winding mainspring into barrel by hand

[Bill2024]

Hi All,

can I just check which part of the mainspring is supposed to engage with the barrel when winding it back in by hand (no other options at the moment I'm afraid). The barrel and spring is from a st3620 (clone eta6498) which I thought I'd have a go at as I've never worked on a barrel before; I wanted to try this on my practice movement before getting into the movement I have been servicing.

I have circled the part of the spring and barrel which I "think" should catch when inserting the spring. I am a bit suspicious though as it seems very difficult to get a start and it feels like there's a very real chance of kinking/breaking the spring just after the circled part. I've had a few little goes at putting it back but didn't want to go too far without sanity checking what I'm doing. Any advice or tips on how to do this much appreciated.

I think the coils went anti-clockwise in the barrel before I took it out (why oh why didn't I take a picture) - does that sound right? As I understand the end of the spring that is not at the arbour end is wound into the barrel against its natural curve. 

 

If the worst comes to the worst and I completely bugger it up would a replacement mainspring for an ETA 6498 fit?

Thanks, Bill

[mbwatch]

A replacement for an ETA 6498 will fit, but might not perform exactly as expected (I have been through this before with a ST36 spring in a genuine 6498). Differences in beat rate between the two will mean differences in spring thickness that need to be accounted for. If it is a 6498-2 spring, I think that 21600bph matches both.

When winding it, the way you have it laid out on the table is correct. Starting from the tail, you have to wind against the spring's coil direction which is not the easiest to get started but you'll get the hang of it. I needs to end up in the barrel with the inner coil pointing left as you have it in your photo, so the whole S curve tail has to be wound backward from the position it wants to be in.

Do not worry about hooking the hook onto the edge of the barrel as you had wind. You don't need to do that for a regular Swiss tongue mainspring end. Just wind it in and eventually when the watch movement is wound up that hook will slide around to catch where it needs to be and hold onto there. (this is different and easier from other mainspring types with a hole in the end or a T shaped end that have to land in place before the lid can even  be closed). As you wind it in, there is a good chance the spring will make its way around the edge to catch the hook anyway. I hand wind springs frequently and can tell you I have never once worried about this.

Plan on the spring jumping out of the barrel at least once or twice as you get the hang of winding it in by hand. When it does, the barrel can fly a long way away so be prepared to search for it.

Edited March 7 by mbwatch

[Bill2024]

46 minutes ago, mbwatch said:

A replacement for an ETA 6498 will fit, but might not perform exactly as expected (I have been through this before with a ST36 spring in a genuine 6498). Differences in beat rate between the two will mean differences in spring thickness that need to be accounted for. If it is a 6498-2 spring, I think that 21600bph matches both.

When winding it, the way you have it laid out on the table is correct. Starting from the tail, you have to wind against the spring's coil direction which is not the easiest to get started but you'll get the hang of it. I needs to end up in the barrel with the inner coil pointing left as you have it in your photo, so the whole S curve tail has to be wound backward from the position it wants to be in.

Do not worry about hooking the hook onto the edge of the barrel as you had wind. You don't need to do that for a regular Swiss tongue mainspring end. Just wind it in and eventually when the watch movement is wound up that hook will slide around to catch where it needs to be and hold onto there. (this is different and easier from other mainspring types with a hole in the end or a T shaped end that have to land in place before the lid can even  be closed)

Plan on the spring jumping out of the barrel at least once or twice as you get the hang of winding it in by hand. When it does, the barrel can fly a long way away so be prepared to search for it.

Thank you so much. Yes tricky to get started but knowing that I was on the right track was a big help. I got it wound back in and the arbour back in.

The idea was to clean it and lubricate but I'll leave that and take the "win" for now! At least I know what I'm aiming at now. Thanks.

Edited March 7 by Bill2024

[JohnR725]

14 hours ago, mbwatch said:

A replacement for an ETA 6498 will fit, but might not perform exactly as expected (I have been through this before with a ST36 spring in a genuine 6498). Differences in beat rate between the two will mean differences in spring thickness that need to be accounted for. If it is a 6498-2 spring, I think that 21600bph matches both.

