M8. Positions Test Figures

What It Is

A watch runs at a slightly different rate in each position, because gravity acts on the balance differently in each. A positions test measures it in several positions and sums the differences in a few figures. The Positions Test pane (chapter 20) gives them, and so does timegrapher session. They follow the figures Witschi’s timegraphers use and the Swiss standard NIHS 93-10 (NIHS: Normes de l’Industrie Horlogère Suisse, the Swiss watch industry’s standards; the project’s math standards list them).

The Positions

CommonWitschiGroup
DU, dial upCHhorizontal
DD, dial downCBhorizontal
CD, crown down9Hvertical
CL, crown left6Hvertical
CU, crown up3Hvertical
CR, crown right12Hvertical

Witschi’s codes count round the dial with the crown at 3 o’clock, so crown up is 3H and crown right 12H.

How It Is Worked Out

Each position is read as timegrapher session reads a stretch: its measured seconds, after its settle, with any masked time left out. A position measured more than once is read together, or averaged first in the indices. A position needs one whole 10 s reading to count.

In one state of wind (all positions at the same hours since full wind), with r each position’s rate and a its amplitude:

X     = mean of r over the positions read            the Average row
XH    = mean of r over the horizontal positions
XV    = mean of r over the vertical positions
D     = largest r − smallest r                       the Largest Difference row
DV,DH = the same among the vertical or horizontal positions
DVH   = XV − XH, and the same for amplitude          vertical less horizontal
Di    = r(6H) − r(CH)                                crown left less dial up

The Largest Difference row gives D for the rate, and the same for the amplitude and the beat error (as a size).

Across two states of wind, full wind and 24 hours after, with each position’s change Δr = r(24 h) − r(full):

Im    = the Δr largest in size, crown right (12H) left out, signed
Im*   = the same over all positions             command line or MCP only
Ie    = | X(24 h) − X(full) |, each X over the positions read in both

Why This Way

  • Means over positions: a watch on the wrist turns through many positions, so the mean rate X is the nearest single figure to its rate worn.
  • The largest difference D: how much the rate depends on position at all. A poised balance with a well-set hairspring gives a small D.
  • Vertical against horizontal, DVH: in the horizontal positions the balance pivots rest on their ends; in the vertical ones on their sides, with more friction and a lower amplitude. DVH sums that split.
  • Crown right left out of Im: as NIHS 93-10 takes it. The Positions Test pane shows Im and Ie; the command line’s session and the MCP server give Im* as well.
  • The same positions in both states for Ie, so a position measured only once can’t move the change in X.

When You’d Use This

  • Regulating: X says where the watch runs on average; D and DVH say how much the positions differ before you adjust the rate.
  • After service: compare the six positions and D with the figures from before.
  • Full wind against 24 hours: Im and Ie say how much the rate moves as the mainspring lets down over a day, in each position and on average.

What It Helps Find

  • A poise error: it can show as one vertical position far from the others.
  • A large horizontal-vertical split (DVH): the balance pivots or jewels.
  • A rate that runs away at 24 hours (large Im or Ie): isochronism (M12).

What It Can’t Tell You

  • Whether the figures pass. The app gives no tolerance, verdict or pass and fail. Compare with the maker’s or the standard’s limits yourself.
  • The cause. A large D or DVH says the positions differ, not why.
  • Precision beyond the readings. Each position is one reading, with its own ± (M7); a short measurement in a noisy position gives a wide one.

Previous: M7. The ± and Correlated Readings · Next: M9. Steadiness Statistics

This page comes from the app's user guide at v1.0.1. See it on GitHub.