Deciphering mainspring specifications…

[Zendoc]

Hello All,

I’m looking for advice about how to interpret the mainspring specs described in the GR documents and in the catalogues of online sellers such as cousins. 

My problem relates to interpreting the value describing barrel size - I’m unsure what the last number REALLY refers to.  

For example:

1.10x.0105x360x11 

(This refers to GR2628-X)

Does the “ 11 “ here refer to the barrel inside diameter of the calibre OR to the diameter of the wound mainspring before insertion?

For example if my measured barrel inside diameter to 11mm, I would normally select a wound spring size a little smaller, say 10.5, wouldn’t I? 

Or, does the barrel diameter described in the catalogue specifications take that into account and I would therefore order a catalogue size 11 for an 11mm barrel ID?

Hope this makes sense.

I can’t find clarification of this anywhere.

Many thanks for any enlightenment.

Cheers,

John

[RichardHarris123]

Yes, barrel size.  A 11mm spring will push out into a 11mm barrel, no need to go 0.5mm smaller. 

[Zendoc]

Thanks Richard.  That’s clear.  

So - another question - how much latitude is there in barrel and spring sizing?

Say I have have measured the ID of my barrel to be 9mm and (all other mainspring specs being equal), the closest match I can find for a replacement mainspring is listed with an 8.5mm barrel size, is it acceptable to fit the smaller spring in to the slightly larger barrel without compromising anything functionally? 

I’m in this situation now with an AS 1686 which has a 9mm barrel, but the closest replacement I can find has an 8.5mm diameter listed in its specs.  I can find an alternative 9mm spring, but would have to accept a different thickness.  I’m unsure how much wriggle room is acceptable, or whether in mainspring sizing, there’s a priority - ie thickness over barrel size etc.

Thanks again for any info. 

Cheers,

John

 

[nevenbekriev]

Hi John,

The most important parameter of the spring is the thickness. The lenght is something that one would not consider as obligatory, as relatively big variations will not affect anything in significant way. The barrel size is related to the barrel and has nothing to do with the spring. Yes, the spring is designed to work in a barrel and the size of the barrel matters, but let this parameter be chosen by the movement manyfacturer, not by the sping manyfacturer who doesn't have idea in what kind of movement the spring will work. Sometimes the washer in which the spring arrives will be bigger than the barrel and there will be no way to just push the spring in the barrel and be happy. Well, if this is a problem for somebody, then probably He is a person with to many problems that will be verry hard to help to in any way.

By the way, why You are changing the spring? Is the old one broken?

[Knebo]

Hi @Zendoc,

I guess you're like me and you don't have a mainspring winder. 

I searched GR mainspring for the AS 1686 mainspring and I now understand your dilemma. The exact one is discontinued. To answer you question more directly, I see two good alternatives for you:

a) GR4485 with 1.60 x .10 x 280 x 8.5 ( I guess this one is the one you found):
https://www.cousinsuk.com/sku/details/mainsprings-by-list-watch-pocket/GR4485

OR

b) GR4477 with 1.60 x .095 x 320 x 9.
https://www.cousinsuk.com/sku/details/mainsprings-by-list-watch-pocket/gr4477

 

As you know, the the AS 1686 spring is supposed to be 1.60 x .10 x 300 x 9.

- both of the above will fit and you'll manage to insert them from the retention ring without need for a mainspring winder (provided you know how to do it).

- GR4485 will give you more power (=higher amplitude), but less power reserve (due to shorter length) than the original.

- GR4477 will give you the same power reserve as the original. However, a bit less power.

 

Personally, I'd take GR4477 for two reasons: 1. Well, the power reserve is better. 2. With modern lubricants and modern mainspring metal alloys, you tend to get a higher amplitude anyways. You'd maybe even end up with too much amplitude (knocking) if you go with the original size and use thin synthetic oils-- see discussion here: https://www.watchrepairtalk.com/topic/10580-omega-calt17-help/#comment-248101

 

 

 

Edited May 13, 2024 by Knebo

[praezis]

10 hours ago, Zendoc said:

 I’m unsure how much wriggle room is acceptable, or whether in mainspring sizing, there’s a priority - ie thickness over barrel size etc.

The diameter is that of the ring - afaik, will have to look it up.

“Wiggle room“ doesn‘t matter as long as the spring length is in the ball park. 

Essential sizes of a mainspring are thickness, width and length. Diameter is just a hint if you can push into the barrel directly or will need a winder, it is no property of the spring.

Frank

[nevenbekriev]

1 hour ago, Knebo said:

- GR4485 will give you more power (=higher amplitude), but less power reserve (due to shorter length) than the original.

Actually one would not claim such thing  about the power reserve without check to confirm this, but if any difference, it will be verry small.

[Knebo]

14 minutes ago, nevenbekriev said:

Actually one would not claim such thing  about the power reserve without check to confirm this, but if any difference, it will be verry small

I'm surprised you say that. 

I'd say that IF (!!) barrel diameter, beat rate and all other things are equal, the length is directly correlated with the power reserve. 

