Budget Watch Collecting/Regulating

From Wikibooks, open books for an open world
Jump to: navigation, search

All you ever wanted to know about regulating your watch yourself...and why you wished you hadn't asked...

I’ve noticed a lot of posts lately from people wanting to know how to regulate a Seiko 7S26 movement, and whilst there have been various posts in the past on this topic, there’s nothing tying it all neatly together in one post. So, having got me a Vibrograf timing machine recently (as if you didn’t know), I thought I’d cover the basics and also demonstrate how the machine works, and what it can do.


First, the basics.

1. Regulating a mechanical watch means making it run slightly faster or slower, by making a small change in the effective length of the balance hairspring. The shorter the effective length, the quicker the balance completes its swing and so the watch runs faster. If the effective length is made greater, the opposite occurs. This change is made by moving the regulating lever, marked ‘A’ in the photo below. Note that there is also a beat adjustment lever, marked ‘B’ in the photo. This lever should NOT be touched, unless you suspect the movement is ‘out of beat’ and you have some sort of instrument such as a watch timer which will allow you to see when it is ‘in beat’ again. I will explain these terms more fully later on.

Watchregulating177f.jpg


Note that regulation is not the same thing as adjustment. To adjust a watch is to compensate for all the various factors that can affect timekeeping. This can be a complex and long-winded process, and is best left to professionals. Regulation is performed as the final ‘tweaking’ after adjustment has been carried out. As far as we are concerned in this post, we shall assume no adjustment is necessary, only regulation.

2. Unless a watch is in good mechanical condition and adjustment, it’s really a waste of time trying to regulate it within a dozen or so seconds per day. By ‘good mechanical condition’ I mean the movement is free of dirt and is adequately lubricated with clean oil, there are no excessively worn bearings, the mainspring can supply a steady flow of power to the gear train, there are no chipped or missing gear teeth or pinions, etc. Similarly, it’s very optimistic to expect a cheap, low jewel-count movement to perform well on a regular basis. However, even an entry-level movement such as the Seiko 7S26 can produce astonishing results, as we shall see.

3. It’s important to realize that there are many factors that conspire against consistent timekeeping when a watch is worn. These include change in position (e.g. dial up, crown down, crown up), change in temperature, change in mainspring state-of-wind, and personal wearing habits of the owner (e.g. worn 24/7 or taken off at night, what position is it left in at night, shocks and bumps from sports activity, etc.). Whilst modern materials have reduced the effects of temperature and mainspring state-of-wind (also known as isochronism from the Greek meaning ‘equal time’), they still come into play, along with ever-present gravity. Thus, in regulating a movement we can only try to compensate for all the ‘errors’ caused by these factors on an overall basis. It’s important then that before attempting any regulation, the watch must be observed over several days of normal wear to establish how much it gains or loses each day, or its daily rate. It is this rate that is the starting point for regulation.


The Vibrograf machine

The Vibrograf was invented and developed in La Chaux-de-Fonds, Switzerland, about 45 years ago by Portescap, the creators of the Incabloc shock absorber. Various models have been produced over the years, and because of its simplicity and reliability the B200 became the watchmaker’s standard workhorse. Although now superseded by more sophisticated (and expensive) machines such as those made by Witschi, the B200 is still more than adequate for simple diagnosis and regulation. In addition to mechanical movements it can, with the appropriate optional circuits, also be used for electro-mechanical, tuning fork and quartz clock movements.


How it works

The basic components of the B200 are an inductive pick-up mounted on a multi-positional arm; an amplifier; a quartz oscillator giving an extremely stable time base; a paper tape printer; and a read-out scale. The pick-up (which is basically a microphone) conveys the sound of the watch movement to the amplifier, and it is then compared to the time base. The printer then records the difference between the two frequencies onto a moving paper tape. The resulting linear trace (called a vibrogram) can then be analysed to determine the watch’s instantaneous rate and also a variety of defects such as beat irregularity, damaged escape-wheel teeth, out of poise balance, wrong type of mainspring, damaged pallet stone, dirty escapement, etc.