14 hours ago, Bill2024 said:

The barrel and spring is from a st3620 (clone eta6498)

The problem is 6498 is not the same watch is a 6498 – 2. Then there's the problem of parts listing later generation ETA watches don't typically have mainsprings listed as you're supposed to replace the entire mainspring barrel. And often times the mainspring listing doesn't seem to show up anywhere.

Although it looks like I'm wrong according the cousins they're exactly the same

By we do end up with this weird problem if the mainsprings are exactly the same then why do we have to have two separate mainspring winders?

Then someone was kind enough to make a nice PDF of mainspring winder sizes snipping out a image brings up the peculiar problem of the center part for the arbor is not the same size as the handles not the same size? I'm also attaching the PDF.

Then the other thing is interesting is for mainspring looking at ETA because of course the Chinese don't list parts anywhere in the universe and we don't exactly know how exacting they clone things. In a case looking up the parts listings we find out that the barrels are different between the two types and the arbors are  different and bestfit disagrees with cousins as to whether there even the same mainspring

Unfortunate problem of best fit on line is you get part numbers you do not get specifications but the indication is the mainsprings a be different.

 

HJ-Mainspring-Winders.pdf

[mbwatch]

Added this PDF to my stash. Sooner or later I will have collected all the world's knowledge of watch parts & tools through @JohnR725's slow drip of forum post document dumps.

I had no idea Bergeon sells separate winders for the two 6497/8 models.

[Bonefixer]

I also wind mainsprings by hand. Finger cots are a nuisance - they get pinched. Doing it ‘naked’ is easier. 

[AndyGSi]

46 minutes ago, Bonefixer said:

I also wind mainsprings by hand. Finger cots are a nuisance - they get pinched. Doing it ‘naked’ is easier. 

You're OK doing it naked as long as the spring doesn't come loose. 

[Neverenoughwatches]

It's possible a small unused appendage could get seriously injured .

[Bonefixer]

8 hours ago, Neverenoughwatches said:

It's possible a small unused appendage could get seriously injured .

Never tried it 'three handed'. Good tip

[mikepilk]

I started out on pocket watches. Winding some of those springs in by hand was a serious workout on the fingers.

[Bill2024]

So thinking about winding mainsprings in by hand made me wonder about whether just cleaning and lubricating a mainspring would have any effect on the amplitude/performance of the watch. 

From what I’ve been reading it’s pretty much standard to replace the mainspring when servicing a watch (assuming a spare is available if course) or even replacing the barrel.

With vintage movements though I imagine cleaning may be the only option. So would you expect a cleaned/lunricated mainspring to improve the watch’s performance, and if so is there any way of quantifying by how much, or would things be pretty much be the same as before cleaning. 

Edited March 9 by Bill2024

[mikepilk]

50 minutes ago, Bill2024 said:

From what I’ve been reading it’s pretty much standard to replace the mainspring when servicing a watch (assuming a spare is available if course) or even replacing the barrel.

If servicing a watch professionally then a new mainspring is a minor cost compared to the service, so will probably be fitted.

I used to nearly always fit new mainsprings to my own watches. But after reading advice, and years of experience, I found it rarely necessary to fit new springs. Unless they are completely set or distorted, I nearly always use the old spring. I'd guess on about 80% of the watches I service, I don't fit a new spring.

[AndyGSi]

1 hour ago, Bill2024 said:

With vintage movements though I imagine cleaning may be the only option. So would you expect a cleaned/lunricated mainspring to improve the watch’s performance, and if so is there any way of quantifying by how much, or would things be pretty much be the same as before cleaning. 

I'd expected cleaning and lubricating to give an improvement and the best way to quantify is before and after results on the timegrapher.

[Bonefixer]

3 minutes ago, AndyGSi said:

I'd expected cleaning and lubricating to give an improvement and the best way to quantify is before and after results on the timegrapher.

But to prove it you'd either have to clean the spring on its own, leaving everything else as is, or clean everything but the spring, test, then clean the spring and test again. 