A random/quick google search found me this quote from Longines "The length of the mainspring directly correlates with the amount of power reserve a watch has. The longer the spring, the longer the power reserve or also the reduction of the vibration, i.e. from 28’800 A/h to 25’200 A/h. (less vibration, less energy needed)." (https://www.longines.com/en-za/universe/blog/what-is-power-reserve-on-a-watch)

A few anecdotal experience on my side would also lead to that conclusion.

Then, of course, it depends what you mean by "verry small". In the particular case with an original MS of 300mm, going to 280mm, I'd expect losing maybe 2h of power reserve. Just a "guesstimate". 

[nevenbekriev]

No, trus me, this is totally wrong.

Firs of all, there is theoretical lenght for a fixed other parameters (spring thickness, barrel internal diameter and arbour diameter), which will give maximal power reserve. This is written in all text books, but I have never seen in  the text books mathematical expression (formula) of the relation lenght/power reserve. I am to lazy to generate that expression now, but will draw for You the graphical representation of it.

As You can see, there is a maximum in the middle, but the slope near the maximum is really small, so even change of the lenght of let say +/- 30% doesn't lead to significant change of the power reserve.

More to that, sometimes movement designers put much longer springs in the barrels than the theoretical lenght for maximm reserve.  This is ecpecially in high grade movements. You will sometimes see that the spring covers not the half, but actually 2/3 of the barrel free space. The designers compensate this loss of power reserve by usage of bigger barrels. This way they achieve much smaller difference in full wind / end of power reserve torque. In such case, reducing the spring lenght leads to increasing of power reserve...

Edited May 13, 2024 by nevenbekriev

[Knebo]

Thanks @nevenbekriev. I did some further reading and I think I kinda understand it now. Basically, if you were to fully fill the entire barrel with a super-long mainspring, you can't actually wind it anymore. Hence zero power reserve. Likewise the other extreme (mainspring too short to be wound around the arbor). This explains in very basic turns why there is a sweet spot in the middle of the curve you drew.

Importantly, this is relative to the barrel diameter (and arbor diameter). In other words, if you have a larger barrel you should have a longer mainspring and hence also longer power reserve. So Longines' statement isn't entirely wrong (longer mainspring = more power reserve). BUT you can only make the mainspring longer if you also increase the barrel diameter.

Thanks again for making me think about this a bit more and learning something. That's why I love this forum!

 

However , there is still some truth to what I said (I think! Please correct me if I'm wrong!): according to my reading, the key parameter is the share of the space between barrel arbor and barrel wall. Half of which should be occupied by the mainspring (based on Theory of Horology by WOSTEP, quote in depth here: https://www.vintagewatchstraps.com/mainsprings.php). 

The space occupied by the mainspring in the barrel is a simple function of mainspring length AND thickness. 

This implies that increasing length, but keeping the same thickness, will lead to occupying too much of the barrel space and hence reducing power reserve. This is what @nevenbekriev 's drawing correctly shows.

However, if you increase length AND decrease thickness in the correct ratio, you can maintain the correct mainspring proportion vis-a-vis the barrel (i.e. occupying half of the space between arbor and barrel wall). 

This would indeed lead to an increase of the number of barrel revolutions (when unwinding) and hence a potential increase in power reserve. However, you loose torque. And a loss of torque will also lead to the watch stopping earlier (when torque can't overcome the friction in the gear train). Thus, these two opposing effects may cancel each other out. Which again makes this statement probably true: 

2 hours ago, nevenbekriev said:

but if any difference, it will be verry small

 

FINALLY, we still want to help @Zendoc with his very concrete decision: 

GR4485 (same thickness but shorter than original) or GR4477 (slightly thinner and a bit longer than original). 

I would still advocate (considering modern lubricants and potentially stronger metal alloys -- and consequently the risk of knocking at full wind) to choose the latter.

 

 

 

 

[nevenbekriev]

OK, a little bit more from today's lecture about the main springs.

Generally, the spring in the barrel is limited and can't  unwind fully.

With the blue line is shown the reserve/torque relation when spring unwinding is limited by the barrel. The green line represents the same when the spring is free to unwind fully And the red line shows the minimum torque that is needed for the movement to keep running. The yelow graph shows how the torque changes (from the blue) when the spring gets weaker (set) after 100 years of work. The purple line represents thinner and longer spring in the same barrel.

As You can see, using thinner and longer spring will increase the power reserve. The 'set' spring will have the same reserve as a new one with the same sizes, only the amplitude will be just a little smaller.  Of course, this is true only when the movement othervice is in good health ( the red line is lo enough)

[Neverenoughwatches]

12 minutes ago, nevenbekriev said:

OK, a little bit more from today's lecture about the main springs.

Generally, the spring in the barrel is limited and can't  unwind fully.

With the blue line is shown the reserve/torque relation when spring unwinding is limited by the barrel. The green line represents the same when the spring is free to unwind fully And the red line shows the minimum torque that is needed for the movement to keep running. The yelow graph shows how the torque changes (from the blue) when the spring gets weaker (set) after 100 years of work. The purple line represents thinner and longer spring in the same barrel.

