Subtle timegraph issues?

[dieale2]

I am using tg-timer to time an Unitas 6340. The ebay seller said it was serviced. Apart from the slow rate, it actually looks pretty good. However, I noticed some subtle problems with the waveform of the graph. First, on the bottom panel, there is an extra noise detected. According to this, something is glazing in the balance. Second, in the middle panel, right before the pallet locks the wheel, there is an extra noise. Does that mean the balance is hitting something it's not suppose to?

I am new to watch repairing so I am wondering if I am reading this graph right.

[dieale2]

Here's another timegraph of a watch. It's a vintage Russian watch that I bought from ebay. It looks like it obliviously needs a rebuilt from the irregular chart. It doesn't say it's been serviced and you can see the results.

[jguitron]

May be the case but there’s not tons of info on sound wave analysis. I’ll leave it to more experienced guys to comment.

I’d stick to the standard rate, BE and amplitude on a few positions.

I would go chasing the sound waveform if all else looks fine.

Of course, just MHO.

Cheers!


Sent from my iPhone using Tapatalk

[wls1971]

There are a number of things that can cause this You need to carefully observe the balance wheel under a loupe is it untrue? and catching somewhere? pallet fork bridge maybe, is the hairspring out of shape and catching the center wheel ? 

careful observation of the balance may make the cause obvious.

Timegrapher's can show you symptoms  but quite often you have to do the diagnosis.

It should also run better than -90 seconds a day

 

[jguitron]

May be the case but there’s not tons of info on sound wave analysis. I’ll leave it to more experienced guys to comment.

I’d stick to the standard rate, BE and amplitude on a few positions.

I would go chasing the sound waveform if all else looks fine.

Of course, just MHO.

Cheers!


Sent from my iPhone using Tapatalk

Sorry, meant “I wouldn’t go chasing...”





Sent from my iPhone using Tapatalk

[JohnR725]

I rather liked something that my instructor said when I was in school don't please the timing machine. Timing machine points you in directions gives you clues but perhaps think of it more as reading a crystal ball it's not always an exact science. Then when you toss in the Oscilloscope part and start trying to make interpretations from that things get much more complicated. If you get obsessed about having perfect waveforms which are going to be absolutely impossible to have you going to be chasing your tail forever.

Then you have a rather impressive amplitude which is a little excessive?

Then realistically to do proper analysis of a watch it needs to be done in multiple positions. Problems don't always show up when the watch is in only one position. So what tends to work well is six position timing. So this is dial-up and dial down and Use of the crown is used as a reference which would be crown up down left and right.

[praezis]

Hi,

I will comment the picture in the first post: 

- the extra noise in the bottom is most probably a ringing hairspring. It comes from sti:cky regulator pins and does no harm (often you can hear this ringing).

- there is no extra noise in the middle, waveform is absolutely normal. The tick noise can contain three, four or more single pulses. If you wonder where they come from, study books dealing with escapements.

What else you can see in the same picture: your TM does not recognize the first pulse of the noise! A common issue of certain TMs,  A Weishi user has no chance to recognize this faulty measurement, your scope picture  fortunately shows it. Hence the too high amplitude value, what JohnR already mentioned.

Frank

[jdm]

1 hour ago, praezis said:

What else you can see in the same picture: your TM does not recognize the first pulse of the noise! A common issue of certain TMs,  A Weishi user has no chance to recognize this faulty measurement, your scope picture  fortunately shows it.

A recent Mark's video documents that: A fault not revealed by a standard time-grapher.