1950 Spencer

Hamilton made watches in all sorts of materials but rarely in "cheap" materials, that is, at least, up until the late 1950s when the watch industry faced significant competition from lower cost global competitors.  For many decades Hamilton used metals that ranged from platinum on the high end, to 10K gold filled on the low end, eventually introducing stainless steel for the most durable models.

Gold filled has a couple of different levels but generally it's 1/20 gold filled - meaning 1/20th of the case thickness is layered in gold - that's 5% and a significant difference over gold plated cases where the gold layer is measured in microns.

Fitting almost squarely in the middle is 10K solid gold - above which you'd have 14K, 18K and platinum and below would be 14K gold filled, 10K gold filled, eventually Rolled Gold Plate (or RGP), and stainless steel.

Solid 10K gold is the most durable of the solid gold options but it wasn't used all that often.  One of the models to feature it was the 1950 Spencer.

The Spencer was produced for five years (through 1954) and featured a sterling solver dial with 18K solid gold numerals.  Tucked inside the case was Hamilton 17 jewel 8/0 sized 747 movement.

My project Spencer reminds me of Jack, from Jack and Jill - it appears to have lost it's crown and the number 7 has come off too.  Fortunately it's caught under the hour hand so it will be easy to replace.

The case back looks fine but it probably has a bit of wear based on the soft details of Hamilton and 10K solid gold.  It will probably look a lot better with some light buffing.

The dial looks to be an older refinish.  It's not bad but I can tell from the finish that it's not an original dial.  It's the correct pattern though and with the 7 put back in place it looks much better.

The hole in the side of the movement where the stem goes makes me wonder if the stem and crown came out together.  That's pretty unusual.  Perhaps there's an issue with the set lever that holds the stem in place.

The case back makes it easy to identify the model - it's clearly stamped right there in the center.

There are numbers scratched into the back of the dial - proof positive that it has been refinished.  I'll put a small dollop of UV glue over the posts that hold the 7 in place - that will secure the posts once I place the dial in the sunshine for a few minutes.

With all the parts stripped from the back of the movement, I can see there's a portion of the stem still holding the winding pinion and clutch in place.  That means the set lever is doing its thing.  I can turn the set lever screw a few turns and slide the remainder of the stem out the hole.

Everything is cleaned and ready to be reassembled.  I like the 747 movement.  It's very well designed and  almost reassembles itself.

The completed movement is ticking away with a good motion.  So it's off to the timer to listen to how it's working.  This photo shows the serial number of the movement - Y149205.  That dates the watch to 1949.  The model years for watches are similar to automobiles, the next year models came out in the Fall of the prior year.  So this watch is likely a 1950 model from late 1949 - perhaps a Christmas present.

It's running a little fast and the beat error is well out of range.  The amplitude of 209 degrees is on the low end.  That will likely come up after I adjust the beat error.  The 8/0 movements actually have several different strength mainsprings.  This watch already had a white alloy spring but it might be the minimum strength option.

Reducing the beat error requires pulling the balance from the balance cock and rotating the hair spring collet in the correct direction the correct amount - so it's not a task for the faint of heart.  One false move and the watch will only tell the correct time twice a day.. as it will no longer work.

Fortunately I chose wisely and went the correct direction the correct amount.  The beat error of 2.3ms is well within my personal specs of 3.0ms and the benefit of reducing it further really doesn't justify the potential risk of goofing it up.  I like the Swiss-made Hamilton calibers more when it comes to adjusting beat error - you can fine tune it to near-zero without risking the balance because the hairspring stud is moveable (most of the time).

A couple of tweaks to the regulator brings the beat right right in line.  Notice the amplitude came up to 251 degrees - that's perfectly fine.

The last thing this watch needed was to have a new stem and crown installed.  Voila - it's now ready for another 75 years of wrist time!

