What lift angle should I set on my timegrapher?

Use the lift angle the movement's maker prints in its technical document. For the movements most owners put on a timegrapher that is a short list, 44, 49, 50, 51 or 53 degrees depending on the caliber, and the chart below gives it for 37 of them. The figure that turns up most often in answers to this question, 52, is the factory default of many timing machines. Not one of the 37 calibers is published at it.

The setting matters for one number only. Rate and beat error come straight from the timing of the ticks and read the same whatever angle you enter. Amplitude is the number the machine has to calculate, and it calculates it from your setting, which makes it the reading owners use to decide whether a watch needs a service.

Lift angle by movement: the chart

Every figure below is printed in the maker's own document: ETA's technical communications, Sellita's technical documentation, the Seiko Instruments technical guides published through Time Module, and Miyota's specification sheets. Where one document or one value covers several calibers, they share a row.

Read by family, the chart is easy to remember. Seiko's NH movements are 53, a figure no Swiss caliber here matches; our NH35 page has the rest of its specification. ETA's 11½ ligne automatics split by family: the 2824-2 and its 28xx relatives at 50, the thinner 2892A2 and the 289x calibers at 51. Sellita follows the same split, with the SW200-1 at 50 and the slim SW300-1 at 51.

The chronographs are where the Swiss makers part company. ETA prints 49 for the Valjoux 7750 and its 7753 and 7760 siblings, while Sellita's SW510, the same 30.00 mm by 7.90 mm as a 7750, is published at 50. The large hand-wound 6497 and 6498 sit lowest of all at 44. Miyota divides by beat rate: the 3 Hz 8215 and the rest of the 82 series at 49, the 4 Hz 9015 and the 90 series at 51.

Why 52 keeps coming up

Because it is what an unconfigured machine shows. Weishi's manual for its No.1000 timegrapher says: "The lift angle is 52° for many common movements, therefore the lift angle value of instrument for factory is set to 52°." The sample screen in its No.5000 manual shows 52.0°. The same No.1000 manual goes on to say the value "must be selected according to the watch movement", and accepts anything from 30 to 70 degrees. A reading taken without changing the default carries 52 with it into every screenshot and forum post it reaches.

Against the makers' documents, the "many common movements" claim does not hold up for the calibers in this chart. The nearest are the 51 degree ETA and Sellita calibers, one degree away. The NH35, the movement most often asked about, sits one degree the other side: SII's technical guide revised in August 2012 and Time Module's 2025 version both print 53, so the answer to "is the NH35 lift angle 53?" is yes, in both.

The 7750 is the case where the gap is large enough to matter, and where search results most often repeat 52 against the maker's figure, the 49 our 7750 page takes from ETA. ETA's technical communication printed 49 degrees in its April 2010 revision and still prints 49 in the 2011 revision. In between, a revision dated 8 November 2010 added one word: the change log reads "Addition Typical 'angle lift'". That is the honest word. Lift angle is a design figure, set by the geometry of the pallet fork, roller and escape wheel, and an individual escapement varies around it. A setting of 52 on a 7750 is 3 degrees, or 6 percent, above it.

What a wrong setting does to the amplitude reading

A timing machine never sees the balance. It hears the escapement unlock and the impulse end, measures the milliseconds between those sounds, and works out how wide the balance must be swinging to cross the lift angle you gave it in that time. For a balance in harmonic motion the relationship is: amplitude = lift angle ÷ (2 × sin(π × lift time ÷ period)). The machine measures the lift time and knows the period from the beat rate. The lift angle is the only figure you supply, and it sits on top of the fraction, so a setting 6 percent high gives an amplitude 6 percent high.

Put numbers to it. A 7750 at 4 Hz swinging a real 280 degrees crosses its 49 degree lift in about 7.0 milliseconds. Tell the machine the angle is 52 and the same 7.0 milliseconds becomes 297 degrees. Nothing about the watch has changed.

