Jack, thank you for this gift, which has given me the mindset for the next days and nights to ponder the on hobgoblins of something as close to the divine as watchmaking can get.
Interesting! A few points in this piece seem to be contradicted by university research (as hyperlinked in a relevant watchmaker interview that I’ve provided a link to below), but perhaps that’s nitpicking as this was a compelling read nonetheless. Thank you! From a Huygens perspective, it seems to me that the key technical difference of “resilience against disruption” favours the Armin Strom much more than you suggest; particularly if one’s goal is to choose a watch that most corresponds to Huygens’ functional ambitions for Resonance. Reading both pieces together provided interesting historical context as well as relevant technical information; something that marketing departments are apparently unqualified to provide :)
Thanks for your comment. Yeah, one thought occurred to me which in retrospect seems so important that I edited the article to include it, is that every single research paper I looked at without exception, modeled the behavior of coupled pendulums and I don't know whether or not those results can be generalized to watches – it seems to me that they should, since the basic physics are the same: driving force, restoring force, amplitude and mass of the balance relative to the total mass of the movement are all known, at least to the people who made the watches. I think that work on this problem in the modern era (from Journe, Armin Strom, Vianney Halter, Beat Haldimann and I'm sure others) is extremely interesting and can produce some beautiful work. But when it comes to evaluating actual watches things get tough. You can't use a standard Witschi machine because you have two escapements the device has to listen to, so the only other solution would be some sort of optical timing system (which is what Don Corson suggested in one of his articles for PuristS.com). You'd have to test the watches in terms of rate stability over a longer period of time, at various power reserves, and in various positions at minimum and then even if you ended up with one system outperforming another in terms of precision, the discrepancy would have to be significant enough that you couldn't ascribe it to sample variation. Perhaps some wealthy collector out there would like to donate some watches, hire a watchmaker so we could disassemble the watches and measure the relevant physical parameters, and then do some actual empirical science on them. Maybe I should start a Go Fund Me 😉
I haven't yet but I've been thinking about making a couple. I'm not sure how my family would feel about the kitchen table in our NY apartment being taken over by a months long clockmaking project though!
Thanks Jack! Quick question re Breguet repositioning the balance cocks. Does this change coincide with the inclusion an easier mechanism to move the right hand balance towards the other?
You know that absolutely never occurred to me despite the amount of time I've spent staring at these movements and you might very well be right. I'll have to take a closer look but that seems the likeliest reason!
Hi Jack. Really interesting to see the result of your immersion in the subject. This might be a fairly irrelevant question, a stupid one or both, but it crosses my mind to ask whether, if coupled oscillators can take any number of forms, resonance could also be observed in two quartz crystals? To what useful end I'm not particularly sure, but the idea of Citizen creating a "Calibre 0100 à Resonance" and the reaction from Switzerland does amuse me slightly.
Well, I had never really thought about it but I can't think of any reason in principle you couldn't have a system of coupled oscillators with quartz crystals – they're piezoelectrically drive but they're still mechanical oscillators. I can imagine – well, a lot of things 😂 but you could have some sort of MEMS oscillator bridge connecting the two crystals, or some sort of electronic feedback circuit. That's a very interesting idea. I wonder whether it would offer any practical benefits – I mean, for all that Huygens' and later systems are interesting in terms of progress in precision timekeeping they seem to have been a dead end, but it would be an interesting experiment. Citizen could certainly afford it.
My quibble with the Strom solution is not the coupling (which seems robust, for better or worse), but just the extent to which both oscillators are compromised by having studs wagging around at every oscillation.
This seems obvious to me to be the equivalent of an escapement error on steroids: the maximum perturbation is occurring at the limits of the amplitude, when it will have the greatest impact on the isochronism (the H2 may suffer from this somewhat too?).
Would love to see the Strom tested for chronometric performance in some laboratory and real-world environments.
As best I can tell, Vianney's DSRT does not employ movable studs like the Strom. I say that because Vianney's piece can operate either in phase or anti-phase, and I don't think a connection via movable hairspring studs would allow anti-phase synchronization. Both oscillators will still be "perturbed" in the sense that some external force is causing them to find synchrony with each other, but not the same kind of dramatic "escapement error on steroids" that the Strom surely "enjoys."
Jack, thank you for this gift, which has given me the mindset for the next days and nights to ponder the on hobgoblins of something as close to the divine as watchmaking can get.
Thank you Rich and many thanks for reading.