1936 Norfolk

In 1935 Hamilton introduced the Elinvar hairspring in men's wrist watches.  Elivar was introduced earlier in railroad pocket watches.  The new hairspring offered more consistent timekeeping over a variety of temperatures.  1935 also introduced a new series of models with 14/0 size movements with Elinvar hairsprings, and the 987-based models were outfitted with the 987E (for Elinvar) which replaced the 987F.  The F in 987F stood for Friction - as the jewel settings were friction fit instead of being held in place with screws in the original 987.

In 1936, another men's model was introduced with an Elinvar hairspring - this time with a movement used in the ladies line-up - the 989.  Officially, the movement was the 989E but Hamilton's catalog just referred to it as the 989.

The Norfolk was the first men's watch in over a decade to not feature a sub-second hand.  I think it's interesting that later in the 1930s Hamilton would introduce another watch without a second hand - the Contour - but that model used the 14/0 caliber with a shortened 4th wheel bit (the second hand attaches to the 4th wheel).  

The Norfolk was only offered for a single year.  So it begs a question - why use a ladies movement?  Was the Norfolk a trial for men's watches without second hands?   Did engineers realize a 980 could be modified to work instead?  Beats me - but the Norfolk isn't overly rare - and Hamilton sold quite a few.

My project watch was sent to me with a description that "it would run for a while and then stop".  That could mean anything when it comes to watches... maybe it's dirty, maybe it has issues.  Running at all is typically a good sign but you never know.  There's a big difference between "ticking" and "keeping time".  

Looking at it, as received it appears to be in good shape.  The case isn't overly worn and the dial has been refinished - so this watch was definitely taken care of.

The back of the case is unremarkable - it shows just a little wear to one of the corners.

I don't normally bother checking a watch on the timer before I work on.  Why bother?  I'm going to take it apart anyway.  However, I've learned to listen to the little voice in my head and today it was saying, "do a pre-check" so onto the timer it went.

With this model, it's hard to tell if the watch is running since there's no second hand to observe but the timer listens to the ticking.  According to the display, the timer is hearing all sorts of noise - so much that it cannot even register how it's running.  This shotgun pattern is never a good sign.

My first though was perhaps it was magnetized.  That's an easy check - all I need to do is pass it though my electric demagnetizer.  This tool generates a magnetic field.  If you touch iron-based metal to the side of the tool it will magnetize it.  If you pass it through the center of the field, it will demagnetize it.

No deal - although the timer is picking up a different pattern of noise now.  There's something definitely going on inside this movement.  We'll just have to see if a cleaning will take care of it.

The Norfolk has a two-piece case and the movement is held in the case back.  With the front bezel removed you can see the dial has a light scrape in the center that extends to the 8 and what appears to be a finger print around the 6.  The font for Hamilton is a little off center to my eye so this refinished dial is nothing to brag about.

The movement lifts out easily.  The 1930s was an interesting period - it's the only timeframe where it's okay to not see Hamilton Watch Co Lancaster PA inside the case back.  Keystone and Wadsworth cases are often seen.  With any other time period you should be suspicious if the case back doesn't say Hamilton.  But not today - this is a legit Norfolk.  You can see the movement is a 989E, just as expected.  

Personally, I don't care much for this movement design.  The balance wheel sticks out like a sore thumb and it's very exposed when not in the case.  One misstep and you'll break a pivot on the balance staff.

The back of the dial has a heavy coat of epoxy securing the numerals in place.  The applied gold numerals have little posts that extend through the dial.  When new, they are burnished in place like rivets.  When the dial is refinished, the posts need something to hold them in place.  Glue is a common choice.

The first thing removed is the balance assembly in order to keep it out of harm's way.  Then I'll start to disassemble the remainder of the movement, piece by piece.

I'll get out my small stash of 989 movements in case I run into something.

I also happened to have two Norfolk dials... one appears to be original with a nice even patina.

The other is refinished, new-looking but without a finger print.  So I have a couple of good options for this project.

Everything is cleaned and dried.  This movement is much smaller than a typical men's movement and it requires a lot less space to dry.  It's the same number of parts though - just a lot smaller and that makes it a little more challenging to reassemble.

