Seeing is believing

Watchmaking is all about precision - and taking accurate measurements during the design and manufacturing process is essential. However, when it comes to the art of horology, often the relationship between arbours, bearings and gear meshing is a matter of 'how if feels' rather then how it should be engineered. As strange as it may seem, often the 'ideal' fit is not the most desired one; theoretical shapes and tolerances are not the most perfect. Why? Because tight tolerances do not necessary translate into long-term reliability or the best timekeeping under stress, changes in temperature and gravity. There is a saying that engineers are poor watchmakers which is equally true the other way around.

Some of you may remember the acquisition of a microscope for our workshop. I certainly do - we  sacrificed a really fine watch in order to acquire this lovely instrument. But it was worth it!  When it comes to the inspection of surfaces, part geometry, and the inspection of cutting tools, a microscope is worth its weight in gold. Trying to make a watch without being able to understand how the Swiss have done it would be impossible. So we observe, learn and try to replicate.

Here are just two examples. The first photo shows the mesh in a high grade Swiss watch. The second one is our attempt to copy the Swiss. A person with a keen eye could clearly see the difference in the meshing depth between the two examples. Without any doubt, our mesh is far tighter, more precise, and follows the 'ideal' calculated point of contact. However, we soon discovered that the sloppier Swiss way was actually more desirable. The winding action was actually smoother than ours! Lesson learned.

The second example: a short video showing the very tip of the balance staff inside the shock absorbing jewel. This is an actual recording of a high grade Swiss watch still in brand new condition. I have no doubt that any engineer (or car mechanic) would be horrified. But this is how it is, and this is how it's done. Check it out: Of course, do keep in mind that you are observing an arbour with 100 microns diameter under 240 x magnification.

CAM tool paths, watchmaking - main plate

Watch main plate [B]

Main plate holder [watchmaking]

The purpose of this quick report is just that - to provide an update to those who follow our watchmaking progress. Also, we feel that sharing our trials and tribulations may one day help someone on their own watchmaking journey.

Last week we started to prototype the main plate. The main focus was on positional tolerances and working out the 'perfect' press fit for the main train jewels. But it wasn't long until we figured out that we are wasting too much material; precisely over 60%. In addition, our work holding method was excellent in relation to rigidity, however, our solution would be very inefficient for mass production - even a quantity of just a dozen pieces at a time.

We had two options: To continue with prototyping or to pause and find a better clamping solution before we go any further. I voted for the first option: to keep learning and not to worry about waste. However, I was quickly outvoted by Josh (who already had Andrew on his side). The final straw was a video which clearly shows a very effective mainplate holding solution on Willemin Macodel CNC, as demonstrated at Baselworld in 2015.  Here is the link to it.  Definitely worth watching:

https://www.youtube.com/watch?v=dyqBtc-qXmE

In just 3 working days, putting in a solid 50 hours of work, Josh and Andrew designed, coded and manufactured our new 'clamp'. The benefits are already paying off: 

- we can use a smaller blank than before, while there is no compromise on rigidity;

- the brass waste is now below 50%

- the new work-holding piece allows for both sides milling plus 90 deg side cut without having to re-clamp the piece;

- and, most importantly, the new blanks could be clamped directly without pre-machining preparation which saves 25 minutes per piece!