Discussion about this post

User's avatar
John Doe's avatar

You really ought to find a technical adviser. Anyone who has any amount of experience with this kind of work can tell you that not a single one of Atelier Wen's claims about the challenges involved in manufacturing the dials are correct.

In general this article is a good summing-up but I want to point out a few things.

The claim that the guilloche machines lose calibration if left unattended for more than 15 minutes is ridiculous. A rose engine is just a lathe which operates on a slightly different principle. It is a machine tool which has deterministic results for given inputs regardless of how long it has been left unattended. The only danger in leaving work unfinished on the machine is that the operator may forget what step they are in the phasing and indexing necessary for generating the pattern. Atelier Wen's claim may be some severe and negligent corruption of this potential issue.

A note on precision in guilloche - The precision necessary in a guilloche pattern is related to the fineness and intricacy of the pattern, and the depth of the guilloche cut. An extremely fine basket weave pattern as in Roger Smith's dials will have a depth of cut of 0.042mm. The depth of cut varies depending on the pattern and the angle of the cutter, in Roger's case a 3 bar basket with a 1.2mm pitch pattern bar, and 160 degrees of included angle on the cutter. Let's presume 10 percent of cut depth is allowable without being noticeable, then 0.004 micron is the tolerable error. Already that is greater than Atelier Wen's claim of 2 micron required accuracy. Looking at, for example, the Perception V2 Piao dial, my estimation on the depth of cut is about .126. This is on the basis of a 22bump rosette and correlating the case diameter to the spacing of the cuts, and taking a conservative guess as to the cutter angle which I estimate at 140 degrees included. 10 percent of this figure is 0.012, 6 times the stated required accuracy.

Of course there are many other types of precision in lathes. Even in very precise manual turning lathes like those used for turning watch parts, like the schaublin 70, 2 microns of runout or centering is considered very good. For a rose engine such figures are unimageable and beyond that far beyond what is necessary.

As for the claims of effort or time requirements - you can take as long as you want to do a guilloche dial. If it is taking his shop 8 hours of continuous engine turning to produce these simple patterns, that is not a reflection of complexity or challenge. It is a matter of pace of work. A positive spin might be to call it deliberate, more realistically it's just slow. Or, someone somewhere is lying, or misunderstanding and re-interpreting something benign for marketing spin.

In general if someone refuses to be transparent about these things then there is something being hidden. If it was actually this difficult they would show the process in full.

Expand full comment
Watchcraazyham's avatar

Superbly put together piece, I salute you. Especially as I was critical and defensive against your critique when we discussed this at the time of the Revo launch. My apologies for that.

Expand full comment
35 more comments...

No posts