The barrel bridge and train bridge are reassembled. That allows me to verify that the train wheels rotate freely before I install the pallet fork.

Everything is looking good.  It's ready for the balance but I'll put the parts back on the front of the movement first - that way the balance will go on last.  It's less risky that way.

I'll put the original dial on in place of the one it came with.  Since the case is so nice, it's better to have a decent original dial with patina than a refinished dial with a finger print.

Okay - ready for the balance to be installed and then it's off to the timer.

Ugh...  the timer is picking up the beat rate but it's running about 10 minutes fast per day.  That's REALLY fast and well outside the regulator's ability to adjust.

A pass through the demagnetizer might do the trick.  The demagnetizer tends to shake springs into place sometimes.  However not this time. now it's running a little faster.

The beat rate is a function of several things.  You could say the "balance" of the balance wheel is related to how well poised it is - perfectly weighted all the way around.  But it's also related to the balance between the weight of the wheel and the length of the hairspring.  Those two factors are critical to how quickly the wheel swings side to side.  If the two factors are perfectly synced, you don't even need a regulator. 

Regulating the beat rate is normally accomplished by moving the "regulator".  The regulator has two posts that straddle the hairspring and effectively adjusts the length of the hairspring.  A little shorter makes the rate faster, a little longer makes it tick slower.

However, it's possible to speed or slow the beat rate by using the timing screws on the end of the balance wheel arms.  This is similar to how a spinning figure skater changes their rate of spin - they pull their arms in to spin fast and then stick them out to slow down.  The timing screws on this balance appear to be screwed all the way in which is the fastest position.  Perhaps someone messed with this balance?   I'll see if moving them out a few turns will make a significant difference.

Well, that's a little better.  I'll try again.

3.5 minutes fast per day is the best I can do with adjusting the screws.  My next option is to add timing washers to slow the beat rate down.  Adding weight to the balance wheel will slow the beat rate.  I just need to add the same weight to each side in the same position, 180 degrees opposite to each other.

If you happened to hear an anguished cry of despair echoing over the Blue Ridge Mountains - that was me.  While installing the second of two timing washers I dropped the balance about a half an inch and that was all that was needed to kink the hairspring.  Hairsprings from this era are unforgiving and they are so small that only and absolute expert can make them correct again - and that's not me.

So I had to resort to one of my parts movements.  After a few attempts I finally got the movement running again.  This time I get 57 seconds fast per day and that's as good as it's going to get.  I've pushed my luck to the limit.  Did I mention I don't like 989 movements?  As Shrek would say, "That'll do Donkey, that'll do".

Alright, this little 1936 Norfolk is finally running and keeping decent time for an 80 year old watch.  I think the new "original" dial is a huge improvement... don't you?

1973 Auto Date Buccaneer

 Hamilton introduced calendar models in the early 1950s but if you if you see a Hamilton with a day and a date then you can be sure it's from the 1970s or later.  There are quite a few of them out there but, to be honest, I don't find them all that desirable.  It's not because I don't need to know what day it is, it's more a a result of the big and chunky 1970s styling.

Back in 2018 I did a post on a watch that looked a lot like an Auto Date Buccaneer.  Today I get to do another post on a watch that is much more likely to be a Buccaneer.  

The Buccaneer was introduced in 1973 and made for two years.  It came in a stainless steel case with a matching bracelet - yours for $110 - that's a little over $700 in todays dollars, and probably more like $1,100 once Trump is finished making America "great".

In 1974 the priced was increased to $120.  The Q/C II means the watch has a quick change setting capability so you can set the day and the date without moving the hour hand.  The stem has the positions... the winding position, then you pull the crown out a little to set the day in one direction or the date in the other.  Pull the crown out all the way and you can set the time.

My project watch arrived sans bracelet but appears to be in decent shape.  Stainless steel cases can usually take a beating and this one shows it's seen some action over the last 50 years.  I can tell from the catalog image that the bezel originally had a brushed finish.  That could be restored but I don't have the wheels to do it.  A good cleaning might be the best place to start.

