Quote:
Originally Posted by HiBoost
I see. 3647 is interesting for sure, and I can see how continuing the trend lines alone might lead to an expected (+5, 290) pairing. But in reality this dataset maxes out at 245 degrees. A better predictor might be to include data you have from a high amplitude watch where I would expect these relationships to stop being linear.
|
I think you are close but did a wrong assumption: we must NOT extrapolate the trendline in this graph.
I extended the graph shown in #3647 by adding the following timegrapher data sets:
(1) NEW: Before Regulation (highest amplitudes), no PR measurements
(2) NEW: After Regulation (reduced amplitudes), no PR measurements
(3) Before Repair (reduced vertical amplitudes), PR measurements
(4) NEW: After Repair (increased all amplitudes), PR measurements
The isochronism fit for (1) would probably look as I have sketched in the graph below (dotted black line). But I did not measure the PR before/after the regulation, therefore only one data point for (1) and (2).
The dotted vertical (green) line at Xrate = 0 s/d indicates perfect isochronism, that means the average amplitude decreases but the average rate remains constant.
Quote:
Originally Posted by HiBoost
Isochronism aside, I don't believe amplitude will continue to increase beyond its full-wind peak based on a simple regulation. Consider the Sea Dweller data you have in that graph. We drop from 245 degrees to 160 degrees and the timing drops about 5 s/d. But surely we would not expect the amplitude to go from 245 up to 330 just by regulating up 5 s/d.
|
Agreed, explained above, one must NOT extrapolate the fit (trendline) of data set (3)
Quote:
Originally Posted by HiBoost
Or to put it another way... I think of isochronism as the movement's ability to resist rate changes as the amplitude drops from its full-wind, full-strength levels. But this does not imply rate is the primary input driving the amplitude.
|
Agreed, isochronism means that the rate remains stable with changing (decreasing) amplitudes. Sure, that rate is not the primary input increasing the amplitude, but I say:
if a watchmaker trims the 32xx movement to obtain highest amplitudes, without looking at the rates, then high rates will be the result. That is the aim of my proposed experiment.