Watchmaking: Machining a Watch Gear

Screw Driver
87 followers
#machinery #motorbikes #occupations #career
If you would like to help support our channel, please consider subscribing to our Patreon page: / chronovaengineering
In this video, Mike machines a gear wheel for a watch.
There's more to machining watch gears than meets the eye.
In this video I will introduce some of the secrets behind how these components can be made,
including my own take on traditional watchmaking techniques.
Unlike the watch screw we made previously, which we mounted in the lathe as bar stock,
our gear starts as a disc blank I prepared off camera. As such we need a different way of holding.
Historically watchmakers used an arbor to which the blank could be glued using shellac.
In the modern age we can use super glue instead, which is much more convenient.
Like the gear itself, this arbor can also be machined on this little watchmaker's lathe.
I'm starting from a brass blank, 25mm in diameter.
I need to drill a hole in the arbor to accept a spigot, which I will machine later.
We'll be gluing the spigot to the brass arbor using a strong metal adhesive.
For the adhesive to work effectively, the gap between the spigot and the brass
needs to be between 25 and 80 microns. Thanks to the invention of the external
and bore micrometer, we can easily machine this accurately.
Moving on to the spigot, I start with the steel blank I made previously.
It's simply a matter of machining this down to the correct diameter,
with the spigot mounted in the headstock.
Now I apply adhesive to the metal surfaces. It's worth pointing out that on a larger machine,
making this super glue arbor from one piece would be the preferred approach.
Now I machine the diameter down to less than the root diameter of the gear.
This will allow the cutter to clear the arbor when we machine the teeth.
In some cases, such as when making escape wheels, the teeth are fragile,
and instead we would leave the arbor larger to support the teeth during cutting.
I have made the arbor longer than necessary so I can reuse it in future projects,
and since there's no point in reducing the diameter at the back of the arbor,
I put on a taper to clear the gear cutter.
I drill a hole in the centre that's larger than the hole in the gear,
and machine air grooves so the super glue can cure.
I start with a gear blank 0.45mm thick. This is 50 microns over the final dimension
to allow some room for cleaning it up later.
The blank was taken from round bar rather than plate
due to the direction of the residual stresses in the material.
This means that as I machine the brass away, warping will be minimised.
Super glue is applied to the arbor. Because I machine the face in situ, I know it is accurate.
I left the blank oversized so I'm not concerned about concentricity as I glue it in place.
Now it's time to put the hole in the middle of the gear.
This is important to get right. If the hole is slightly off centre,
the gear won't mesh smoothly and the watch simply won't work.
Due to the small scale of these watch components, precision is key.
Drill bits are a good way to make holes, but they are certainly not accurate enough here.
We need to bore the hole using this micro boring bar.
I made this simple holder so I can mount the round boring bar in my tool post
accurately on centre height.
I bore the hole just shy of my final target size which is 2mm.
I'm lazy so rather than boring to the exact size, I take it to within tolerance with a reamer.
Now it's time to machine the outside to diameter.
The final diameter is determined by the number of teeth, in this case 75,
and the module of the gear, in this case 0.25.
The module of a gear is a metric that quantifies the size of the teeth.
Perhaps surprisingly, the exact final diameter is not critical
since we will be using a process called depthing to determine the correct meshing distance
before drilling the watch plate.
Mass produced mechanical watches however are made on precise CNC machines.
In this case the wheels have little variation and can be positioned based purely on measurement.
Here I am setting up for dividing.
I custom machine the spindle for my watchmaker's lathe to fit a super
precision rotary table I procured from an old optics lab.
To fit this system I remove the drawbar from the back of the spindle
whilst taking care to minimise disturbance to the gear blank.
I slide the rotary table onto the back of the spindle and use an ER16 collet to grip it.
One thing I love about this setup is that I can use the rotary table in its manual mode
or I can connect the stepper motor to a magic box of electronics designed by Hazel,
our electronics expert.
For cutting the gear teeth I use a system I designed and built myself,
incorporating a micrometer stage and a small DC motor to drive the spindle.
I check the attachment is bolted square to the bed with a tri-square.
For cutting the gear teeth it is possible to purchase a commercial cutter such as the one
you see here, but, you guessed it, for this example we use a shop made cutter,
which is much cheaper than these commercial options.
Its disadvantage is it only has one cutting face,
so I will have to take the feed slower than I would with a multi-point cutter.
The cutting tip is aligned to the centre of the axis
and the cross slide is locked for the remainder of the tooth cutting process.
I gradually lower the cutting tip until it just starts to scrape the blank.
From this point I can add the required depth of cut which I measure with the micrometer barrel.
Now the teeth are cut I clean off my writing on the wheel since I'm far
enough through the process to remember which wheel I was making.
To provide a sense of scale, here I show the size of the teeth compared to a matchstick.
Under the scanning electron microscope it looks completely different,
all the imperfections can be seen. Nevertheless, these teeth have turned out satisfactory.
At this point you would be forgiven for thinking it is time to call it a day.
The wheel will work like this, but spokes really are aesthetically necessary.
For faster moving wheels, the spokes have added functionality of reducing the moment of inertia,
improving the efficiency of the watch train.
In the horological world, cutting the spokes is called crossing out.
According to George Daniels, wheels this size are crossed out with a saw and files,
and this does indeed work just fine, but perhaps it's time to try something new.
Here I have a go at milling out the spokes with the wheel still mounted in the lathe.
I can swap the spindle round in my gear cutting attachment to turn it into a
milling and drilling fixture instead. In its new configuration,
I mount the attachment onto the cross slide and centre it on the wheel.
For the milling cutter I use this 0.5mm tungsten carbide end mill.
This is quite small, so care is required when using this cutter.
It's crucial the cutter is held concentric, otherwise it will break instantly.
Thankfully my watchmaker's collets are accurate enough for the job.
The geometrical calculations were harder than I expected, or at least that's what I told myself
when I made the initial cuts in the wrong place. Luckily this was rectified with a second pass.
Once the spokes have been cut, I move on to the rim before finishing up with the hub.
Acetone softens the superglue and the wheel can be removed from the arbor.
I use some 2000 grit emery paper to remove burrs left from the machining.
I rinse off the grit in solvent, but I will be ultrasonically cleaning the
wheel before fitting it into the watch. The milling was a success, but it isn't perfect,
therefore I need to clean it up with fine escapement files.
A great way to protect the wheel when holding it in the vice is to line the jaws with leather.
Leather is grippy so holds the wheel firmly with minimal jaw pressure.
I begin by filing the spokes. These only need a light touch up with a barrette file.
The barrette file only cuts on one side and has angled safe edges on the other.
This allows me to get into the corners without disrupting the other edges.
This file also works well on the hub. I use the tip of the file only due to the
restricted space between the spokes. A nice way to hold the wheel when working on the rim
is to hold it by hand against a cork. Champagne corks work well, partly because they're bigger
than wine corks, but their convex surface also makes it easier to work on the component.
The wheel teeth dig into the soft surface without worry of damage.
The wheel is now ready to be packed away until the pinion is made and ready for staking onto
the wheel. Membrane boxes are perfect for storing delicate watch components.
Thank you for watching. If you enjoyed this video please consider subscribing.
See you in the next one.
