A watch that is "out of beat" takes significantly different time to "Tick" than to "Tock". A pocket watch usually "Ticks" and "Tocks" 5 times a second or 0.200/second per cycle. When it is "In Beat" it should be less that 0.001 seconds difference between "Tick" and "Tock". Anything more than that will cause a deterioration of accuracy.

While re-assembling a watch, I always assemble the Balance wheel to the appropriate movement plate to check "free swing" in all 6 positions before all the rest of the stuff is there. Most of the "3/4 Plate" watches are "wide open" at that time which also makes it very easy to get a pretty good idea if the wheel is "in beat" when at rest. There is an imaginary line you can visualize as going through the Balance Wheel staff and the Roller Pin Jewel . . .

When the Wheel is in the Bare Movement plate you should be able to extend the sight line of the Roller Pin Jewel and Balance staff through the Banking pins. If that line equally bisects the Banking pins then the watch will be very close to "in beat", and if they do not, this is a good time to adjust the hairspring collet so they will.

Thanks Dave, with two pictures and an explanation I now understand something that's been confusing to me in the past..Your willingness and ability to teach is outstanding. Thank you,
Ray

Nice pictures Dave.
That's the best explanation I have seen on this subject. It will help a lot of people.
Brian C.

David,
Did you straighten that bent pivot before you installed that balance?

Hey David,
Your explanation makes it easy to understand how to get "in beat". I have been doing it by "eyeballing" since I got started. They say there are no dumb questions,so here goes. I know that "in beat" is important,but,wouldn't the short "tick" make up the difference in time from the long "tock",thus making the timekeeping difference miniscule? I have no training of any kind,and only about 6 years experience,most of which was trial and error,but if I may I would like to recommend a book. I bought it off Ebay when I first started out,knowing nothing,and it helped me a lot. It is still available,and only about 20 bucks. lots of information,in plain language,including how to make many specialty tools. I know it won't be of much use to you professionals,but the novices might learn a lot from it. I sure did.

I should add,the book was published in 1948,but some things never change.

First, "Beat Error" is based on the different time needed for the balance (hair) spring to Compress (Tick) and Extend (Tock) and then recover from those oscillations in each direction and pass through the pallet fork to unlock it from the escape wheel thus propelling the balance wheel through another period of oscillation.

Sadly, the propulsion time is always negatively altered as there will never be a better "kick" from the escape wheel than that of a watch in perfect "beat" due to the unlock-lockup cycle the banking pins limit to full pallet fork rotation.

As the Roller Jewel is necessary to "unlock" the Pallet Fork causing the escape wheel to "flip" the Pallet stone outwards and impulse the roller jewel (and balance wheel) to continue their journey around the wheel axis, "beat error" can only increase deviations of the swing cycle of the wheel.

i.e. when it unlocks early, it cannot kick any farther or longer because the pallet fork is stopped from its "throw" by the banking pin. But on the return, because it unlocks later, it will shorten the maximum "kick" because the roller pin is exiting the pallet fork before the full "kick" is transmitted, causing a shorter motion "swing" of the balance wheel.

In fact that is WHY there is Beat Error, due to this "short motion" on one side of the balance wheel oscillation cycle.

i.e., if there is 0.010 sec beat error, this will be magnified by a factor of 4, and become a difference of balance wheel swing of about 20%, that calculates on a watch with a "good swing" to about 50 degrees less rotation in one direction, and sadly, that error is NOT made up for on the "backswing".

The temperature changes that cause the balance spring and compensated balance wheel to work in opposition to keep the watch on time are in greater harmony with 0.001 sec beat error and keep better time. With 0.010 sec beat error (20%) beat error, what could be a few seconds a day on a well timed watch becomes 30 seconds to a minute on a watch simililarly timed without eliminating that beat error simply due to temperature fluctuations.

Add gravity problems with a badly poised wheel (dirty) and that can increase to minutes per day and be very position sensitive too.

As for the bent pivot, when I drew the white lines, I caused a "trick" optical distortion. The pivot is fine. (btw this is a double signed Sears Roebuck Hunter cased Getty, s/n 1551435 in great shape!)

Thanks David. I actually understood that.

Dave, what tool do you use to adjust the hairspring collet, and how do you stabilize the balance assembly while twisting the collet?

The experienced watch maker uses a small screwdriver sometimes specially shaped at the tip. With that they can usually reach the collet for adjustment while it is in the movement. Lacking thise skills, I use a tool made to reach into the balance wheel through the Hairspring and go over the pivot so that its small "spur" can access the collect "split" and allow you to rotate the collet ever so slightly. Figuring out the direction is another story, but this is the tool, and I use my Watchmaker's pin tool to hold the Balance cock, and then I hold the Wheel in my other hand (instead of the camera when I was taking this picture).

David, Thanks for the great explanation and fantastic photos. Sometimes little procedures like this are hard to come across in books. You presented it in a very clear manner, and it was much appreciated. P.S.- How is that West Troy Eckert coming along?

David,
Another great lesson with great visual aides.
Thanks,

Dave that is a wonderful set of photos and an explanation that is outstanding. Thank you.

No offense, but I don't like the looks of this procedure going down the coils while still on the balance cock. Especially since the picture shows the hairspring out of the regulator pins. You simply can't do that without distortion of some kind. I would just as soon take the balance off, make the adjustment and put it back on. To me, there is much less chance of deforming the hairspring. Of course, that is just my personal opinion. To each his own.

Roger, I was answering a question about using an adjustment tool, and I was showing it's use on a parts movement Balance Cock/Balance spring assembly without holding the Balance wheel, as I had the camera in my hand as explained in that text.

I apologize if this offends you and agree totally that in real life this adjustment is made with only about 25% of the balance spring extension pictured and the wheel is carefully supported by my right hand.

When the human race evolves into a three or four handed species then I would take a better picture.

Hi,

Dave is well authoritative in horology so my modest opinion is really not needed but he has shown and explained the correct procedure and the right tools to adjust the hairspring to ''beat''
and that in accordance with books, courses, experience and tools made available by the Industry. All other procedures can be considered a second option....
rgds

Enzo

David...I'm sorry if my post sounded offensive. My concern is that someone with little experience attempting this would deform the hairspring. There are many legitimate ways to accomplish the same tasks and many of them are less likely to cause a problem for the less experienced.

I don't think Dave needs for the cavalry to come to his rescue yet, he can very well hold his stand on his own.

Rather, here is something I personally consider to be one of the major improvements in mechanical horology...In many modern watches (wrist watches too) the hairspring pinning point is located on a rotating platform, coaxial to the regulator.

The watch is installed fully assembled on the timing machine and both the beat and the timing can be adjusted in real time...And that, my friends, is real progress.

Regards
Peter