I bought a fairly cheap G shock 5610. Looks like the watch I had in college. Solar powered and atomic timing. My automatics just sit there dead because as pretty as they are they suck at time keeping. I can’t rationalize the disparity between cost and actual performance. This cheap ass Casio is dead nuts on every day. Day after day.

I love the nice watches but I’m becoming too pragmatic as I get old.

Originally posted by pedropcola:

quote:

I bought a fairly cheap G shock 5610. Looks like the watch I had in college. Solar powered and atomic timing. My automatics just sit there dead because as pretty as they are they suck at time keeping. I can’t rationalize the disparity between cost and actual performance. This cheap ass Casio is dead nuts on every day. Day after day.

I love the nice watches but I’m becoming too pragmatic as I get old.

I get where you’re coming from. Having an automatic watch that is more than say +\- 15 seconds a day does not provide me much joy of ownership.

Now on the flip side of that there are a few automatic watches I can wear for a week or two and still be within 4-5 seconds from atomic time if not dead on.

It’s sort of like owning a Honda Civic and a Ferrari 360 Modena Spider with a 6 speed manual. They will both get you from point A to point B, but one will do it with boring reliability while the other does it while also moving the soul. There is a certain engineering mastery and artful element of excitement and passion that you either love and appreciate or you don’t.

For a long time I didn’t appreciate it myself.



Konata88, any update on the Orient? Did your watchmaker manage to get it dialed in any better?

This message has been edited. Last edited by: stickman428,

I'm still somewhat characterizing after the second visit (in which he demagnetized only).

This is what I think I'm seeing:
1. Watch, on the average, is fast by about 12-15 sec per day.
2. It gains more during the day but loses some over night so that the net is about +12-15 per day.
3. I believe that it gains more over the course of the day while I'm wearing the watch. For example, it might gain 2 sec per hour for the first few hours then about 5 sec after wearing the watch for the next few hours. Not sure if it's because the spring is getting wound or because of the average watch orientation (goes back to some of my expected behavior questions above).
4. Not sure and maybe this supersedes speculation above - watch may gain more time in the afternoon / evening regardless. Like there is something mechanically wrong (as suggested by Mars?) like a burr or something that leads to increase in rate in the afternoons.

Against prior recommendations and at the cost of perhaps investing in a new watch, I've invested in some tools to help me better understand the Kanno. It has value as my first automatic.

I have coming a case wrench and a 1900 timegrapher.

I want to characterize the Kanno further:
1. Rate for each of the 6 primary orientations
2. Rate for various states of wind (amplitude)
3. Rate for various period in the day.

One thing I'm realizing is that a rate error of zero doesn't really mean that the watch will be accurate while wearing - only for that orientation. I need to figure out what the net is for a typical day for me. And then figure out how to adjust the rate to optimize for a typical day.

After a week or so of measurements, I think I'll be ready to try my first attempt to regulate.

You’ve already answered most of your own question.

Does it gain 12–15 seconds each day, on average? That is “within spec” for your movement but I understand if that’s not satisfactory (I wouldn’t be either).

If it consistently gains roughly that per day, then you or a watchmaker can open it open, regulate it a bit slower, close it up, and see how it does with its net daily rate. You should be able to drop that number closer to 0.

I’m wearing a Seiko 5 (7S26) as I type this. On the wrist for 14 hours a day or so, it nets under two seconds a day. I put it crown up when off the wrist. When it gets too far ahead, I leave it crown up an extra day off the wrist (example, I’ll take it off at 8:00 pm Friday, and if I need it to lose time, I won’t put it back on till Sunday morning). In that extra day of crown up off the wrist, it’ll lose 12–14 seconds. Doing it this way I can wear it most days (like 8/9 or more) and keep it within 5–10 seconds of zero all the time.

No regulation needed…just playing with positional variance.

Woohooooo. Timegrapher arrived. Feels like Christmas.

Reading the manual and I’ll try my regulating attempt later today. It may not matter but I’ll regulate in the afternoon when I suspect it’s running faster and also in the dial up position. Other choices were 6 o’clock up or crown down positions.

I've had one watch adjusted by a watchsmith, it was the '80s and they used a timer to check the beat - it was less than a 5 min procedure.

I'm sure the OP can do a lot better over a few weeks.

Here's a thought - at least with the quartz models, it would seem we could do the same with an external interface, ie, you tell the watch how many seconds fast or slow every 30 days and at the end of the year it should be spot on. With an auto, it would be potentially a simple reduction gear train allowing extremely fine adjustments. Planetary gearset might be one way.

Which then goes to why not already?

Okay, I took readings in the six positions, before winding and after winding. Rate, Amp and BE doesn't really seem to change based on winding status.

Here's the funny thing: rate for all 6 positions are reading between -1 to -16 although I'm net positive about 12-15 sec/day. I'm going to guess that there is an offset in measurement error and still adjust down by about 10 sec. Otherwise, I'm not sure how I'm ending up fast per day unless there is a period of time where the watch speeds up for some reason - period in the day other than right now.

I'm also going to assume that regulating one position affects all other positions equally. So, I'm just going to pick one position and go from there. If I should pick a specific position in which to regulate, I'll need more data to decide upon which position to select.

ETA: why do they make the regulate lever so sensitive? Do watches really need the range afforded by the regulation lever? Accuracy of adjustment is in the tens of seconds. I needed to go back and forth until I got close to my target regulated level.

So, net is that I've slowed the rate by about 6 sec/day based on the dial down orientation. We'll see where that lands me in real daily usage.

