Jump to content

Balance wheel beat adjustment--unfriendly collet

LittleWatchShop
 Share

Recommended Posts

LittleWatchShop Distinguished Member

LittleWatchShop

Posted
Posted

I finished an ETA 2834 the other day.  When timing it, I moved the stud to the end of adjustment range to adjust the beat and then had almost no more range left to move the regulator to get it timed properly.

So, I removed the balance with the intent to rotate the collet.  Well...I had not seen a collet like this before and there was no slot for me to insert a screwdriver and give it a little twist.  The picture here came off the interweb--it looks just like mine.

I assume that the proper method must be to just remove the hairspring and reinstall with a little rotation.  Am I correct?

The watch is running nicely with a little beat error, so I don't plan to do this...for now at least, but I would like to know the next time I run into this.

2021-12-16 06_49_40-Window.png

mikepilk Distinguished Member

mikepilk

Posted
Posted

I recently had exactly the same problem with an ETA 2789 movement. I don't understand how it ended up so far out - a new balance would be close enough? See here :

https://www.watchrepairtalk.com/topic/18380-does-hairspring-on-eta-2789-just-lever-off/

As the collet doesn't have a slot to rotate, I didn't know if they are easy to lever off. Which is why I posted the question :

The answer (thanks @nickelsilver ) is they just lever off as usual. 

  • Like 1
mikepilk Distinguished Member

mikepilk

Posted
Posted

As I posted in my question, I did consider turning the roller, but then you have to push it on with the hairspring in place.

It came off as easily as other collets

Nucejoe Distinguished Member

Nucejoe

Posted
Posted
43 minutes ago, LittleWatchShop said:

I thought about doing that but would not have anticipated the poise issue!

Beart error is a small angle, so the effect on poise is minimal.

Nucejoe Distinguished Member

Nucejoe

Posted
Posted
26 minutes ago, mikepilk said:

As I posted in my question, I did consider turning the roller, but then you have to push it on with the hairspring in place.

It came off as easily as other collets

Is this a modern collet? 

Ronp Member

Ronp

Posted
Posted

I would check the spring itself first before moving the collet or roller. They are normally very tight so the change they slipped is really minimal.
is the bridge for the palletfork oke?if one side is off, damaged you also create a beat error, same if one pallet has moved in the fork. I would change this too first with another set before turning the collet etc.

 Afterall, you not want tp correct an error by making another error. 

  • Thanks 1
mikepilk Distinguished Member

mikepilk

Posted
Posted

On my watch the spring looked good, and I could see nothing wrong with the pallet. Which is why I was puzzled as to how the beat could be out by so much.  Everything looked normal on the timegrapher.

As I mentioned, it was easy to lift the collet - less risky than moving the roller with the hairspring in place.

JohnR725 Distinguished Member

JohnR725

Posted
Posted
4 hours ago, mikepilk said:

I did consider turning the roller, but then you have to push it on with the hairspring in place.

I think they actually make a special tweezer for holding the roller otherwise you can put it in a lathe and just turn it no need to remove it. But it does screw up the poise. Plus it is on silver type that could be an issue

then I see in this discussion we of two separate watches with the exact same problem I don't suppose both of you could give us a decent picture looking straight down on the balance assembly just because I'm curious about something? Yes I can look up your watches and find pictures when I want to see your watch not the watch online. Plus the background history of your watches is always interesting like have they been worked on before. In other words how did they come to be in the condition therein? As they probably didn't leave the factory that way did they?

nickelsilver Distinguished Member

nickelsilver

Posted
Posted

If turning the roller to correct beat error I'm usually chasing down a degree or two of movement which would be immeasurable on the poise. There are tweezers for this, with round cutouts both sides and they accommodate about any roller. Joe mentioned my tool from a long lost post, this allows moving the roller in very tiny and precise increments and was a lifesaver on a job I used to do.

 

But on this stuff, figure out about where you want the stud, lift collet and set it there. Bang.

 

mikepilk Distinguished Member

mikepilk

Posted
Posted (edited)

The stud is in a similar position on my watch

I bought the movement without a case. The dial has 'Garrard' on it, so that usually means a good spec movement (I believe that the curved balance arms mean it's a 'Top' spec?)
Nothing else unusual.

