Yes, I know, not desperately exciting, but it cropped up recently, and so possibly worth documenting. I’ve had a few Seiko watches pass through my hands fitted with the ubiquitous 7S26 movement, and have had cause to notice, on occasion, that watches fitted with this movement can require a bit more wrist time to dial in a decent amount of power reserve. I had always assumed that the gearing of the autowinding mechanism on these watches was tuned to the more active type and gave it not much more thought. Recently though, I had cause to look more closely at what might be the cause of this apparent lack of efficiency.
The 7S26 and 7S36 movements are unusual in comparison to the 6-series automatics, in requiring the winding weight, the rotor, to be fitted with the weight aligned in a particular way with respect to the reduction wheels that transmit its motion into power reserve in the mainspring. I suspect it may not be that uncommon for the weights to be refitted post-service, incorrectly, or indeed, perish the thought, in the factory (although I would hope this unlikely). Somewhat curious about the effect of rotor alignment on winding efficiency, I recently performed a rather unscientific test with a watch whose rotor was about 30 degrees out of alignment. This next photo shows the rotor of a 7S26B fitted to a mid-size dive watch, more or less in such a state:
A good solid 8 hours on my wrist was sufficient to dial in about 17 hours of power reserve, more than enough to get through the night and certainly enough not to cause any sort of concern that something might be awry. If the watch was being used regularly by someone whose level of activity was comparable to mine, then a full power reserve would build within two or three days wear. If we take off the rotor
we see the wheel on its reverse that meshes with the first reduction wheel (below) whose rotating motion causes the pawl lever, whose axis is mounted off-centre of the first reduction wheel, to move back and forth rotating the second reduction wheel which then winds in power to the barrel and mainspring:
The beauty of this arrangement is that the pawl lever rotates the second transmission wheel in the same direction regardless of the direction of the rotor or indeed the first reduction wheel. This magic lever arrangement has been used widely in Seiko automatics since the early 1960’s. Refitting the rotor, we follow the instructions given in the 7S26 technical manual:
We see that with the winding weight set so that the midpoint of its arc lies along the stem direction, the hole in the first reduction wheel should point towards the hole in the balance cock.
When the first reduction wheel is in this position, the pawl lever is precisely at its turning point, i.e. the lever is poised so that depending on the direction in which the first reduction wheel turns, either one or other of the two arms will move to engage with and rotate the second reduction wheel. My guess, and this is only a guess, is that it is supposed to be aligned thus, because wrist movement is more likely to swivel around the axis defined by the arm which itself aligns along the stem and crown axis. With the weight at its midpoint when it sits over this axis, rotating the wrist will cause it to ‘fall’ one way or t’other so winding the watch. Moving the arm up and down will rotate the watch about an axis perpendicular to this one which will be less likely to unbalance the weight.
With the rotor in its correct position, 9 hours on the wrist resulted in 37.5 hours power reserve. A bit of a result. The conclusion therefore is that a misaligned rotor is only likely to cause issue in people who aren’t especially active during the day but could result in the watch stopping overnight following a day’s wear.
Adventures in Amateur Watch Fettling 
I believe you have followed the instructions too literally. The ideal position for the rotor is ‘down when the watch is worn on the owner’s arm which, in turn, is hanging freely by their side’ or, more succinctly: 3 o’clock for the left wrist and 9 o’clock for the right. The key is, as you say, that the maximum amount of work is transferred either side of the position where the dots line up on the first reduction wheel, so we must, therefore, ensure the rotor is at its lowest point when this is so.
Of course, if you have 30º wrists, the 4 o’clock alignment is correct, and I apologise 😉
The 7S26A and 7S36A technical manual is pretty unambiguous in its instructions on refitting the rotor, leaving little room for concern that one might be following instructions too literally. It states that one should ‘set the middle point of the oscillating weight’s arc towards the winding stem…’ It makes no distinction between watches with the crown at 3 and those at 4 in spite of making reference to both variants elsewhere in the manual. In this instance, though, the proof of the pudding is in the eating: prior to the realignment, 8 hours on the wrist yielded 17 hours of power reserve; after refitted, 9 hours on the wrist yielded 37.5 hours of reserve.
As I said in the original post, my speculation about why the weight should be aligned thus was based on guesswork. Regardless of how it is fitted, it will naturally fall to its lowest position when one’s arm is pointing downwards, but it is not at all clear to me that during a typical day, my arm spends any more time in that position than it does elsewhere! Even at rest in a standing position, the natural bend at the elbow means the wrist is in any case probably at 30 degrees to the vertical. While I would be interested to hear further suggestions about how and why the weight alignment in this movement matters and on others it does not, I certainly won’t be losing any sleep over it 🙂
The ratio between the amount of rotation of the pawl wheel against the first reduction wheel varies according to the position of the first reduction wheel in a sine wave pattern. Making sure the rotor is positioned so that the first reduction wheel is at the top point of this wave when the watch is at rest ensures the maximum efficiency of the system.
Yes, I am sure what you say is right and in fact it’s consistent with what I said in the penultimate paragraph about the pawl lever being at its ‘turning point’ when the weight is positioned as instructed in the manual. What I struggled with was how the ‘resting position’ of the weight is defined. It may well be crown down but of course that takes no account of a watch worn on the right hand. Thanks for the insights.
Thanks for a wonderful blog with excellent pictures.
I agree with JS. The instruction manual simplifies things too much and assumes all movements are installed on a case with crown at 3, worn on left wrist. Even my brand new Seiko 5 with crown at 4 has the weight aligned for crown at 3. Which is understandable for the movement that is possibly the most produced in history.
Pingback: A Seiko 5 Sports 7019-6030 from May 1970 | Adventures in Amateur Watch Fettling