Regulating Regmariner 2813

[hermesgroove 0]

Hey guys,

 

I bought a submariner from reg and it looks and feels great. It looses about 2 mins per day though. I bought one of these rubber ball case openers and tried to adjust the 21j movement myself. I didn't find a ton of guidance online so I wanted to ask you for some advice.

 

There is two levers, one with one dot and the the second with two dots. I moved both levers slightly to the left. Now I think my watch is quite a bit too fast. It has only been half an hour but it seems to have gained already some seconds. Can really only tell after a day I guess. Should I have left the lever with one dot alone and only more the two dot lever?

What's the difference between the two? Thanks for any help!

Edited August 19, 2017 by hermesgroove

[Diver Dave 0]

First,  you need a regulation instrument of some sort. If you use an iPhone, "Hairspring" is one that works. These systems "listen" to the watch so you can see what you're doing in real time. 

 

The lever with one dot is the one you use to get the beat "regular", meaning that each beat is spaced correctly compared to the others. It's the Isocronism lever. Step one is to get the watch beating correctly using this. 

The one with two dots is the rate lever. Second step is to use this to adjust the rate. Use a magnifier and a toothpick. If you can see it move with the naked eye, you have moved it way too far. 

Temperature, mainspring state, and position all affect accuracy as well. Tossing a watch randomly on a nightstand at night is a guaranteed surefire way of making a watch run irregularly. 

I wrote the following for buyers of my custom watches, which use Seiko movements. The information is identical for all other mechanical watches as well. It might be of interest:

 

++++++++++++++++++++++

 

Mechanical watch accuracy:

A mechanical watch is a precision machine, but it is after all simply a machine. Think of it as nothing but a spring that is allowed to unwind at a very specific rate, with us being able to watch it unwind. The self-winding mechanism is re-winding the spring constantly. Machines are affected by temperature changes causing expansion and contraction of metal, lubrication, tolerances of fit, friction, and gravity. Each one of these things affects the RATE of a watch. There is a difference between PRECISION and ACURACY, so let’s examine these in a bit of detail.

RATE: This is the daily change from “perfection”. If a watch has a rate of +2 seconds a day, that means it’s running fast by that amount. On a ship, where mechanical chronometers were used for navigation, a log was kept and daily an entry was made to calculate REAL TIME as a function of CLOCK TIME corrected for RATE. It was not unusual for the ships chronometers (three were carried) to be off by several hours, but by knowing WHEN it had been set at the nautical observatory and what the daily rate was, accurate time could be derived by simply adding or subtracting the daily rate multiplied by the number of days since the chronometer had been set.

PRECISION is the ability of the watch to HOLD a rate over a long period of time. If a watch has a variable rate, you cannot calculate it and take it into account. PRECISION is more important than accuracy. ACCURACY is the ability of the watch to actually read true time by looking at the dial. It’s not as important as precision: After all a PRECISE watch may never show the actual time, yet a stopped clock is perfectly accurate twice a day!

What would be perfect would be a watch with zero rate, which maintains that over a wide variety or temperatures and positions, which has been set perfectly. That’s not something possible with a mechanical watch. More likely your watch will settle into a predictable rate, and that’s all we can ask for. If you know your watch runs two minutes fast a week, reliably, then make it a habit to accurately re-set it once a week on Sunday and know that it’ll be one minute off on Wednesday. That’s the use of a reasonable RATE, relying on PRECISION (predictable rate) to know the time ACCURATELY.

There are a variety of factors that affect rate. First, do not expect a settled rate for a few weeks after you start wearing the watch. It takes time for the mechanical bits to settle in, for lubricants to spread themselves evenly, and for the watch to simply break in. Do not be surprised if the rate varies during this time. It’ll settle out after a period, and begin its decades long career of beating 21,600 times per hour for many decades. Yes, the watch movement steps TWENTY ONE THOUSAND, SIX HUNDRED times per hour, year after year after year. Think about that for a minute.

You may have seen a watch described as “Adjusted in 5 positions” or “Adjusted at three temperatures”. What does this mean, and why is it important to us?

POSITION: The watch is a machine. Gravity affects the rotating masses, and every watch will run at a slightly different rate in different positions. It may run slightly fast when laid on your table face up, slower when face down, faster when sitting crown-up, etc.  Watchmakers aren’t dumb, and all good watches are REGULATED (internally adjusted) to be most accurate in an average position when worn on the left wrist. That position is an average of crown-down (imagine walking), and 12-down (think of sitting with your arms on a table). In other positions the watch may not be as well regulated. A knowledgeable left-handed watch wearer of taste and wealth, buying a $100K custom watch, will let the watchmaker know he is left handed and the regulation will be done accordingly. Why is this important to us? Because if we are someone who takes our watch off at night and lays it on a table, the position you lay it in will affect the rate. If you do it the same time, the rate average over 24 hours will be the same. If you randomly toss it on the table each night face up, face down, and on its various sides, you will have unpredictable results and may think the rate is not precise. The watch will be most accurate when it’s worn continuously. The randomization of movement, even when you are sleeping, will always be averaged out and the watch will keep reliable time.