[Neverenoughwatches]

2 hours ago, Bill2024 said:

So would you expect a cleaned/lunricated mainspring to improve the watch’s performance, and if so is there any way of quantifying by how much, or would things be pretty much be the same as before cleaning. 

That's a question that can't be answered without knowing the condition of the spring before it's cleaned and lubricated. A dirty gummed up spring  vs a cleaned correct lubricated spring... has only one answer.

[AndyGSi]

1 hour ago, Bonefixer said:

But to prove it you'd either have to clean the spring on its own, leaving everything else as is, or clean everything but the spring, test, then clean the spring and test again. 

Yes

[Bill2024]

1 hour ago, Neverenoughwatches said:

That's a question that can't be answered without knowing the condition of the spring before it's cleaned and lubricated. A dirty gummed up spring  vs a cleaned correct lubricated spring... has only one answer.

Yes I’m sure cleaning/lubricating isn’t going to make things worse - as long as nothing gets damaged doing the hand winding. I have a movement that has been cleaned apart from the barrel so I’ll give the barrel a once-over and see what difference it makes. 
I’d like to try a mainspring winder but from everything I’ve read the Chinese ones are not much good and the Bergeon ones cost a kidney. 

Edited March 9 by Bill2024

[AndyGSi]

4 minutes ago, Bill2024 said:

I’ve read the Chinese ones are not much good

It all depends which ones you get and what the arbor is made of.

[Bill2024]

2 hours ago, mikepilk said:

If servicing a watch professionally then a new mainspring is a minor cost compared to the service, so will probably be fitted.

I used to nearly always fit new mainsprings to my own watches. But after reading advice, and years of experience, I found it rarely necessary to fit new springs. Unless they are completely set or distorted, I nearly always use the old spring. I'd guess on about 80% of the watches I service, I don't fit a new spring.

Interesting to hear that. Is it only the circular springs that can become set, or can it also happen to the ‘s’ shaped ones?

I’ll be doing everything I can to reuse the springs in the watches I have lined up to practice on - mostly ones from the 70’s. 

[JohnR725]

6 hours ago, Bill2024 said:

So thinking about winding mainsprings in by hand made me wonder about whether just cleaning and lubricating a mainspring would have any effect on the amplitude/performance of the watch. 

From what I’ve been reading it’s pretty much standard to replace the mainspring when servicing a watch (assuming a spare is available if course) or even replacing the barrel.

With vintage movements though I imagine cleaning may be the only option. So would you expect a cleaned/lunricated mainspring to improve the watch’s performance, and if so is there any way of quantifying by how much, or would things be pretty much be the same as before cleaning. 

If you look at modern Swatch group servicing procedure they typically replace the entire barrel. his is why find it amusing that Omega has a working instruction on recycling a mainspring barrel. In other words taking it apart checking out the mainspring making sure it meets the requirements making sure the barrel and everything else basically making sure everything looks fine and then it tells you how to put it all back together again which for my understanding of people at Swatch group they never do this at all may just replace the entire barrel because they have a infinite supply of them.

That of course if you look at newer documentation versus older documentation for the same caliber of watch the older documentation will often times list mainspring or the newer documentation does not.

I'm attaching a PDF on how mainsprings are made. On PDF page 3 we get the quality and maintenance. In that particular section in the first paragraph it talks about the various qualities of spring is which is unfortunate problems for us in that it would only apply if we were purchasing directly from the factory. So for instance it indicates blue ring is better than red.

One of the reasons or bring up the ring color is often times the mainsprings that I receive are not packaged in a GR package they been repackaged by someone who possibly purchased from them and other people. Previously I wasn't paying attention to the ring color I was paying attention to the quality of the mainspring when I pushed it out of the ring as to how nicely it does not typically look. So for a while now I've been noticing that some of the mainsprings and I can always tell because the spring steel itself is a yellowish color and the last couple of them came in a yellow ring. Then a lot of times the mainsprings get repackaged by whoever's the distributor is in the region so it's hard to tell exactly where mainsprings come from so the articles amusing in that we should always get a specific spring only works if we can actually order a specific spring

