Mystery - Regulation thrown off by case back?

[Simeon]

I have just picked up a vintage soviet Zarja watch and I have been inspecting it prior to putting it on the 'to be serviced' pile.

When I measured it with the timegrapher app, I found it was gaining loads of time. Taking the case back off, I find it's a tiny movement within a relatively large watch and it has a separate steel cover over the movement which presumably holds it in place against the caseback.

 

Checking the app again without the cover, I find it is actually running reasonably well for an untouched old watch. Putting the cover and the caseback back in place and it reads as massively gaining time again. This behaviour is consistent every time I take the cover on and off.

I have checked the balance cock is fixed and the screw is tight, so I assume that the cover and case back are pinching the movement in some way and making it run fast. Not sure I can see how that works because if the cover is touching the balance, surely it would make it stop not run faster? It can't be touching just the hairspring to make it effectively shorter. 

Placing only the cover in place or placing just the caseback in place (not even snapping it shut) both cause the fast running. Pretty sure there are a couple of millimetres clearance between the case back and the movement when the cover isn't there. I have also checked the watch, cover and case back for magnetism with another phone app but none of them seem to have an appreciable magnetic field. Perhaps I need to preemptively demagnetise and get a real compass to check them with. 

I appreciate that this is just an app using the built in microphone on my phone - could it be misreading a ringing / reverberation of this cover to give the wrong result?

Observing the watch just running on the bench closed up, it does seem to be appreciably gaining time. I will wind it, set the time and leave it running overnight and see how it does with no cover / back in place. Perhaps I will also see how it performs on the wrist to rule out misreading by the app 

 

What am I missing here?

[tomh207]

I think the app is misinterpreted the noise because of the extra case construction 

 

Tom

[RichardHarris123]

I think the metal cap for want of a better word is soft iron to protect from magnetism.  It is also the only thing holding the movement in the case. 

[Knebo]

5 hours ago, tomh207 said:

I think the app is misinterpreted the noise because of the extra case construction

That is definitely a possibility. Even with a proper timegrapher, you sometimes need to adjust the "gain" settings to get a proper reading depending on the thickness of a case. 

Still, I wouldn't rule out magnetism. Also sounds plausible to me. The Chinese blue demagnetisers are cheap and worth getting (IMHO). 

[spectre6000]

Easy test for the case noise theory: The app told you how much it should or shouldn't be gaining over the course of a day right? Set it, and see how much it gains. If it goes super fast, something is up with the movement. If it's only kinda fast, it's noise.

[nevenbekriev]

Try to demagnetise the whole thing, especially the back and the springy cover.

[Simeon]

I left it running dial up, fully assembled for 24 hours+ and it maybe lost 30 seconds. Not bad for an unserviced old watch.

I am going to call it for 'app misreading the watch'

[mikepilk]

This is a common problem,  miss-readings when the watch is cased, especially if the movement is not held rigidly in the case. When testing cased watches in my Weishi 1000 (you cannot adjust the gain), I put the crown against the outer metal post  of the holder, as the stem and crown can transmit a clearer sound from the movement. Even so, I have some watches (like a Seiko Divers watch)where I cannot get a sensible reading from cased movement.

Edited June 5, 2024 by mikepilk