TEMPERATURE: Similarly, there are a number of reasons that a watch will change its speed when its temperature is varied. All of the metal elements change their dimension slightly, and lubricants change viscosity. Temperature changes adversely affect the all-important rate of the watch. Keeping the watch at one temperature is best. The easiest way to keep it at constant temperature is to WEAR IT CONSTANTLY. Your skin will keep the entire thing close to a constant temperature, and this also assists reducing position error. Take it off for the shower so you do not heat it up more than needed, also knowing that keeping soap out of the rubber seals will increase their lifetime.

 

The best advice I can give to you is to simply put the watch on, and wear it all of the time.

By doing this you will:

 (A): Keep the mainspring fully wound (rate is affected at different mainspring tensions),

(B): Randomize the position, and lastly,

(C): Maintain nearly constant temperature.

 

What rate/precision/accuracy to expect?

Once the watch settles down and is broken in, day to day accuracy should be excellent. The Seiko 7S26 movement is regarded as one that is extremely reliable, resistant to shock, and able to run accurately for many years without service. Many watches using the predecessor to this movement were purchased by soldiers at military exchanges during the Vietnam War, and are still running well today. If you allow the watch to break in, keep it fully wound, randomize the positions it’s used in (or store it at night the same way every night), and keep it in a fairly constant temperature environment, it’s not unlikely that it will keep time +/- 2 minutes per week. Learn the rate, correct for it during the week, and re-set weekly.

If your watch is running a bit fast or slow, you can “cheat” it by selectively changing how you lay it down at night, assuming you do not wear it, as follows:

Face Up = Gain a few seconds per night (if your watch is slow)

Crown down = lose a few seconds per night (if your watch is fast)

Crown up = Lose rather more seconds per night (if your watch is too fast)

My advice? Put it on, wear it, and only take it off for hot showers.

 

 

 

 

Edited August 20, 2017 by Diver Dave

[tabuktime 54]

Wise words indeed Dave. 

[hermesgroove 0]

Very nice write up, thanks! I found watch check for android and will give it a try.

[nofitstate 27]

Have a look a clock tuner too in the play store. Found this to be the better of apps for timing watches.

 

[black263 228]

Problem now is that having moved the lever with two dots, you have no way of checking if it is correct without the use of a timegrapher or similar, which shows beat error.  As a basic rule, never touch that one without such a machine.  The one with a single dot is much more forgiving.

 

Edit.  Note to self.  Don't post pissed.  As has been pointed out it's the other way round.  Don't fiddle with single dot without a timegrapher.  Two dots can be adjusted by trial and error.

[Nikosaldente 630]

36 minutes ago, black263 said:

Problem now is that having moved the lever with two dots, you have no way of checking if it is correct without the use of a timegrapher or similar, which shows beat error.  As a basic rule, never touch that one without such a machine.  The one with a single dot is much more forgiving.

It's the other way around!

The one dotted lever should never be touched! It's the isochronism lever.

The two dotted is the regulator and is more forgiving!

Moving it towards the one dotted lever, makes the watch slower, whereas moving it away makes the watch faster 

[black263 228]

On 20/08/2017 at 21:55, Nikosaldente said:

It's the other way around!

The one dotted lever should never be touched! It's the isochronism lever.

The two dotted is the regulator and is more forgiving!

Moving it towards the one dotted lever, makes the watch slower, whereas moving it away makes the watch faster 

You're absolutely right.  Shouldn't post advice when pissed.

Fact remains that having moved both, the OP now needs a machine to get it right again.

 

 

[Diver Dave 0]

18 hours ago, hermesgroove said:

Very nice write up, thanks! I found watch check for android and will give it a try.

 

There are other (better) software systems as well for laptops, where you can either use the laptops built in microphone or an external microphone to provide the acoustical input to the software. The isochronsm adjustment is one where you want to see a fairly long "span of time", to look for any variation in the beat *over a period if time* (like 30 seconds to a minute).  That's a bit hard to do on a smartphone. You can do it... it's just harder on a smaller screen. Do a little Google and you might be surprised at what's there for free. 

