degree of error can tell you exactly where the problem is. You don't even have to see the watch.

This watch gained 22 minutes per 24 hours. There are 64 teeth on the center wheel (typical of American pocketwatches). The passing of each center wheel tooth against the pinion of its neighbor, Mr. 3rd wheel, represents 1/64th of a minute. 1/64th of a minute error per hour (each turn of the center wheel) adds up to 22.5 minutes error in 24 hours.

Inspection of the center wheel reveals that I missed a slightly bent tooth. Each time this part of the wheel came around, it was not enough to stop the watch, but it did pop over to the next tooth, skipping exactly one, per hour. The problem is visible in the above image.

Fortunately, the damage to the wheel was not severe. I was able to nudge the tooth careful back to its correct position, so the center wheel would turn smoothly.

The lesson is not to get lazy about checking each wheel, especially on a watch like this that did have damage to the train to begin with.

Job Number 180062

Now to test the rate for a few days... It looks pretty good on the machine anyway. We'll see.

Job Number 180062

This one is a Waltham 1899 model, 16 size, 17 jewels, made about 1907.

Here's that classic Waltham motor barrel.

And the Waltham shipper, with all the associated keyless works components...

Ticking nicely...