My Seiko Automatic has started running fast.

[Zundfolge]

I popped in here to ask if anyone here had this problem and how to fix it. I have an older Seiko automatic (self winding) watch and the last few days it has been running fast (gaining several seconds a minute or about 5 minutes an hour).

Well I thought I'd do a quick bit of googling first and low and behold I found a solution ... one that was simple enough I figured I'd just post the answer to my own question here just in case anyone else here has had this problem.

I am not a watchmaker, so follow this advice purely at your own risk. But, it worked for me.

I received my first new Seiko, a Black Monster, back in October. Out of the box, it was gaining 10 seconds a MINUTE. I was disgusted, but I hate sending stuff back, so I did some research. It appears that the hairspring of the 7S26B can get hung up on the regulator pins if it takes a sharp impact. Mine was definitely not magnetized - it had no effect whatsoever on the needle of a good Suunto compass.

Based on posts on this forum, and others, it appears that the fast-out-of-the-box problem is common. However, the 7S26B seems to be very reliable in actual use. My theory is that the dynamics of the balance wheel are such that the hairspring is vulnerable when the mainspring is unwound and everything is "loose", but better controlled once the mainspring is driving everything.

Back to the solution. I tapped the watch, on a cloth-covered wooden table, alternately on the caseback and on the edge opposite the crown. Based on pictures of the movement, this seemed the logical direction to get the hairspring back in place. I have no way of calibrating the force used, but I would call it fairly hard "taps", but short of "banging" (all with wrist and finger motion, not moving my elbow). A level of force that I wouldn't expect to damage anything, and that certainly wouldn't mar the case in any visible way (in case I ended up sending it back after all).
The result: on about the 6th tap, the second hand visibly slowed, and the watch began to keep time. I have worn it daily since, with about +4sec/day accuracy on average and no problems at all. Your mileage may vary.
As far as the demagnetizing suggestion, that might actually work, in an indirect way. Demagnetizers, at least that I am familiar with, also make ferrous metals vibrate. This might shake the hairspring back in place, even if the watch was not magnetized to begin with. I would add that it would be fairly difficult to magnetize a watch in typical shipping packaging, with the box enforcing several inches of effective air gap around the watch itself. Also, the 007 is an ISO diver, with a fair degree of antimagnetic protection (not quite up to Milgauss or Gaussman level, but still more than an unprotected watch).

This watch is one of the watches I got when my dad passed away back in June (the other is my grandfather's early '60s Rolex Oyster Stainless that isn't running at all so I need to eventually take it to a shop). I remember my dad tossing this watch into a drawer back in the '90s complaining that it was running fast but he didn't want to spend the cash to take it to a watchsmith. Too bad all he had to do was smack it (if I had a nickle for every time he smacked a non working appliance or TV ).

[davsel]

Have you tried a heavier buffer?

[Zundfolge]

I just held it in my hand and tapped the backplate with my finger (rather hard, but not so hard it hurt) and then tapped the side of the watch on my mouse pad 5-6 times and it worked fine.

[TFOGGER]

I wear a Casio Solar/Atomic that sets itself to the NOAA atomic clock six times a day. Not as classy as the Seiko you have, but dead accurate and utterly reliable. My Seiko dive watch is quartz/battery, but I only wear it on dive trips or with "dressy" clothes (weddings and funerals).

[BPTactical]

Have you tried a heavier buffer?

He said it was an automatic.

Not without NFA stamps!

[Great-Kazoo]

He said it was an automatic.

Not without NFA stamps!

But you can work on them provided the customer has a copy of their NFA stamp, correct ?

[Ah Pook]

If it's an autowind and is getting wound up too much, causing it to run fast, maybe switching hands will fix the problem.

Just take the Swiss Army knife out and start adjusting.

[theGinsue]

It's not broken. Everyone knows time goes faster as you get older.

[BPTactical]

It's not broken. Everyone knows time goes faster as you get older.

Well, aren't you a bucket of good cheer?

[ruthabagah]

If colorado springs is closer to a black hole than denver is, then the time dilatation is normal, yet its a paradox.