What’s on your watchmaker's bench today?

I actually might skip... all of the sewing needles in my wife's needle drawer are WAY too large to spin the regulator. THis one might just be a 'runs not bad' for now 😀

In a moment of ~bordem~ ~stupidity~ inspiration, I decided to try adjusting the collet without removing it from the balance cock. I just put it on a spike, and carefully put a screwdriver down from the top and tweaked it. It took about a half dozen tries, but I managed to just get silly--lucky all at once.

Dial up is ~+10s/d, -15amp, and 0.1ms beat error (I figure the two are a rounding error apart). So way better than I am typically capable of 😀

Good notes about the screwdriver!

I've seen those tools! I actually tried to pick up a few about 4-5 months ago and it was either "wouldn't ship because europe" or out of stock. I should go looking again. Obviously they are pretty expensive. Seems easy enough to make from tube stock! Though this is a task I've done pretty rarely, so shrug

Esslinger sells a version of the collet turning tool Al posted above here: https://www.esslinger.com/watchmakers-hairspring-setting-tool/?searchid=16413773

Comes in three sizes:

• 1.10mm; hole size: 0.45mm
• 1.60mm; hole size: 0.55mm
• 2.10mm; hole size: 0.75mm

ooh, neat! Any idea which ones are a 'good' size? It would be nice to not buy the whole set particularly if 1 size was a 'never use'.

Good question that I don't have an answer for, but I'm curious to know as well. Maybe Al can share which size he uses for vintage Omegas like the 50X movements.

This the Bergeon version of the tool from Cousins. I'm assuming they sell the 1mm and 1.2mm sizes more than the others.

Something a bit different. Cleaning up a mechanical bird mechanism, Oiling it and dusting things off. I have a number of these, This one I keep uncovered for when I would give lectures on automata.

Tomorrow I am giving a talk titled "A child's history of AI." A subject I have been researching since the maker faire last fall.

This one is sponsored by the local maker space, and called a S.T.E.A.M discovery festival.

I also have my crank organ which I have been bringing for over a decade.

I would do a bit at the Dickens fair noting how Ada Lovelace's abstractions evolved the music box into a cell phone. I would start by stating she told me to connect my music box to the telegraph?

... the wires would get all tangled, would be my implied response.

Her counter argument would be that she wanted to study with Michael Faraday who said electricity and magnetism were the same thing and the music could be sent wirelessly.

The next part would be about sending pictures through the transatlantic cable. And so on. Relating how the punched holes in the paper could not simply represent words. They could be sounds or colors.

For this lecture I want to explain Time space transforms. Which involve the sums of sin and cos waves. So the birdie is the perfect example. In a lot of way neural network activation functions can be modeled as cam follower systems.

To the Romans and the 18th century audience, what rediscovered the technology, Such toys were much like AI is now. They seemed alive.

1655 with the electronics disassembled.

Looks like the issue was water damage. There was still liquid between the display and the reflector. I do not think it is the nematic crystal fluid as the glass seals look to be intact.

The gunk on the battery was cardboard used to wedge it in. One screw which hold the + battery contact and retains the battery is rusted in place. I put some penitrating sili-kroil on it. This also holds a strip of plastic coated insulating paper, which is basically parchment paper.

Part of the plastic retainer frame, what holds the display, has disintegrated. Sadly this was the part where the serial number sticker was. Not sure why they did not engrave this onto a metal part. Now this information is irrevocably lost. Part of the retaining clips also seem to be missing.

The spotting on the display does seem to be between the glass plates. If I can figure out a way to connect tiny probes, I can see if the nematic fluid is still active. If I hold the glass at an angle I can see the plate contacts.

The whole point of this project is to revive my old LCD testing setup. Where I was probing unknown displays. That is a subject unto itself.

The circuit board looks to be in better shape than the one offered on line. (I see the seller made me an offer on theirs.) There is some obvious rework on one of the traces. Everything is gold plated, although this would amount to a microscopic amount of gold.