Okay continuing on in that section on quality and maintenance the next two paragraphs talks about Teflon and the person who wrote the article experience of lubrication did not seem to matter at all or help at all? Which does bring up the problem of the Teflon is basically almost invisible except hinting of whitish powder. Depending upon where the mainspring was purchased from they may have other lubrication's including Omega at one time had a liquid lubrication and they did say yes you could use standard other lubrication's with without a problem here the indication is it didn't make an improvement at all? But we also have other companies that had other substances on their mainsprings

Oh which would bring up the Omega working instruction where they recommend wiping with a soft cloth lint free no solvents. Except a lot of the mainsprings we would like to reuse have been prelubricated by others before us with sticky grease so in which case you would have to use a solvent and if there is some sort of magical dry lubrication or Teflon on the spring using a solvent wipe it off and then you will need a lubrication. Unless of course and there is a theory that some of the steels used don't actually need a lubrication at all.

Then PDF page 4The at the it interesting problem of stress of mainsprings. Some more is at a lecture once where the comet made was that even using a decent mainspring winder because you have to wind it into a barrel relatively small and inserted into a watch that there is a time limit of how many times you can do that before he distort the spring. Plus of course it's generally frowned upon for hand winding Springs because that also has the possibility of distorting the spring and then it becomes cone shaped which is a problem or can become that's the fear.

Then the word set? We really need to define what set means which is the mainspring loses its spring characteristic it become soft and visually looks set. But even this brings in a problem of looking set might not actually indicate that it's soft at all. 

5 hours ago, mikepilk said:

Unless they are completely set or distorted, I nearly always use the old spring. I'd guess on about 80% of the watches I service, I don't fit a new spring.

Interesting your percentage is much higher than my percentage. because typically am working on vintage pocket watches. With typically all the mainsprings that are bad would be blued steel. Although I was working on a stopwatch last week with a nice  shiny white steel and it was still set. But I would say the percentage of mainsprings that I change or light 90 some percent of them only occasionally do I find that the spring inside it is looking good And I can keep it in

2 hours ago, Bill2024 said:

Interesting to hear that. Is it only the circular springs that can become set, or can it also happen to the ‘s’ shaped ones?

I think basically anything that is a mainspring whether it has a beautiful S-curve or doesn't have an S-curve at least one is brand-new conceivably will no longer function as a spring and can be set.

Yes like everything else in watch repair it brings up all sorts of interesting problems with lots of variations. Like for instance if you had a whole bunch of pictures of mainsprings and envelopes what exactly did the envelope sitting on them? For instance the lecture I once gave an mainsprings this was the first image of the lecture.

As you're looking at all as envelopes we notice things like unbreakable, self lubricating and of course unbreakable and anti-magnetic and non-setting. To basically any of the modern white spring is typically are saying guaranteed for life fever watch I wonder how we can make good on that warranty?

3 hours ago, Bill2024 said:

I’d like to try a mainspring winder but from everything I’ve read the Chinese ones are not much good and the Bergeon ones cost a kidney.

Somewhere in the group they been discussed before the unfortunate problems the Chinese products are they can range from very good the very bad and you may or may not have any way of knowing which one you're going to get. But I do think there were some better Chinese mainspring winders out and.

Then you just have to be vigilant as often times for instance Bergeon Up until relatively recent times they do make their tools when in the past and probably still do they purchased their tool. So the mainspring winder that costs well can be expensive brand-new occasionally has been sold under other brands because  other dealers were selling the same sets and no I have no idea were usually made them. So yes occasionally will come up for sale on eBay they don't have that nifty name and they don't come in a yellow box and you'd have to recognize what they are but there are examples that they have come up. Or if you're in the right part of the universe may be could find a watchmaker going out of business and Make a lucky purchase of tools.

4 hours ago, Neverenoughwatches said:

That's a question that can't be answered without knowing the condition of the spring before it's cleaned and lubricated. A dirty gummed up spring  vs a cleaned correct lubricated spring... has only one answer.