Interesting! A few points in this piece seem to be contradicted by university research (as hyperlinked in a relevant watchmaker interview that I’ve provided a link to below), but perhaps that’s nitpicking as this was a compelling read nonetheless. Thank you! From a Huygens perspective, it seems to me that the key technical difference of “resilience against disruption” favours the Armin Strom much more than you suggest; particularly if one’s goal is to choose a watch that most corresponds to Huygens’ functional ambitions for Resonance. Reading both pieces together provided interesting historical context as well as relevant technical information; something that marketing departments are apparently unqualified to provide :)
https://monochrome-watches.com/in-depth-resonance-christiaan-huygens-the-longitude-problem-interview-with-armin-strom-claude-greisler-on-creation-of-mirrored-force-resonance-watch/
Thanks for your comment. Yeah, one thought occurred to me which in retrospect seems so important that I edited the article to include it, is that every single research paper I looked at without exception, modeled the behavior of coupled pendulums and I don't know whether or not those results can be generalized to watches – it seems to me that they should, since the basic physics are the same: driving force, restoring force, amplitude and mass of the balance relative to the total mass of the movement are all known, at least to the people who made the watches. I think that work on this problem in the modern era (from Journe, Armin Strom, Vianney Halter, Beat Haldimann and I'm sure others) is extremely interesting and can produce some beautiful work. But when it comes to evaluating actual watches things get tough. You can't use a standard Witschi machine because you have two escapements the device has to listen to, so the only other solution would be some sort of optical timing system (which is what Don Corson suggested in one of his articles for PuristS.com). You'd have to test the watches in terms of rate stability over a longer period of time, at various power reserves, and in various positions at minimum and then even if you ended up with one system outperforming another in terms of precision, the discrepancy would have to be significant enough that you couldn't ascribe it to sample variation. Perhaps some wealthy collector out there would like to donate some watches, hire a watchmaker so we could disassemble the watches and measure the relevant physical parameters, and then do some actual empirical science on them. Maybe I should start a Go Fund Me 😉
Great job! Have you purchased two cheap identical pendulem clocks, hung them side by side and seen the effect for yourself?
I haven't yet but I've been thinking about making a couple. I'm not sure how my family would feel about the kitchen table in our NY apartment being taken over by a months long clockmaking project though!
Very interesting. As always you incorporate a bit of poetry into a very technical subject. Thank you.
Thanks Jack! Quick question re Breguet repositioning the balance cocks. Does this change coincide with the inclusion an easier mechanism to move the right hand balance towards the other?
You know that absolutely never occurred to me despite the amount of time I've spent staring at these movements and you might very well be right. I'll have to take a closer look but that seems the likeliest reason!
Hi Jack. Really interesting to see the result of your immersion in the subject. This might be a fairly irrelevant question, a stupid one or both, but it crosses my mind to ask whether, if coupled oscillators can take any number of forms, resonance could also be observed in two quartz crystals? To what useful end I'm not particularly sure, but the idea of Citizen creating a "Calibre 0100 à Resonance" and the reaction from Switzerland does amuse me slightly.
Well, I had never really thought about it but I can't think of any reason in principle you couldn't have a system of coupled oscillators with quartz crystals – they're piezoelectrically drive but they're still mechanical oscillators. I can imagine – well, a lot of things 😂 but you could have some sort of MEMS oscillator bridge connecting the two crystals, or some sort of electronic feedback circuit. That's a very interesting idea. I wonder whether it would offer any practical benefits – I mean, for all that Huygens' and later systems are interesting in terms of progress in precision timekeeping they seem to have been a dead end, but it would be an interesting experiment. Citizen could certainly afford it.
Thank you for this, Jack - a very interesting and engaging piece, and one which I think will repay being read a few times over to fully appreciate.
Most excellent and entertaining read!
My quibble with the Strom solution is not the coupling (which seems robust, for better or worse), but just the extent to which both oscillators are compromised by having studs wagging around at every oscillation.
This seems obvious to me to be the equivalent of an escapement error on steroids: the maximum perturbation is occurring at the limits of the amplitude, when it will have the greatest impact on the isochronism (the H2 may suffer from this somewhat too?).
Would love to see the Strom tested for chronometric performance in some laboratory and real-world environments.
Would you imagine this perturbation to happen with Vianney Halter's Deep Space Resonance Tourbillon as well?
As best I can tell, Vianney's DSRT does not employ movable studs like the Strom. I say that because Vianney's piece can operate either in phase or anti-phase, and I don't think a connection via movable hairspring studs would allow anti-phase synchronization. Both oscillators will still be "perturbed" in the sense that some external force is causing them to find synchrony with each other, but not the same kind of dramatic "escapement error on steroids" that the Strom surely "enjoys."