Any person working with watches as more than a casual hobby will have a Timegrapher, which is a machine that is designed to do movement analysis as it's routine job, They can show a number of different statistics about a movement, and are used when doing regulation and adjustments. I use mine daily and it's paid it's way back 1000 times. 

If you're CONUS and want to have me stick it on the machine for adjustment as a favor, let me know. I take enough value from the forums to want to pay things back by paying a little forward, so consider it a favor. I'm not a TD so if I sell your watch and retire to the islands on the proceeds, you're fucked. I haven't done that yet though. ;-)

 

DD

 

 

 

Edited August 21, 2017 by Diver Dave

[hermesgroove 0]

Haha that's a great offer and I'm CONUS yes! Let me try myself first, sounds like fun to figure this out. But if I'm getting nowhere don't be surprised if I ask you for your help 

Edited August 23, 2017 by hermesgroove

[Diver Dave 0]

28 minutes ago, hermesgroove said:

Haha that's a great offer and I'm CONUS yes! Let me try myself first, sounds like fun to figure this out. But if I'm getting nowhere don't be surprised if I ask you for your help 

 

Or grab one of these: (every watch enthusiast should have one)

http://www.ebay.com/itm/Watch-Pocket-Watch-Timing-Machine-Multi-function-Timegrapher-1000-by-ACE-TIMER-/191216929225?hash=item2c856ac9c9:g:cRkAAOxy-WxTBHgb

[hermesgroove 0]

Just tried clock tuner and I must say it's a great recommendation!

It auto detected the 21,6k beats and showed immediately how much the watch was off which was consistent with what I experienced. After 1 min it got a bit more precise but wasn't much off from the initial measurement.

After trial and error I got it to +10 seconds, took me 6 tries. So far I only tested it four hours in and it's keeping time perfectly!

 

Very happy with this free fix, thanks guys!!

[Diver Dave 0]

^^ Bravo! Well done. Once you have it basically dialed in, try it in several positions and see how it changes (or doesn't), and then wear it for a few hours to get it temperature stabilized (and fully wound) and take another look at it. You'll enjoy the process. 

 

[hermesgroove 0]

Thanks for all your insights! It's still losing a bit of time in 24 hours but I'm sure I can get it closer now once I have a bit more data over the next couple of days.

[Leslie 62]

On 2017/8/20 at 08:52, Diver Dave said:

If your watch is running a bit fast or slow, you can “cheat” it by selectively changing how you lay it down at night, assuming you do not wear it, as follows:

Face Up = Gain a few seconds per night (if your watch is slow)

Crown down = lose a few seconds per night (if your watch is fast)

Crown up = Lose rather more seconds per night (if your watch is too fast)

My advice? Put it on, wear it, and only take it off for hot showers.

Is this true for each and every SEIKO 7S26 movements?

Is it also true for other movements?

Thanks in advance.

[Diver Dave 0]

15 hours ago, Leslie said:

Is this true for each and every SEIKO 7S26 movements?

Is it also true for other movements?

Thanks in advance.

 

 

It's good advice for mechanical movements in General. Rolex in fact included this advice for years when they sold actual tool-watches to working guys who were advised to cheat their watches into better accuracy by doing this. at night. You can Google for the old Rolex brochure with the information. I basically stole it from there and included it in my write-up. 

 

Dave

Edited August 24, 2017 by Diver Dave

[Leslie 62]

Thanks, Dave, for the information on the regulating by different resting positions.
I still have some doubts, and hope you or someone can kindly help me.

I have seen the Rolex's "Simple Regulating" before:

After examining the placement of the balance wheel and escapement wheel, 
I can guess why resting positions have different rates -- the gravity/friction on balance/pivot are different for various resting positions.
And, I guess all movements with the same relative position of the balance/escapement (or the pallet lever) should behaves the same.
But, for the movements with different allocation (say, Unitas 6325 that has its balance unit behind the 9 o'clock dial position, or Seiko 7S26 with its balance unit behind the 10 o'clock; while the Rolex/ETA 2824 has the balance unit behind the 12 o'clock), the effect of vertical resting positions should be different.
Is what I said correct or not?

Of so many watches I have, most run fastest on the flat (dial facing up or down) position, which is reasonable.
But for vertical positions, I failed to get a regular pattern, even for watches with the same kind of movement.
Maybe these watches (made in the 70-80's) were mostly not well-adjusted during their lives, and so there are other factors affecting the running rate.

[Diver Dave 0]

^^ Boy, I dunno. A micro-mechanics engineer can probably solve the riddle. Me? I'd just play with what you have and learn its nature. 

 

Dave