Testing would become interesting? If you were going to overhaul a watch remove the mainspring popped the lid if it's a blued steel spring or if it's covered in's nasty sticky grease you can then do an experiment if it looks nice and clean you'll need another watch for the experiment but blued steel would be the best.

Then it reassemble your mainspring barrel but the watch back together and see if you can test it? It's an ideal testing would be at least three positions on a timing machine following a procedure. Procedure would be a wind the watch up and let it run for 15 minutes minimum up to about an hour I'd be consistent with this. So if the watches running you can check it in the three positions record which you get then wait 24 hours because this is the more important part of the test can the mainspring and the existing condition of the watch even run 24 hours?

Then you would service the entire watch not the mainspring barrel. Repeat the experiment the whole thing wound up timing 24 hours later with your bad mainspring bear. Once you've completed that now you can clean and loop the mainspring make sure you're really careful with cleaning because a lot of times lubrication's really sticky and nasty. I've noticed that occasionally even on mainsprings that I've reused that whatever they lubricate it can be quite sticky. Now you put in your nice clean mainspring lubricate repeat the test and get back to us with the results.

The overall prevailing theory of a set mainspring in other words one that no longer a spring characteristics would be that the watch may not actually run the 24 hours.

 

 

2000-08-web-2 mainspring -1.pdf

[Bonefixer]

8 hours ago, Bill2024 said:

Yes I’m sure cleaning/lubricating isn’t going to make things worse - as long as nothing gets damaged doing the hand winding. I have a movement that has been cleaned apart from the barrel so I’ll give the barrel a once-over and see what difference it makes. 
I’d like to try a mainspring winder but from everything I’ve read the Chinese ones are not much good and the Bergeon ones cost a kidney. 

I'd recommend avoiding the Chinese winders. I bought a set, ensuring that they had steel arbors. Still no good. They're movement specific, and even when I've used them with the exact movement they don't work. 
 

I instead bought a few Bergeon ones - sizes 5-8 - and have had far more success, but even they're not guaranteed to work. Hand wind or new spring are I believe the best options. 

[JohnR725]

9 minutes ago, Bonefixer said:

I'd recommend avoiding the Chinese winders. I bought a set, ensuring that they had steel arbors. Still no good. They're movement specific, and even when I've used them with the exact movement they don't work. 

The problem is with the mainspring winder sets you can't have enough of them. Then there is a problem of when the Chinese cloned a set they cloned the ETA set of specific calibers. It would've been nice if they had clone the general-purpose set. Plus even if they did that it brings up the minor issue of the handles themselves.

The problem with having a Winder for specific caliber is it should only fit that specific caliber so like the handle would be the same size as the arbor more or less and that might not correspond to any other mainspring other than that specific one. As opposed the general-purpose set where they're in different increasingly increment sizes. But sometimes when I use one of those sets typically for smaller pocket watches I might have to do a mixing and matching of handles and the outer barrel part to get something that's going to work. In addition to back she make three different handles or did make three. In other words had the standard right-handed and left-handed which you need for water to Seiko watches may be a few other things perhaps. Then the vintage handles that have a much more aggressive hook part that sticks out for the blued steel Springs which is why the newer handles are sometimes extremely problematic for blued steel Springs

 

 

 

[RickTock]

These 3D printed mainspring winders have become my go-to. My nice Chinese set with steel arbours is just collecting dust.

[Bill2024]

On 3/10/2025 at 7:43 AM, Bonefixer said:

I instead bought a few Bergeon ones - sizes 5-8 - and have had far more success, but even they're not guaranteed to work. Hand wind or new spring are I believe the best options. 

I will look out for Bergeon winders but I probably can’t justify the cost (even for just a few of them) just yet.

Mostly I will try to clean and reuse mainsprings. If a new one is needed however, I found an interesting article/video on watchrepairtutorials about sizing a generic mainspring to a barrel if you can’t get the exact one for whatever reason. 

https://watchrepairtutorials.com/how-to-size-any-watch-mainspring-to-a-barrel/

[AndyGSi]

Most new mainspring come in a washer used to insert into the barrel without needing a winder.