Modern chip packages can be moisture sensitive. So are baked before re flow soldering. This is more to keep them from blowing up when the steam escapes. Not sure if the epoxy potting compound is water resistant. Most of the water damage is around the chime speaker and the transistor what powers it.

From skimming the literature, this is an hourly chime and not an alarm. There are quite a few functions this can do as there are 5 contact buttons. There are also a number of test points on the side of the PCB.

The analog side looks to be in good condition. There is no signs of rust or water damage. The coil seems to be in good shape. I have not measured it yet as it was pretty late by the time I got this apart.

Finished up (I hope!) my Rolex Date project, and started a new one! An Omega 865 Chronostop.

I managed to end up with TWO Chronostops, this is a grey dial "Driver" model. I also have an incoming white one that also needs a seconds hand, but I have two new ones showing up!

Fun movement so far, pretty easy to tear down. I'll have a bit of work to get it together, but I ran out of time not long after I took it out of the cleaning machine.

Yeah it is a mess. This is where I do my micro electronics rework.

Finally got around to testing the 1655 circuit board. I got a new frame and clip. The LCD is kaput. I also got some 1365 watches the first of which arrived yesterday. (it works.) These calibers share parts (but not electronics.)

The 1655 was badly corroded. (I have threads planned for detailing the teardown etc.) So I did not have too much hope for the electronics. I did make some quartz test fixtures (subject also of a future planned thread.) Which is part of this mess.

The new frame would not fit as the alarm capacitor was mounted in place crooked. The board has a lot of signs of prior rework. Before removing this and the broken "earth" pins. I wanted to power up the chip. Normally the contactor plate supplies + voltage to different parts of the board. This though blocks some of the test points I wanted to scope out. I was able to use some test probes to simulate the connector points.

I did not have much hope, so I was glad to see the quartz was oscillating. I spent the last few days researching pierce oscillators in case I needed to check the crystal outside the circuit.

The next probe was the LCD backplane signal. (subject of the not yet posted LCD thread.) This was present indicating that the chip is probably good.

The remaining test was the motor driver for the analog side. Since the pulses are once a minute, this is a bit awkward. The maintenance sheet does have a section which says to apply + voltage to the corrector contact. On the 1365 this makes the movement spin fast. So I can see the pulses on the 1655 board. (this means the analog side may work. which is what we hope for when bidding on such things.)

Having tested the chip I set about the rework. Starting with what are called 'earth pins,' but they carry the alarm signal through the dial to the piezo speaker. I bought one from Canada, not realizing there are two. The remains were pretending to be blobs of rust.

The alarm cap will be tricky to source as this is not a footprint what became popular. My guess it is a small electrolytic. There is also a driver transistor (or FET) what is connected to this. Since the speaker is part of the crystal, which along with the bezel is missing, It will be some time before any of this is needed.

The kaked LDC though is a bit of a disappointment. I have a whole write up on that. As far as I can tell the epoxy failed and all the Liquid crystal nematic fluid leaked out. So I will need to source a new LCD glass. Which I have noted is a subject unto itself.

edlt: the thing with the beer taps is my automated APS film processor upside down and acting as a scope table.

I scored another rust bucket.

Do not use acetone to remove the PCB from a motor coil !!!

It reacts with the wire bonding agent and releases the tension on the winding.

I also noticed there seems to be two designs for the motor. One has a central core which can be pressed out. The other has the core spot welded in place.

I think the only way to fix these is to wind a new coil. I have a winding machine for pipe organs. This though uses a much thicker wire which is visible in the photograph (AWG39). On the other hand giggle AI says that the gauge should be around 42. Such wire is used for guitar pickup coils, so is fairly cheap.

There are no 1340.9400 listed on eBay at the moment. Plenty of watches and movements. One listing has the coil dismembered from the can.

Getting another parts watch is not really a practical solution, since these 4 movements/coils are from parts watches.

Have the same issue with 330 bumper hairsprings. Lots of balances, no hairsprings.