1931 Perry

Life is fully of ironies and watch collecting is no different.  My favorite models tend to be from the 1930s but they are also my least favorite to work on.

Well, that's not entirely true.  My absolute least favorite are the silver-colored microrotor movements from the TM-line... the 628 and it's kin.  They have a knack for loose cannon pinions where the movement runs fine but the hands don't keep up.  Parts are nonexistent.  I don't even bother working on them anymore.

But, back to my original thought, I hate that I love the 1930s models.  They're the best the world had to offer at the time but, like cars or planes, watches from this era are much less robust than watches made 15 to 20 years later.  Technology and materials improved quickly over the 1930s and 40s and, compounded by decades of use, early Hamilton watches can be extremely temperamental.

Don't get me wrong... they're excellent time pieces and when properly maintained they are stunning.  It's just that use and abuse or even one false step can make them untrustworthy time pieces.  

I always think about how many old black and white movies from the day had a scene where someone asks, "Do you have the time? My watch has stopped".  In the 1930s, there was a watchmaker in every jewelry store.

My latest project is a good example of this paradox - its a beautiful 1931 Perry.  The 1931 Perry was the first of three models to be called Perry.  It was produced for four years.  It was offered in white or yellow 14K gold filled.  It's barrel-shaped case is accented by six ridges or bands around it's middle.  It was offered with a luminous dial, a black enamel dial, or a dial with applied gold numerals (or raised gold figures as it was marketed at the time).

It's interesting to note that the price for the Perry decreased slightly over it's production run - considering it was the height of the Great Depression.  $55 in 1933 was roughly $1,300 by today's standards, so the Perry was not an inexpensive watch.  I think the white Perry is a little more scarce than the yellow.  When paired with the black enamel dial, I think the white Perry is my favorite option.

My project watch is the luminous version and it's missing a little of the luminous paint from the hour hand.  This case is in remarkable condition.  It doesn't show the typical wear through on the corners of the bezel that most Perry's will show after decades of shirt cuffs rubbing across it.  The crown is considerably more worn than the rest of the watch.

It's not unusual for the case back to show a lot of wear around the perimeter and, like the bezel, this back is in great shape.

Uh oh... here we go.  The movement is fairly clean.  I can see there's an incorrect screw holding the barrel bridge near the crown.  What's more concerning is the regulator is set to "super slow" and, if you look closely, the hairspring is not centered, it's shifted down and to the left in the photo below.

I usually don't test watches before I service them because it's like measuring how dirty you are before you take a shower... it doesn't usually matter.  What matters is how clean you are when you're finished.  In this case my senses tell me this watch has some issues beyond just being dirty, so a baseline can confirm my suspicions. 

Sure enough, it's running extremely fast - so fast that my timer thinks its well beyond the 18000 beat per hour rate that is expected.  I guess that's why the regulator is set to super slow.  

Oh well, it doesn't really change the path I have to take.  It just tells me what to expect along the way.

After disassembling the movement, I can see the mainspring is already a "lifetime" white alloy design.  That's good.  It's very dirty but it can be cleaned up and reused.

Everything is cleaned and dried.  The bezel and crystal are sitting outside in the sunshine.  The crystal was loose and I had to reglue it with UV glue - the sunshine will secure it in place.

The movement is almost reassembled.  I wound the mainspring up and it's now the moment of truth.  Will the balance run?

It's ticking away, so that's a good sign.  There's a huge difference between ticking and keeping accurate time.  You really need a timer to listen to the ticking and provide an assessment.  In fact, you and I just hear "tick tick tick tick" but there are actually several noises in the watch that are used to determine how it's running.  I set the regulator to the center to see how things are as a baseline.

Well... it's running about 10 minutes fast per day.  That's a brisk pace, for sure.  I'll move the regulator to slow and see what happens.

At full slow, the beat rate drops to about 6 minutes fast per day.

Balances from this era are very complex.  Keeping accurate time is more difficult than you might think... things like temperature, air pressure, position (dial up, dial down, etc), and mainspring tension can all impact the beat rate of the movement.  Eventually changes in materials would provide improvements but at this point in time, the balance had several design aspects to help.  One of them are the timing screws at the end of the balance arms.  If you screw them inward, the watch will run faster.  This is not unlike a spinning figure skater... when the draw their arms in, the spin quickly and when they push there hands outward they stop spinning.  The timing screws need to be adjusted the same (in or out) but they can slow or speed up the beat rate quite a bit.

This balance already has the timing screws all the way out.  So I'm walking in someone's footsteps here - as the regulator set to super slow attests.  The next option is to add weight to the balance wheel.  That's done by adding tiny discs to the various screws on the balance wheel.  The weights come in various sizes.  This is very tricky business though - you need to add the same weight on opposing sides of the wheel so that it stays balanced.  Since I'm running 6 minutes fast, I'll start with adding two 2 minute weights and see where I wind up.

Well, as expected, the movement is running roughly two minutes fast per day now.  Another issue with this balance is the off-center hairspring.  The hair spring is very precise coil that it bound on two ends... at the center (balance staff) and the outside end (the balance cock).  Ideally it will coil evenly from the outside to the inside, centered around the balance staff.  When properly centered, the hairspring will "breath" open and closed as it rotates back and forth. 

Since this hairspring is obviously not centered, it's "breathing" can be constrained, especially if the movement is on it's side.  There's always a little more friction with a movement on it's side so a little variation is anticipated but in this case, the variation could be a lot more than the +/- 30 seconds per day that Hamilton originally specified.

At this point, it's hard to say what to do next.  There are several options.  First, I could try to adjust the hairspring - which is easier said than done.  Not only does it need to be centered, it also has to be perfectly level.  A second option would be to replace the hairspring with a different hairspring and hope they are close to the same length - as the hairspring is matched to the mass of the balance.  A third thing is to replace the balance wheel as an assembly.  That would require having a good balance wheel to use, which after almost 100 years they are getting hard to find.

I thought this would be an interesting project to share.  Readers of the blog might presume that every project goes smoothly and restoring watches is "easy".  The reality is this type of work can be tricky business and, although it's straightforward, there can be a lot of curves that have to be navigated.  Projects like this remind me that I also enjoy fishing... maybe I should do that instead.