Converting 2824 to low beat movement

[vcelkamaja]

I'm working on a project where I need a low beat 21600 movement. I'm using ETA 2824 with:

- balance wheel assembly

- pallet fork

- escape wheel

from low beat ETA 2846.

It seems it all works fine and the movement runs. However I can see some non stability on the timegrapher. It happens when the movement is positioned dial up - timegrapher rate start jumping up and down (not stable). If I move the timegrapher holder just eg 15-20 degrees horizontal or vertical, I'm getting a straight line with +/-5 seconds / day.

It only happens when the movement is facing up. I tried to analyze the problem a bit and it seems it is the balance wheel assembly causing it (if I replace it with the balance wheel from original 2824, while using low beat pallet and escape from 2846, it is perfect - of course running fast, but otherwise beat it a straight line on the timegrapher). I also tried more balances from from different 2846 movement and it is always somehow unstable.

 

Any suggestions what to do?

[nickelsilver]

Sounds like you have too much amplitude and the watch is rebanking. 2824 has a mainspring thickness of 0.125, the 2846 is 0.105mm, that's a big difference (the higher beat rate takes more power). If you have the 2824 spring in there I would be surprised if it didn't rebank.

[vcelkamaja]

Thanks a lot. Yes, there is an original 2824 spring there. Amplitude on the timegrapher shows 355. But why is it only happening with dial up and when I turn it by eg 20 degrees it runs well?

Btw I just found that if it runs for some time (hours) it seems to be ok with dial up. But when I wind it, it starts again jumping.

Is replacing the spring the only option or any other thing which I can fdo about it?

Thanks

 

 

[nickelsilver]

Amplitude should be pretty close to identical in the two horizontal positions, so why it does it dial up and not down is anyone's guess (dirty balance jewel, too much endshake, and a dozen other possibilities). When the watch is in the horizontal positions the weight of the balance is on the convex tip of the balance staff pivot, so friction is very low, so you get max amplitude. Any deviation from horizontal puts more loading on the hole jewels, which means more contact area, so lower amplitude.

 

You could increase the lock by pulling out your pallet stones a little, which will drop the amplitude a bit, but this is introducing an error to correct another error (and you would need to be adept at such an adjustment- a movement of 0.01mm is already a lot- and you can run into other issues pulling them out willy-nilly). The best would be to get the correct mainspring.

Edited November 3, 2022 by nickelsilver

[vcelkamaja]

I have modified two 2824 and it seems to behave similar for both. When dial up - timegrapher is not stable, when I rotate the movement a bit in the timegrapher holder it gets stable.

 

I also checked the amplitude and it is slightly lower when I turn it (from 355 to 338)

 

I think touching the pallet stones is too dangerous for me so I will try to replace one mainspring and see if it helps.

[vcelkamaja]

Ok, so I replaced the mainspring with the one from 2846 and the result is:

 

- amplitude decreased from 355 to ~ 210 - 250 max (I tried to play a bit with stud rotation)

- I'm able to regulate the rate to pretty low +/-2 sec / day - good news

- there are no jumps or non-stabilities - line in the timegrapher but - good news but...

 

- it is very sensitive to the position - if I change the angle of the movement by rotating timegrapher holder the rate changes - eg from +2 sec/day to +40 sec / day

Can this be due a lower amplitude?

Any sugesstion what can try to make it more independent on the movement position / angle?

 

Thanks

[tomh207]

I’m no where near as smart as nicksilver, however I do believe 355 degrees of amplitude is way, way too high. Can you share with us your objective with this modification?

[vcelkamaja]

A friend of mine has an old watch parts and sked me to help him to put it together. He wants it run slow beat and that is why I'm trying to do it.

 

Btw, while thinking about it - if the amplitude is too high with 2824 mainspring (355) would it work to reduce it by twisting slightly the regulator pin on the hairspring? I tried to do it and it seems it somehow works - amplitude went down from 355 to 260-270 and it seems to run ok - flat line on the timegrapher, still somehow sensitive to movement orientation.

Is this a way how to reduce the amplitude or it is a bad idea?

 

In parallel I will also play a bit with the other movement where I replaced less strong mainspring from 2846.