Every second counts. Or does it?

Last week, I got an email from a watch owner who was fairly disturbed. He took his Tudor watch to a Tudor service centre, and they refused to help. He called later that day. "What was the problem, and why could Tudor not help?" I’ve asked. The watch was losing time. Precisely half a second per day. The caller was annoyed with the fact that the maker dismissed his complaint by simply stating that the timekeeping is well within the chronometer specification for a five year old mechanical automatic watch. And I could not agree more. Actually, I pointed out that he should consider himself lucky - the timekeeping is absolutely perfect, and there is no need for adjustment. He disagreed. "I would be much happier if the watch was gaining one second per day rather than losing half." Let me put things into perspective. It is like a person earning $46,200 per day asking their boss for a pay rise of 50 cents. I am not going to bore you with details, but the longer we talked, the more determined he was that he was 'in the right' and that I should do my absolute best to solve his problem. Thirty years ago I would have ended the conversation by simply hanging up. Twenty years ago, I would have made a sarcastic remark, and then hung up. But nowadays, I am a different man: a mellow, woke wimp. "Sure, bring it in, I'll fix it". The watch arrived on my bench yesterday morning. A Tudor Pelagos with an in house movement. Assessment The first step, prior to any adjustment was to place the watch on auto winder, and let it settle at room temperature for 24 hours. After that, the watch was placed on the timing machine, in various positions. The goal was to establish whether the customer’s observation matched the actual timekeeping recorded by the timegrapher. The watch showed a constant rate in all positions of +0 seconds per day, and a heathy amplitude. There was no deviation except for dial down, showing -1 second per day. However this value is within the tolerance and resolution error of the timegrapher.

Pre-adjustment The watch strap was detached, case cleaned of gunk, rinsed and dried. The watch was then left to recover for another hour at room temperature.

Adjustment In order to gain access to the regulator, the watch case had to be removed. Tudor's new calibre MT5612 has a free-sprung balance with four adjustment screws located radially to the balance wheel, and not accessible from the top of the movement. The mechanism had to be extracted from the case.

The adjustment screws were factory set in position so that the balance wheel is poised perfectly. Theoretically, the time adjustment would require changing the position of at least two screws to prevent a degradation of poise. Also, the screws were machined with custom-design heads so a custom design (in my case modified) regulating tool is required. The first step was to experimentally determine the amount of rotation of each screw that would correlate to an adjustment of 1 second, which was approximately 15 degrees. After that, the watch was re-adjusted to +0, then two opposite regulating screws were adjusted (by moving them in towards the centre) by approximately 7 degrees. Such a fine adjustment is done purely by ‘feeling’. After the adjustment, the timegrapher read a daily error of +1 to +2 seconds.

Post-adjustment The movement was returned to it's case, secured, and once again, the watch was placed on the timing machine. Often, compressing the movement into the case could cause variation to the adjusted value. However in the case of this Tudor mechanism, the rate remained the same, with only a slight improvement in amplitude. The watch was then water pressure tested. Time keeping verification The time was set to mobile phone data network, and placed on the winder for the next seven days when the actual timekeeping error will be determined. The expectation is that the watch will keep time at a rate of around +1 second per day. Now, if you think that the purpose of this write up is to impress you with my determination to showcase the ability to squeeze a 1 second improvement out of an already perfectly factory tuned watch, then you are grossly mistaken. I just need to hear from you how much should I charge for all this work, consisting of 45 minutes behind the bench, 20 minutes of email and phone consultations, 15 minutes of in person consultation, the admin work related to record keeping and for providing insurance coverage on a watch left in my care? And was it worth it? You tell me.