BWCVib200 pickup.jpg

Multi-positional inductive pickup

BwcVibrograf1162.jpg


How it’s used

First, the beat frequency of the movement to be tested (e.g. 21,600 bph) is selected by means of the black push-buttons on the front of the machine. Almost any standard frequency between 3,600 and 36,000 bph can be selected. Next, the watch is fixed (dial up or dial down) to the inductive pick-up by means of the built-in spring-loaded clamp. (For watches with thick cases or plastic movement rings, better results are achieved if the crown is in contact with one of the prongs of the clamp.) The amplifier is then switched on by means of the round black knob on the front panel. This knob is also the volume control. Nothing will be heard at this stage, unless headphones are plugged into the back of the machine.

To start the paper tape recorder the red push-button is depressed, and the volume knob is rotated until the sound of the movement ‘ticking’ can be heard. At that point a linear trace will appear on the paper. There are actually three major sounds associated with the escapement, but the amplifier is designed to use the most distinct one, which is the sound of the impulse pin on the balance hitting the notch of the lever fork. This is also known as the ‘unlocking’. When enough trace has been produced the large black knob on the sloping face of the machine is rotated so that the engraved lines on the clear plastic reading disc (to which it is geared) are parallel to the trace. The instantaneous rate can then be read directly from the dial under the large black knob. In the photo the rate is +18 sec/day in the dial up position.

The above operation is repeated with the watch in various positions (e.g. dial down, crown down), to see what the positional variations are. A difference of about 30sec/day between horizontal and vertical positions can be considered as ‘normal’ in an average-quality movement. If the positional differences are really large it can indicate basic faults within the movement, which will work against consistent timekeeping.


Interpreting the results

The trace produced by the machine is usually a single continuous straight line that may slope to the left or right, depending on whether the rate is losing or gaining. Sometimes the trace shows two parallel lines a small distance apart. This indicates the movement is ‘out of beat’, which means that the length of arc the balance swings through in one direction from the midpoint is not the same as in the other. Put simply, the length of the ‘tick’ is not equal to the length of the ‘tock’. Ideally, to achieve the best regulation possible the movement should be put ‘into beat’ by moving the beat lever (sometimes called the stud carrier), so that the balance swings through an equal arc on both sides. The effect of moving this lever is easy to see on the trace, but hard to distinguish without a machine of some sort. A movement that is ‘in beat’ starts easily and, provided there is enough power from the mainspring, is difficult to stop. Conversely a movement that is out of beat runs irregularly, is hard to start and is prone to stopping.

Traces that display a recurring irregularity or a sinusoidal effect (wavy line), divergence of the trace lines or sharp changes in direction when moving from a horizontal to a vertical position and vice versa all indicate certain defects that require correction before successful regulation can be carried out.


Practical Examples

1. Brand new Seiko SKX779 ‘Black Monster’ This watch utilizes the common or garden 7S26A 21-jewel movement that is regarded as the entry-level for Seiko automatics. Seiko themselves don’t hold out any great expectations for it, not even being game enough to quote acceptable accuracy limits in the instruction booklet that came with it. It’s known as a workhorse rather than a thoroughbred. After a couple of weeks’ wear it was running at +15 seconds/day—not bad but capable of improvement, I was sure. This was with being worn during the day (so the mainspring was nicely wound), and left dial up at night.

I popped it on the Vibrograf and got the following results: Dial up +30 s/d Dial down +30 s/d Crown up +30 s/d Crown down +36 s/d Crown left +27 s/d

To say I was surprised at the consistency of these results is an understatement—I was gobsmacked. Was I looking at a 7S26 or a classic Omega Observatory calibre? This augured well for a good result.