 

1.thumb.jpg.47e782fb2f46845cbb0739fcb4bda225.jpg

movement-017.thumb.jpg.b3d26837f6c98751ab61f9675be1cccd.jpg

Edited by mikepilk
JohnR725 Distinguished Member

JohnR725

Posted
Posted

Interesting they both look like they're in the wrong place but they really can't be where there supposed to be can they? What I wanted was to look at your pictures versus like this link but it's picture really sucks The second links a little better but the studs definitely are not where you'd normally think they would be our they?

http://www.ranfft.de/cgi-bin/bidfun-db.cgi?10&ranfft&0&2uswk&ETA_2834

http://www.ranfft.de/cgi-bin/bidfun-db.cgi?10&ranfft&0&2uswk&ETA_2789

Probably what's happened is somebody's replace the balanced staff and was sloppy about it I just didn't care about things. It's amazing what people get away with out there. I must of course something isn't tight and it's rotating all by itself I've seen that occasionally but not very often.

Then I keep thinking may be perhaps there's a tool because I vaguely recall seeing one someplace but memories are a funny thing that way. Although the second link hints may be there is a tool perhaps?

http://shop.hegel.ch/produkt/hairspring-collet-virole-greiner-typ-4/

Another unhelpful link it's probably pricey but there is a tool for correcting and maybe that's what you need

https://www.greinervibrograf.ch/en/Hairspring-tools.htm

On the other hand it's good that you can get the hairspring off there's some of them you can't get off if you want them removed you have to remove the staff by pushing it out and then starting over again.

 

 

mikepilk Distinguished Member

mikepilk

Posted
Posted

I thought that the stud looked to be in the wrong place - but I Googled images of ETA 2750 (the base movement) and all the studs look to be in a similar position to ours. 

LittleWatchShop Distinguished Member

LittleWatchShop

Posted
  • Author
Posted

I dunno...there is a part of me that wants to go back in and get a perfect beat with the regulator at dead center, but then there are a hundred (well maybe 20 or so) watches in the queue.  It is running nicely keeping good time (have not worn it yet) And...none of these watches will leave the house after being finished I guess--who knows.  Just a personal journey.

The long game is to just keep working on these relatively simple watches, honing my fundamental skills so that sometime next year I will be confident enough to fix my Valjoux GMT Chronograph which was given to me by my Dad and broke it last year.  It is not the end game but it will be a local maxima.

The tool links are nice.  I want more tools.  Pondering a Sherline milling machine.  The beat goes on...

mikepilk Distinguished Member

mikepilk

Posted
Posted

I once calculated that, with a 18000 /h balance, 1ms beat error is about a 5 deg angular error (less for a faster balance) - which is about the angle between the two red lines.  So it's a very small adjustment !

Without an adjustable stud holder, I set it up by eye, looking along the pallet to the impulse jewel. Despite the jewel looking dead centre, I can be out by over 1ms. If you look at the distance between the lines at where the impulse jewel would be, that explains why !  

2.thumb.jpg.31437c27c340d10f7364ab494439010e.jpg

JohnR725 Distinguished Member

JohnR725

Posted
Posted
2 hours ago, mikepilk said:

Without an adjustable stud holder

I am being confused with your pictures? Just to clarify the watch that does not have an adjustable stone holder is the 2789?

Ronp Member

Ronp

Posted
Posted

If you at the hairspring the endhook is looks  flatter than it should be, the bent should be much more pronounced with 2 relative sharp angles, here it is almost fluent. This obviously will make the zero position of the balance different, hence impact the beat, likely causing the observed non-correctable a beat error with the stud alone.

mikepilk Distinguished Member

mikepilk

Posted
Posted
44 minutes ago, JohnR725 said:

I am being confused with your pictures? Just to clarify the watch that does not have an adjustable stone holder is the 2789?

The 2789 does have an adjustable stud holder. I was pointing out the difficulty of setting the beat accurately on movements that don't

JohnR725 Distinguished Member

JohnR725

Posted
Posted
2 minutes ago, mikepilk said:

The 2789 does have an adjustable stud holder. I was pointing out the difficulty of setting the beat accurately on movements that don't

I'm happy you know that I was being really confused here.

Then yes the visual method is interesting in that as you already noticed visual perfect is not timing machine perfect. I remember when I got my witschi timing machine I was all excited and had a Hamilton deck watch which is a really beautiful timepiece and visually it was perfect in beat. But the timing machine was disagreeing with my visual assessment by giving me numbers that were very unhappy numbers. I remember it took me at least 45 minutes to please the timing machine and yet visually it still looked the same.

This also came up at work once where with a little tiny Omega watch visually it looked fine but once again the timing machine was laughing hysterically at us. But on some watches Older watches in particular you might find a little tiny dot On the balance rim. Is there to tell you where the stud is supposed to be. In the case the Omega we just rotated it so what was where was supposed to be than the timing machine was happy. But visually you can see no difference at all.

 

  • Like 1
 Share
Go to topic listing
×
×
  • Create New...