Adjusting a 1937 Hamilton 980 “Clark” Wristwatch to Six Positions

      Comments Off on Adjusting a 1937 Hamilton 980 “Clark” Wristwatch to Six Positions

Watch collectors aren’t as obsessive as home chefs or Paleo dieters when it comes to blogging, but we’re close. Of the many watch blogs, one of my favorites is Dan’s Hamilton Chronicles, a blog devoted to Hamilton watches.


One of the many things Dan’s posts cleared up for me is the identity of this little watch. Hamilton made two models—Clark and Yorktowne—that look nearly identical. I could never tell them apart, but apparently only the Clark has a 14K gold-filled case and arrow-shaped hands. So it looks like this one is a Clark, but I’m sure Dan will let me know if I’m wrong.


Nothing says 1930s like a long, curved wristwatch with applied gold numbers. This watch is from 1937, according to Pocket Watch Database. Inside, we have one of my favorite wristwatch movements: a Hamilton 980. This is a tiny movement (14/0 size) with 17 jewels. The 980 has two bigger brothers—the 19-jewel 982 and the 982M, a fancy “Medallion” grade.

The National Association of Watch and Clock Collectors‘ online store sells a cool poster of an early technical drawing of the 980 that they found in their Hamilton archives. I have one hanging over my watchmaking table. The 980 I’m working on, needless to say, is somewhat smaller.

Cleaning and Servicing

This watch came from eBay, the wellspring of all grimy watches. It wasn’t running at the time, and it was nasty looking—just my kind of watch.

Disassembly, cleaning, and servicing found no gremlins or bugaboos. It got a new white alloy mainspring, a trip to the demagnetizer, and an overnight rest for the oils and mainspring to settle in. The initial amplitude was huge, a common thing for these 980s—they clean up nicely.

Adjusting Mr. Clark

What kind of accuracy can we expect from this watch, and what standards should we use? On the plus side:

  • The 980/982 movements are high quality and have great potential.

On the minus side:

  • The 980 lacks some features that make precision adjusting easier. It lacks adjustable mean-time screws on the balance rim for adjusting the rate, so making big changes is harder. And it has a plain regulator index instead of the micrometer screws found in pocket watches, so making tiny changes is harder.
  • The rates are slightly unstable. Over a 60-second period, the rates would vary by around 6 to 10 seconds.

Our post on unstable rates and wavy traces describes some common causes of short-term variability in a watch’s rates. The variability seemed to have a 60-second period, which would suggest something associated with the 4th wheel that carries the seconds hand.

Whatever the cause, I couldn’t crack it. I examined the likely culprits, re-cleaned and oiled, and evaluated the rates with a few different 4th wheels I had handy. No luck. Some old watches are just a bit wavy. (Or some adjusters are just dense and missing something—I shouldn’t blame the watch.)

Stable rates are a prerequisite for ultra-fine adjusting. As a result, we’ll use looser standards for this watch. I’d be happy to get all the rates within +/- 10 seconds from zero.

Adjusting to 6 Positions

Hamilton originally adjusted this watch to 3 positions at the factory: for a wristwatch, dial up (DU), dial down (DD), and pendant (PD) are the big 3. We may as well aim for 6.

I centered the regulator index, fully wound the watch, and took the baseline rates. This watch is speedy. The DU and DD rates are close enough, and that’s a relief: aligning the DU and DD positions is usually most of the hassle in adjusting to positions. Otherwise, the 4 vertical positions are way out there.


If you haven’t yet, you might read the “Dynamic Poising How-To” series, which explains the nuts-and-bolts of positional adjusting using a timing machine. In short, we need to find the heavy spot on the balance wheel, and it is easy to find when the watch’s amplitude is low.

I let down the mainspring and measured the rates at an amplitude of around 160 degrees. Here’s what we have:


Notice that we didn’t measure two positions (shown with a period) because you can measure the main four, find the fastest one, and then measure the two flanking positions to see if one of them is even faster.

The watch ran fastest in one of the diagonal positions.


As we’ve said before and will say again, the heavy spot is the spot directly below the balance staff in the watch’s fastest position. That means that the heavy spot is here, at the red spot. The light spot, naturally, is exactly opposite it, at the green spot.


Because this watch is running fast, we will add weight. Adding weight moves the DU and DD rates closer to zero, and it makes the vertical positions more similar. (If we removed weight, the DU/DD and vertical rates would grow apart.) Remember: if fast DU/DD, add; if slow, remove.