Interestingly, this case back has the number 820002-3, which is different that the watch I did in 2018.

Based on the deep scratches on the case back, I'm going to assume the case back in on tight so I'll hold the case in a vice and used extra downward pressure on the case wrench.

Here's another interesting find, this watch has an 826 movement in it and not the 825 that I found in the 2018 watch.  All these ETA movements look very similar and share a lot of parts but they are not all the same.

Looking closely at the hands, they were very close to the dial.  That was a little odd.  I have to pull them anyway to remove the dial.  The dial is held on by two clamps that look like a backwards question mark.  You rotate them out of the movement and that releases the dial feet.

Uh-oh - the other side is missing it's clamp.  So there's nothing holding the dial in place.  The hands were close to the dial because the dial lifted off the mainplate.

With the dial out of the way you can see the calendar and date wheels.  They take up of the entire surface of the main plate so I'll have to figure out some way to secure the dial without using sticky "dial dots".  You can tell from the two wheels that the date rotates clockwise and the day rotates counter clockwise.

The day wheel just lifts right off to reveal the complications that change the day and the date.  There's a lot going on under the dial, that's for sure.

Disassembly starts with removing the hour wheel - this is what the hour hand is attached to.  Once its out of the way you can see the cannon pinion and minute wheel.  The cannon pinion drives the minute hand and the minute wheel synchronizes the hour wheel.

Piece by piece, all the parts are removed.  At this point you can see the little u-shaped spring that presses on the index lever on the left side of the movement.  This spring has a tendency to fling off into the ether and disappear forever - great care has to be used to not lose it.

Finally - the majority of the parts are removed from the front of the movement.  I'll take out the balance jewels once I remove the balance from the other side

The rotor and framework are held on with two screws.  Once they are out of the way the movement starts to look like your garden variety ETA movement.  Should be smooth sailing from this point.

Here's another interesting twist... there's no markings on the mainplate to what ETA caliber this is based on.  I expected to 2780 but there's nothing at all.  I can also see the post that the missing dial foot lever would attached to is also missing.  So I may need another mainplate to address the dial foot situation.

I have another Auto Date movement to see if I can use as a donor.

There's no rotor on it so I don't know if this is an 826 but it sure looks similar.

Nope - this setup is different from my project watch - I'd rather not get change out all of these parts.  That could be a larger can of worms.

While everything is in the ultrasonic, I'll install a new EverTite crystal with a silver (white) reflector ring.  The original crystal had some deep scratches that would be difficult to polish out.  A new crystal will be a nice improvement.

Everything is cleaned and ready for reassembly - there are a lot of parts in this movement, half of them are just for the day and date complication.

The basic movement is back together and running with a nice motion.  Let's see what the timer thinks of it.

It's running a little fast but the amplitude is great.  The beat error is well within my personal specs but I should be able to easily adjust it.  Notice the beat rate is faster than the usual 18000 beats per hour most pre-1969 models have.

Well, it turns out the hairspring stud won't move any further in the direction it needs to go so if I want to reduce the beat error I'll need to pull the balance from the balance cock.  That would really tempt fate so I'm going to let sleeping dogs lie and just slow the watch down a smidgen.  The beat error will have to do.

I advance the time until the date changes and then can set the hands at midnight.  After working with the movement and hands, I can see the dial is too loose without something to hold the dial feet.  So I need to think of something else to try.

The movement has two half-rings on the circumference that hold the dial off the movement.  I'll use a few sticky dial dots to hopefully hold the dial down on the side without the dial foot lever.

You can see the four round dots applied to the movement.  I'll fold them over to double up the thickness.  Once the dial is pressed down, it should be enough to hold the dial in place.  The round hole between the dots is where the dial foot goes through the movement.

Well this was more of a project than I was expecting but I think it turned out well.  Only time will tell if the dial stays in place.  If it moves the day of the week wheel will probably stop working and eventually the hour hand will hit an hour marker.  We'll just have to see... there's not much I can do without another 826 movement.