Bwc7s26Monster1168.jpg

. Since the timing machine showed the watch to be running fast by about 30s/d, and when being worn it gained only 15 s/d, my “personal wear error” was a loss of 15 s/d. I therefore had to bear this in mind when regulating the rate. (NB It is conventional wisdom that watches lose about 30 s/d when being worn, which may be why Seiko send them out from the factory running fast.) I undid the bracelet, carefully removed the caseback and put the watch dial down on the pickup. After re-checking the rate I gave the regulator lever a tiny nudge in the loss (—) direction to reduce it by 15 seconds. I let the movement settle down, then checked it again. Oops, 15 s/d too far. Another fractional nudge in the gain (+) direction and a recheck. Spot on—the machine said +15. Combined with my personal error of –15, the watch should in theory now neither gain nor lose whilst being worn in my usual manner. I set the Monster to within a few seconds of atomic time and made a note of the exact offset. After wearing the watch for 4 days I checked it again against atomic time. The result? It had gained 1 second total over the 4 days. Can’t complain about that!

2. 1979 Seiko 6309-7040 17-jewels This recent acquisition had an unknown service history but was running strongly at –24 s/d in daily wear. The vibrogram looked OK but showed the following: Dial up +4 s/d Dial down -20 s/d Crown down -56 s/d

Hmmm, quite a bit of positional variation, possibly indicating considerable wear and/or an unpoised balance. This one wasn’t going to be so easy as the new Monster.

The regulator lever turned out to be very touchy and it took me quite a few tries to get near my target of +4 s/d (dial down). In the end I had to settle for +2 s/d as being ‘near enough’. After 4 days of normal wear the 6309 had only lost 3 seconds, which in my book is more than satisfactory (and rather surprising, given the positional variations).

BwcSeiko6309mvmnt.jpg

The trusty 6309 17-jewel movement


How to regulate without a timing machine

Although having some sort of timing machine speeds up the process, it is perfectly possible to regulate your watches yourself with a little patience. Here’s how:

1. Wear the watch for several days in your normal manner, and make a note of the average gain or loss per day (your “personal error”) against atomic time or another known accurate time source such as an eco-drive or other quality quartz watch.

2. Remove the caseback carefully using a well-fitting opener (you may want to use masking tape on the caseback to avoid scratching it), and locate the regulator lever, screw or index. (See photos below for examples of the various types of regulator. They can be marked + and - , F(ast) and S(low) , or A(vance=fast) and R(etard=slow.)

BwcEterna-mvt1a.jpg BwcRtmvt7sa.jpg BwcMM mvta.jpg BwcOmega911a.jpg 321 balance a.jpg BwcPresmatic.jpg


3. In the case of a lever or index, hold a toothpick or the outside edge of a pair of watchmaker’s tweezers (e.g. #5’s) vertically, and gently nudge the lever so the pointer moves to counteract your personal error (p.e.). For example, if your p.e. is +20 sec/day, move the lever in the –ve or loss direction. If you’re unsure which is the correct direction, just remember that moving the regulator lever TOWARDS the stud carrier will effectively LENGTHEN the hairspring and thus SLOW DOWN the rate. Moving the regulator AWAY from the stud carrier will effectively SHORTEN the hairspring and thus SPEED UP the rate.

BE VERY CAREFUL NOT TO TOUCH THE HAIRSPRING—IT IS VERY DELICATE AND EASILY DAMAGED. Also, you only need to move the lever a very small amount—a little movement can make a big change in the daily rate. It helps to use a loupe (jeweller’s eyeglass) whilst doing this. I prefer to hold the toothpick/tweezers vertically as it is easier to control the movements of the lever. Pushing on the lever horizontally can lead to ‘overshooting the mark’ and the risk of the implement slipping off and damaging the hairspring. In the case of a microregulator with an adjusting screw, use a well-fitting screwdriver in the screw’s slot to turn it slightly in the appropriate direction.

4. Replace the caseback finger tight, making sure any seal or gasket is replaced correctly in its groove and is not pinched. Do not put the watch near any moisture or dust-laden environment until the regulation process is complete.

5. Wear the watch normally for 24 hours, noting the new daily rate. If necessary, repeat steps 2-5 until you are satisfied with your watch’s performance.

6. Tighten the caseback fully with the wrench, again making sure the seal/gasket is in good condition and correctly fitted. If you intend wearing the watch in water it is highly recommended to fit a new seal properly lubricated with silicon grease. Good luck and happy regulating!

External Links[edit]