There must be some trick to moving the lever more precisely and deterministically. How do they do at the factory? How do watch makers do it?

What little I've adjusted the lever was sensitive and the results were always a lot more than I anticipated. As for the timegrapher showing one thing but the watch acting the opposite it goes to the orientation - that's why the Swiss put it in 6 different positions over 45 days. Up vs Down vs Stem vs etc each has it's slightly different resting speed and its an overall adjustment which is accomplished. Then the owner buckles up and voila, you get a wide range of motion, plus some body heat - 20F+ warmer - and some report another adjustment comes into play.

It seems all watches are prone to this in some way, altho those with minimal gear trains and a solenoid have less mechanical issue. That's where some Quartz have an adjustable coil to regulate voltage (or something) and you minimally turn it.

There is also the issue of the watch varying somewhat irregularly over a longer period of time - a snapshot view of an hour or so will give you just that, over a day or more it could go up or down +/- 5 to 10 easily, seemingly at random. That cause may be variable friction in the drivetrain due to ? Interleaf friction in the mainspring, geartrain, a microscopic burr on an hour wheel that only affects a few minutes at a time yet the rate takes a dump then catches back up. The adjustment has it fixed overall, the actual speed runs slightly fast then stupid slow then slightly fast leaving the overall almost correct. That would cause problems timing runners in a marathon where they might cross the line for hours. Hence, Rolex went to quartz to maintain more precision in Olympic timing.

A high end watch company put a lot of time and effort into 1 (one) show watch with all the bells and whistles in friction, temp and production finesse and got a huge increase in runtime from one winding. Retail it was over $25,000 with all leading edge tech not in production IIRC. That was a very interesting read which got me wondering why we don't have watches with an external method of correction for the balance wheel. It would be a selling point for automatics and with "Black Thumb" bombproofing an owner wouldn't be prone to literally screwing it up. Unlike stems, which rank in the top three watch stoppers.

We buy watches to impress ourselves, I'd be pretty impressed owning a watch running +1 a year with no 2d party correction. I'd probably also be insufferable.

Interesting that Rate can be modulated by temp (body temp). Makes sense conceptually - one wonders on the sensitivity (ie - Rate changes by how much for a given degree?).

I've been speculating about time of day and mechanical aberrations and the influence that may have on daily error - again, I think late afternoon may be a time period of interest. Need more data. But pragmatically may not be necessary, only academically. One would still try to average it out during regulation.

The watch kept to zero error until I took it off (about 5 hours later). Face down, it lost about 5 seconds overnight. We'll see what the net is by end of day today. Seems like, using orientation overnight, I can lose 2-10 seconds. As long as net during the day is less than that, I think the watch can be fairly accurate.

I would LOVE an automatic that has an external dial for regulation. I guess most wouldn't need it - expensive watches and COSC, whatever. But for those of us on a budget, I think it would be a great feature to dial in accuracy post-sale.

Did you set the lift angle correctly for your movement?

There was one watch that had external access for regulation. A funky-looking thing on the 9 o’clock side opposite the crown. I’m trying to remember who made it...Bulova maybe?

quote:

Originally posted by Sir Guy:
Did you set the lift angle correctly for your movement?

I think I read somewhere (hard to find) that it is 52 degrees. So I left the tool on default setting.

Very possibly, my knowledge of older American models is limited. Bulova was a big innovator in the day, and they imported a Japanese brand, too, Citizen, who bought them in 2008.

I zero'ed the watch about 2 days ago after regulating.

So, after about 1.5 days which includes 2 nights:

1. lost 5 sec over night (face up)
2. gained 17 sec while wearing
3. lost 20 sec over night (noon high)

So, when I woke up, I was net -3 sec. I presume that I'll gain 17 again today (net 14 sec). Then can lose 15 sec overnight (using a different orientation).

If so, daily swings of +- 15 sec but averaging around zero (+- several sec). If so, call it good?

quote:

Originally posted by konata88:

quote:

Originally posted by Sir Guy:
Did you set the lift angle correctly for your movement?

I think I read somewhere (hard to find) that it is 52 degrees. So I left the tool on default setting.

Do you know what movement is in the watch. This is one of the websites I use for lift angle info.
https://watchguy.co.uk/cgi-bin/lift_angles

I have an Orient Kanno with F6922

I couldn’t find it in the list.

quote:

Originally posted by konata88:
I have an Orient Kanno with F6922

I couldn’t find it in the list.

This another sight I use.
https://calibercorner.com/orient-caliber-f6922/
All tho they are not listing it specifically in the technical info some one in the comments thinks it is 52 degrees.

Thanks. Yea, that was my basis for assuming 52 degrees. I read somewhere that measurements weren’t sensitive to a degree or two.

Okay - here's my experience after a few days.

I think I gain about 8-15 seconds while wearing the watch.

I can lose about 2-15 seconds overnight depending on orientation.

So, at this point, after a few days of monitoring, I think the watch is oscillating up to +15sec during the day and up to -15sec at night. Effectively, I'm netting close to zero each day.

If this persists, I don't think I'll need to adjust the time on the watch anymore (I've been adjusting it weekly to accommodate the net +10sec it had been gaining daily).

To me, this is amazing for a device that is based on mechanics only.

The timegrapher was expensive but worthwhile purchase. If only I had more watches.

I know I got lucky with the regulation - the lever is way to sensitive to be practical. But I think it's in a sweetspot now.

Thanks again Stickman, for the watch and for all the advice and information. And to all others who helped me here as well. It's been an interesting experience and education.