View Single Post
Old 20 March 2023, 05:02 AM   #3647
"TRF" Member
saxo3's Avatar
Join Date: Dec 2020
Location: .
Posts: 2,391
32xx movement problem poll and data thread

Isochronism characterization of 32xx movements

The measurement of amplitudes and rates (5 positions) along the power reserve, i.e., from full winding until the movement stops, delivers an insight about the 32xx caliber, which has not been presented in this thread and nowhere else in a watch forum.

The key word is isochronism, which describes the correlation between the amplitude and rate of a mechanical device such as a pendulum or a watch movement.

The desired feature is that the rate (s/d) should change very little while the amplitude (degrees) decreases with time during a power reserve measurement. In analogy to a mechanical pendulum where the rate is independent (small angle approximation) of amplitude, see my post 3338 (page 112).

How can one measure this for a watch movement? Rather simple: Amplitudes and rates are measured (after full winding) every "few" hours with a timegrapher, as shown in many posts and graphs in this thread.

The new part is that we do not plot amplitudes and rates as a function of time but study how the movement average rates (Xrate) change with the average amplitudes (Xamplitude), the latter one naturally decreases during the power reserve measurement when the watch is not moved or wound.

For a theoretically perfect mechanical movement, the amplitudes decrease while the rates remain constant. I call this the perfect isochronism.

Of course, such an amplitude independent stable rate situation will only be possible down to a certain minimum amplitude, when amplitudes further decrease, the rates will strongly deviate to (very) negative values, i.e., -10, -15, -20 seconds per day.

I have studied this effect (32xx isochronism behaviour) for about one year now.

The graph below compares the isochronism analysis results (Xamplitude vs. Xrate) for the following watches:

GMT-Master II, Ref. 126710 BLRO, caliber 3285, owner: EasyE
GMT-Master II, Ref. 126710 BLNR, caliber 3285: owner: EasyE
EXPLORER II, Ref. 226570, caliber 3285, owner: CharlesN
SEA-DWELLER, Ref. 126600, caliber 3235, owner: saxo3

The numbering (1,2,3,4,5,6) for the GMT BLNR indicates the sequence of timegrapher measurements, done (by EasyE) from full winding (1) towards the last data point (6) taken 60 hours after full winding.

One can see that for all 32xx watches (shown in this graph) the average rates decrease steadily while the average amplitudes decrease, as expected, during the power reserve measurement. It is amazing how linear this isochronism behaviour is, even to very low amplitudes. The blue shaded area indicates the -2/+2 sec/day zone.

The visible straight lines are linear fits to the data, which provide the slope values m. Compare these coloured lines with perfect isochronism case, which is sketched by the dotted green vertical line (no data) at Xrate = 0 s/d.

The best and most healthy 32xx calibers have the highest "m" value, which correspond to the best isochronism situation. Or, in simple words, for the steepest curves the movement rates are more stable while the amplitudes naturally decrease with time. This is the best isochronism as described above: rates are (more or less) independent of amplitude.

The graph also shows that, with respect to isochronism, the best watches are the SEA-DWELLER followed by the GMT BLNR and BLRO. The EXPLORER II contains the 'worst' caliber in this comparison, it has the lowest slope (m) and therefore rates change more quickly (to negative values) during the decrease of amplitudes.

I measured my Sea-Dweller very frequently (using an automized system) within the same PR scan, which explains the large quantity of data points.

I hope this post is understandable and triggers some discussion.
saxo3 is online now   Reply With Quote