I’m always happy for a chance to commune with my inordinately large stash of timing washers. We’ll use this set, which has small washers for tiny watches like this one.


The watch’s poise seems way out there, so I’ll start with a bigger one: a washer from bottle #10. What was the result? Here are the new rates when the watch is fully wound.


A big difference, but not big enough. We’re off once to again to identify the heavy spot by running the watch at low amplitude. I suspect that the heavy spot hasn’t changed—I should have used a heavier washer—and I was right.

I put a timing washer under a different screw this time, one next to the screw weighted in the prior round. I opted for a heavier one (bottle #7 in the picture), and that certainly helped. The watch is still fast in all positions, but all 6 are converging ever closer together.


I suspect we’ll need only one more round involving a very light washer. Once again, we measure the rates at low amplitude, and this time the heavy spot has shifted to pendant up (PU), which is this position.


Because the watch is still a bit fast, I’ll add weight again. And because the rates are fairly close, only a small amount of weight seems called for.

I place the lightest timing washer (from bottle #1) under the screw at the light spot, and everything looked good. The watch was around 15 seconds fast DU—with a slight nudge of the regulator, the final times were solid.


In short, this watch needed only 3 rounds of dynamic poising. Each time, we added a washer, so this is an example of adjusting without removing weight.

This is pretty good, but keep in mind that the rates are worse than they look. Because they are unstable, each number is an average with +/- 3 or 4 seconds attached to it. If this were an Illinois or Hamilton pocket watch, I’d try another round or two, but these rates will do for this slightly wavy watch.

Cleaning the Case & Wrapping It Up

wp-1464527186617.jpgWhat about the rest of the watch? There’s surely more to restoration than fiddling around with timing washers.

The case, as one would expect, was crusty. Cleaning gold-filled cases is easier than it looks. If the case is especially gunky, you can soak it in denatured alcohol and scrub it with a toothbrush.

If pure alcohol isn’t your style, a toothbrush with hand soap works well, too. You can remove the old glue from the bezel with toothpicks.

After scrubbing off the worst of the grime, I cleaned the case and crown in the cleaning machine along with the watch movement. The cleaning solutions work wonders, but you can get most of the benefit with inexpensive alcohol or a soapy toothbrush.

After a soak and scrubbing, the case parts looked much better. Oddly, the case back has a dotted residue. The back didn’t fit snugly, so perhaps someone taped it together at one point.


And then the case paid a visit to Mr. Buffer. For gold-plated and gold-filled cases, you should avoid polishing compounds that will “cut” into the thin layer of gold on the case. We’re looking only to give it a clean, consistent, shiny finish. I like to use the Menzerna Yellow P175 compound. It gives a striking luster to a gold case. The base layer is less noticeable if it has the same glossiness.

Polishing makes a huge difference, but there’s no getting around the case’s wear.


wp-1464527311362.jpgThe dial was in terrible shape. I like weathered and aged dials, but some simply aren’t legible or salvageable. This one had lost nearly all the text and finish, a couple numbers were missing, mildew was taking over, and the dial had some dents and chips.

I sent it off to be repaired and refinished by International Dial, and they did a nice job. (You might notice that the seconds track is incomplete. The dial had been seriously damaged there, perhaps from hand removing gone seriously awry back in the past, and Int’l Dial did a great job with what they had to work with.) It looks sharp.

wp-1464527133911.jpgAnd, of course, the watch needs a new crystal. It came with a glass crystal that looked like the wearer had been dragged through a cactus field, so I replaced it with a new acrylic crystal.

I confess that I prefer the look of plastic crystals on these old watches—the greater height makes them visually striking.

And the End

This was an unusually straightforward project. There were no unpleasant surprises during servicing and adjusting, and the watch now runs pretty well.

Wavy traces are vexing but common in old watches. When rates show near-term instability, it is futile to chase after extreme accuracy. You can achieve incredible positional timing on almost any good watch that has straight, parallel, stable timing traces, as our best performing example watches show (e.g., the Hamilton 912, Illinois 405, Bunn Special, and Illinois Burlington). But when the rates vary, it’s best to simply get the rates as tight as you can so that the watch serves its wearer well when worn.

And I’m looking forward to wearing this once I get a band. In a cruel twist of fate, this narrow, curved watch takes a tiny band: 14 mm in width at the lugs. Sounds like a trip to eBay is in order.