The pattern is one-sided for the Swiss calibers. Every ETA, Sellita and Miyota figure in the chart is below 52, so the default flatters all of them, from 2 percent on a 2892A2 to 18 percent on a 6497, where a real 270 degrees appears as 319. A tired 7750 swinging 250 shows 265, which looks like a watch with years left in it.

The NH35 runs the other way, and at 52 its error is small, about 2 percent low. Its larger risk is the opposite mistake, setting an NH35 to a Swiss figure. At 49 a real 270 shows as 250, the reading that sends people to a watchmaker for a service they do not need, as our timegrapher reading explainer works through alongside what rate, amplitude and beat error each tell you.

Correcting a reading you already took

Because the error is proportional, any reading can be fixed after the event. Multiply the amplitude by the true lift angle and divide by the one that was set. A 7750 that showed 297 on the default: 297 × 49 ÷ 52 = 280. An NH35 that showed 250 at 49: 250 × 53 ÷ 49 = 270.

The same arithmetic applies to old notes and to the timegrapher photographs sellers post, provided the setting is visible on the screen. A pre-owned 7750 chronograph listed with 300 degrees and the screen showing 52 is really at about 283. If the screen does not show the setting, the amplitude in the photograph tells you nothing, which is the point our lift angle glossary entry makes about readings copied from forums. Rate and beat error need no correction either way.

Measure the way the maker does

The makers' documents also say how they take the reading the figure is meant for, and they do not agree with each other. Time Module's NH3 guide measures in three positions, dial up, 9 o'clock up and 6 o'clock up, within 10 to 60 minutes of a full wind, for at least 20 seconds, after letting the watch settle for at least 20 seconds in each new position, on a Witschi Watch Expert. All measurements are made with the calendar out of action.

Sellita's SW200-1 documentation recommends 20 seconds of stabilization and 40 of measuring, takes its full-wind checks 1 to 3 hours after winding, and makes all checks with the calendar not in function and the chronograph uncoupled. Its grade table puts amplitude between a maximum of 315 degrees dial up at full wind and a minimum of 200 degrees with 6 o'clock up after 24 hours, and notes that 95 percent of the pieces in a delivered lot must fall inside the limits. Those amplitude limits assume the 50 degree setting; on the default, a movement at the 315 ceiling reads 328.

At home, the useful part is the discipline: a full wind, a fixed wait, the same positions every time, and the setting written down with the result. Our guide to timing your own watch covers the bench routine and the seven-day wrist test, and the home accuracy test covers what a timegrapher can and cannot tell you about a watch on the wrist.

Movements not in the chart, and why

Seiko prints no lift angle for its own 4R35 and 4R36. Our 4R36 page explains that they are the same base design Time Module sells as the NH35A and NH36A, so 53 is the figure to use, taken from the NH document. For the 6R35, the Grand Seiko 9S65 and Rolex's 3235 we found no public maker document giving a lift angle, so they are not charted; Rolex does not publish even the 3235's diameter.

Omega's 8800 uses a co-axial escapement, which sounds different to a microphone from a Swiss lever, and no lever chart applies to it. ETA's Powermatic 80 is left out because its lift angle is not in a public ETA document we could retrieve; do not borrow the 2824-2's 50 for it. And two movement types have no lift angle at all: Spring Drive regulates a glide wheel against quartz rather than swinging a balance through a lever, and Eco-Drive is quartz.

When this chart stops applying

• The brand has changed the escapement. A caliber sold under a house name, built on an ETA or Sellita base with its own escapement parts, takes the brand's figure, not the base caliber's.
• The figure is "typical". ETA says so for the 7750 family and the 2896. A single movement can sit a little either side, which is why watchmakers who need the absolute number measure the swing optically and adjust the setting until the machine agrees.
• You only compare a watch with itself. A wrong but constant setting still shows a falling trend correctly. It is the absolute number, and any comparison with someone else's reading, that the setting decides.

Once the amplitude is right, the next question is what to do about the rate, which our answer on whether a watch can be regulated covers.