As You can see, using thinner and longer spring will increase the power reserve. The 'set' spring will have the same reserve as a new one with the same sizes, only the amplitude will be just a little smaller.  Of course, this is true only when the movement othervice is in good health ( the red line is lo enough)

I think we've talked about this before, using a new mainspring to initially overcome the effects of more movement friction than there should be. The aim is to reduce all going friction not to push through the friction with more torque, that in the long run will create more wear.

[Knebo]

@nevenbekriev That's a very nice overview.

3 hours ago, nevenbekriev said:

The purple line represents thinner and longer spring in the same barrel.

As You can see, using thinner and longer spring will increase the power reserve.

As I said, right? 

 

 

 

 

2 hours ago, Neverenoughwatches said:

think we've talked about this before, using a new mainspring to initially overcome the effects of more movement friction than there should be

I don't think anyone is suggesting that here, @Neverenoughwatches. If anything, the opposite (taking a weaker but longer mainspring than the originally specified one). 

[Neverenoughwatches]

2 hours ago, Knebo said:

@nevenbekriev That's a very nice overview.

As I said, right? 

 

 

 

 

I don't think anyone is suggesting that here, @Neverenoughwatches. If anything, the opposite (taking a weaker but longer mainspring than the originally specified one). 

Looking at Nev's graph the thinner longer spring is delivering less torque but over a longer period of time. For this to happen and completely unwind itself increasing the power reserve the friction through the train must be less than if a specified spring were fitted.

[Bonefixer]

9 hours ago, nevenbekriev said:

No, trus me, this is totally wrong.

Firs of all, there is theoretical lenght for a fixed other parameters (spring thickness, barrel internal diameter and arbour diameter), which will give maximal power reserve. This is written in all text books, but I have never seen in  the text books mathematical expression (formula) of the relation lenght/power reserve. I am to lazy to generate that expression now, but will draw for You the graphical representation of it.

As You can see, there is a maximum in the middle, but the slope near the maximum is really small, so even change of the lenght of let say +/- 30% doesn't lead to significant change of the power reserve.

More to that, sometimes movement designers put much longer springs in the barrels than the theoretical lenght for maximm reserve.  This is ecpecially in high grade movements. You will sometimes see that the spring covers not the half, but actually 2/3 of the barrel free space. The designers compensate this loss of power reserve by usage of bigger barrels. This way they achieve much smaller difference in full wind / end of power reserve torque. In such case, reducing the spring lenght leads to increasing of power reserve...

I think your axes are mislabelled here. They need switching around. 

[nevenbekriev]

3 hours ago, Bonefixer said:

I think your axes are mislabelled here. They need switching around. 

Yes, correct. Stupid mistake...

[Zendoc]

Hi All,

Just wanted to say thanks to everyone who has contributed a reply to my original question. I’ve learnt a lot and I’m much clearer about the options for replacing the mainspring in my project.   It’s been a very fruitful discussion.

Cheers,

John

[Neverenoughwatches]

1 hour ago, Zendoc said:

Hi All,

Just wanted to say thanks to everyone who has contributed a reply to my original question. I’ve learnt a lot and I’m much clearer about the options for replacing the mainspring in my project.   It’s been a very fruitful discussion.

Cheers,

John

Here's a little more to add from an old book that contains articles from Emil Borer once head of Rolex's technical department and designer of their movements at Rolex's factory ( which apparently they had but not really until the early 2000s )  so i guess the mislead publisher meant Aegler's factory

Edited May 15, 2024 by Neverenoughwatches
Purposeful misspell of Aegler but no one picked up on it or were very polite in not correcting me ( my excuse anyway 🙂 ).

[Knebo]

On 5/15/2024 at 12:42 PM, Neverenoughwatches said:

 articles from Emil Borer once head of Rolex's technical department 

Wait, am I reading this correctly?? 

So, the former head of Rolex technical department suggests (in the last paragraph of your pictures) that - paraphrasing:

If you put a mainspring that is too strong (1st mistake), just file off bits of the pallet fork horns to avoid knocking ??? 

The Rolex master must be one of the greatest watch destroyers then, hahaha. 

[Neverenoughwatches]

1 hour ago, Knebo said:

Wait, am I reading this correctly?? 

So, the former head of Rolex technical department suggests (in the last paragraph of your pictures) that - paraphrasing:

If you put a mainspring that is too strong (1st mistake), just file off bits of the pallet fork horns to avoid knocking ??? 

The Rolex master must be one of the greatest watch destroyers then, hahaha. 

So firstly this book was first wrote almost one hundred years ago. Secondly Emil Borer was more likely in charge of design and research at Aegler's factory, the factory that Rolex claimed belonged to them, but Gruen also apparently made that same claim. And lastly Emil Borer was way way ahead of his time and very very probably knew exactly what he was talking about. 

2 hours ago, Knebo said:

you put a mainspring that is too strong (1st mistake), just file off bits of the pallet fork horns to avoid knocking ??? 

You must be on tenderhooks to read page 25 ? I'll release it mid July 2024