Strange behavior in rete

[szbalogh]

This is my chinese GMT Matster homage watch wearing it all day long now for 3-4 months. Inside is a DG2813 with nonadjustable 24h hand and date complications. Serviced myself, which does not men it is properly done Moreover, it has built-in autowinding mechanism which means i have to align 5 pinions properly at the same time. So, it was tricky putting it together. Also broke one of the icabloc springs which i replaced with a different shaped one.

The problem is that the watch is keeping excellent time for 3-4 weeks and then comes a big loss in the rate. Then i adjust the hairspring preload. This happened 3-4 times as i remember. Last week it made -2,5 s/d in average, deviation was ~15 s/d. Now today it fell down to -80 - -100 s/d. What can do such sudden jumps in the rate? Watch seems healthy.

[clockboy]

Its a strange one as the amplitude is good. When you fitted the shock spring did you check it is secure i.e. holding the cap jewel in all positions. 

[szbalogh]

7 minutes ago, clockboy said:

Its a strange one as the amplitude is good. When you fitted the shock spring did you check it is secure i.e. holding the cap jewel in all positions. 

How can i check the cap jewel? It was ok as i remember. 

But it seems that it is more stable now in the different positions. 

 

[jdm]

1 hour ago, szbalogh said:

Then i adjust the hairspring preload. This happened 3-4 times as i remember. Last week it made -2,5 s/d in average, deviation was ~15 s/d. Now today it fell down to -80 - -100 s/d.

You adjusted what? There is no preload in the airspring. If the rate changes that means the effective hairspring length has changed. 

[szbalogh]

29 minutes ago, jdm said:

You adjusted what? There is no preload in the airspring. If the rate changes that means the effective hairspring length has changed. 

Ah sorry, not perfect in english, google suggested me "preload"  I wanted to say "effective hairspring length". 

The regulator is hard to move, i am sure it couldnt not move. 

[jdm]

12 minutes ago, szbalogh said:

The regulator is hard to move, i am sure it couldnt not move. 

It must move, otherwise the rate can't be regulated. Also, if it has an etachron style pin, that is used for regulation and adjustment also.

Edited May 14, 2016 by jdm

[szbalogh]

4 minutes ago, jdm said:

It must move, otherwise the rate can't be adjusted. Also, if it has am etachron style pin, that is used for regulation and adjustment also.

No etachron pin, simple as possible movement. 

Meanwhile made more measurements and now it has simply no deviation in rate in the different positions. Day before yesterday i had this report:

It seems that the jewel or the bridge was moved indeed. Maybe the balance was not touching the cap jewels with both ends. 

Assembled the measuring system this week and now i see strange things

[szbalogh]

The exact same rate as yesterday. Definitely better stability than before.