How much mainspring deformation is too much?

[pubudeux]

Hey all!

After diving into the theory of mainspring barrel disassembly and assembly for a while, I decided today was the day to try it in practice. I disassembled and assembled a mainspring on a Seiko 6309A scrap movement a few times.

This movement is an automatic one, so it has that stiff bridle at the outer edge of the spring. 

I realized after the second disassembly that it looks like my first attempt deformed the mainspring a bit, as a bunch of the curves were less "curvy" than I would like them to be. 

Does this mean I've ruined that particular mainspring?

Or, could I treat those little deformations with some pliers or manually? Or is it something that would be fixed when the mainspring is wound and settles inside the barrel? When it was wound inside it looked perfectly round, so I wonder how much of a difference this would make. In any case I'm hopeful that the cause of it was my shoddy technique on the first attempt.

[pubudeux]

Answering my own question:

Found a really great resource that covers the logic behind different lubrication points on mainsprings, and photos and examples of all kinds of damage and situations and examples of when a replacement is warranted.

 

https://omegaforums.net/threads/basic-watchmaking-tips-oiling-part-2-the-mainspring-barrel.71246/

[oldhippy]

Watch springs/clock springs if they get deformed in any way I would always replace them. The thing that can happen with a bad spring is it can play up the movement when it comes to time keeping, also springs in barrels can cause wear in the barrel. I haven't looked at your link.

[nickelsilver]

3 hours ago, pubudeux said:

Answering my own question:

Found a really great resource that covers the logic behind different lubrication points on mainsprings, and photos and examples of all kinds of damage and situations and examples of when a replacement is warranted.

 

https://omegaforums.net/threads/basic-watchmaking-tips-oiling-part-2-the-mainspring-barrel.71246/

Archer (from the link above) is an outstanding watchmaker and is very generous with his time and knowledge on at least a couple of forums. I will say though, that the spring he shows that is "out of flat" and thus needs to be replaced, is absolutely A-OK. If you have a spare spring around, nice and flat, put it in the barrel, grab the arbor with a pinvice and wind it up without the cap on. You'll find that it wants to go everywhere except around the arbor. This is why there is wear on barrel bottoms and caps, and why the sides of high grade mainsprings are well finished, one reason why the spring needs to be lubricated, and why some really high end makers actually add hardened steel discs above and below the spring (I think one even used sapphire discs). Of course this has been an argument I've had with pro watchmakers over the years, to the extent of purposely distorting a mainspring and comparing amplitudes and running reserves to prove the point. Yes, when very distorted in the flat it can pose a problem, but a slight out of flat laying unwound on the table is no problem.

[CaptCalvin]

Don't bother trying to bend it. You'll only be doing harm by weakening it. No matter how much you tweak the shape once you wind it back into the barrel you'll see that it settles right back to the shape you started if you take it back out again.

Edited July 21, 2020 by CaptCalvin

[watchweasol]

Hi There is only one answer to a distorted mainspring and that is change it, unless you are prepared for trouble, which it will most likley cause in the future. With the amount of GR mainsprings on the market there is no excuse really.

[spectre6000]

Interesting mix of opinions here from experienced people... "Any time there is a disagreement between two rational parties, the answer lies in the middle". Also, "everything is gray, there is no black or white."

Since the responses (here and elsewhere) range from "any deformation means replacement" to "don't worry about it you're fine", my takeaway is a little deformation isn't going to affect anything, but there's a curve of diminishing return with an assumed point where the cost of a new spring and the wear/performance intersect. Since there's a cost component, it's specific to each watch by value (market or sentimental). With an expensive watch with a high accuracy requirement or a sentimental watch without easy access to replacement parts, err on the side of a new spring. If it's a cheap watch that's not capable of especially high accuracy and the wear isn't that big of an issue (due to either a ready supply or replacement parts or a relatively low value), don't worry about it. Then draw a line between the two, and figure out where you fall on that line.

[oldhippy]

One reason I always replaced a bad m/s was I didn't want the watch or clock coming back or the customer complaining. I had my reputation to think of.  

[pubudeux]

2 hours ago, spectre6000 said:

Interesting mix of opinions here from experienced people... "Any time there is a disagreement between two rational parties, the answer lies in the middle". Also, "everything is gray, there is no black or white."

Since the responses (here and elsewhere) range from "any deformation means replacement" to "don't worry about it you're fine", my takeaway is a little deformation isn't going to affect anything, but there's a curve of diminishing return with an assumed point where the cost of a new spring and the wear/performance intersect. Since there's a cost component, it's specific to each watch by value (market or sentimental). With an expensive watch with a high accuracy requirement or a sentimental watch without easy access to replacement parts, err on the side of a new spring. If it's a cheap watch that's not capable of especially high accuracy and the wear isn't that big of an issue (due to either a ready supply or replacement parts or a relatively low value), don't worry about it. Then draw a line between the two, and figure out where you fall on that line.

This is a great point. One thing I'd also love to understand are the consequences.

What does it do to the performance to have a non-flat spring, for example? My thought would be that it might cause certain parts of the spring when unwinding to drag on the bottom or top of the barrel, leading to drops in amplitude over time.

If that wild speculation of mine is true, it would also make me think that on the horizontal plane if the lines are not 100% curvy it wouldnt make too big of a difference when wound, as it shouldnt cause undue drag, but maybe additional friction of the spring rubbing against itself.

[pubudeux]

4 minutes ago, oldhippy said:

One reason I always replaced a bad m/s was I didn't want the watch or clock coming back or the customer complaining. I had my reputation to think of.  

This is a good point, but from the hobby perspective its a little difficult as I dont have a reserve of new mainsprings. I think if I was honed in on a specific type of watch and could predictably source them I, too would probably replace it every time.

[jdm]

2 minutes ago, pubudeux said:

What does it do to the performance to have a non-flat spring, for example?

That the watch, before any in deep technicalities, will not perform the best. As an old watchmaking say goes, adjustment starts at the barrel.

[pubudeux]

2 minutes ago, jdm said:

That the watch, before any in deep technicalities, will not perform the best. As an old watchmaking say goes, adjustment starts at the barrel.

Thanks jdm but im asking specifically how that performance impact looks in a physical sense

[spectre6000]

Irregular power delivery from the mainspring due to irregular drag on the barrel/lid over the course of the spring's unwinding (in addition to the wear over time) would cause the amplitude of the balance to vary over the course of the mainspring's travel as a subset of the energy intended for the balance is lost to friction within the barrel. In short, isochronism suffers.

[pubudeux]

48 minutes ago, spectre6000 said:

Irregular power delivery from the mainspring due to irregular drag on the barrel/lid over the course of the spring's unwinding (in addition to the wear over time) would cause the amplitude of the balance to vary over the course of the mainspring's travel as a subset of the energy intended for the balance is lost to friction within the barrel. In short, isochronism suffers.

Makes sense to me. I have a watch that I serviced without touching barrel and is in beat, but has some weird variations in timegrapher readings over time. The periodicity of these differences doesnt seem to match any of the wheels in the train, but I think it might be explained by a gunky mainspring. I'll test it out and see if that's the case.

This seems to me to be the prime candidate for those kinds of wildly curving or seemingly random direction switches on a timegrapher reading when the beat remains regular and clean.