Weird timegrapher reading

[IFELL]

Hi, I’m servicing vintage Venus 203 movement, triple calendar with moonphase, and on reassembly I get this reading:

it is like 4 rows of the usual read... I demagnetise again and get the same reading, then I slow down the balance and now I see this:

Any clue?

Thanks.

[Klassiker]

I would have suggested knocking, but if your timegrapher is to be believed, then it's not that with only 229 to 241 degrees amplitude. Maybe the hairspring is rubbing? Is it the same in all positions? What did you do to slow down the balance?

Out of interest, why are you running your timegrapher on monochrome? It's a Weishi 1900 with colour display, isn't it?

 

[IFELL]

Hairsping seems fine, but it may be bit old. I will check other positions, my main suspect is now the balance staff. The balance wheel seems to be doing some up and down movement, I am suspecting that the staff’s pinion is short or damaged.
I slowed the balance down by moving the regulator as usual.

What do you mean by ‘running timegrapher on monochrome’? It is Welshi 1900.


Sent from my iPhone using Tapatalk

[Deggsie]

He means that the weishi model you have is capable of showing a colour display. You appear to have it in mono. Makes no difference when problem solving your issue though. Check the machine settings are ‘auto detect beat rate’. What you are seeing there looks like the machine has mistakenly sensed the movement as 18000 bpm. Please let me know. Regards Deggsie

 

 

Sent from my iPhone using Tapatalk

 

[Nucejoe]

Your suspicion of the staff may well be correct.Timegrapher readings of  Face up&down, stem up positions reveals more.

With this type of end stones ,Easier to just remove the balance for visual inspection of possible pivot damage.

I can,t tell by the pic if your device is high magnification capable, if you have no access to high magnification, you might want to try this approach I know to check for pivot damage. Moving upper and lower jewels inward can compensate for shortness of pivots, again if don,t have tools to move the jewel, I know of this technique for moving the jewels, 

Shimming the cock outbound is worth giving a try.

[Klassiker]

According to Ranfft, this movement does run at 18000 bph, but the traces are what you get if you are slightly out on the setting. For example, I've got a 18000 bph movement on my Weishi 1900 here, and I set the beat on the machine manually to 17786 bph. The display looks very similar to the second picture in the original post. Doesn't help with the problem, but interesting all the same.

[Deggsie]

Ok, back to basics then - without the timegrapher, simply measure the accuracy over say 3 minute period. What are you obtaining please? What you see on display is a clear ‘tick and tock’ of the escapement but the frequency is way out, as if the rate is extremely high. By the way I would adjust the movement first then worry about setting the rate. That is to say the beat angle needs to be closer to 0.0 ms than it is currently. Let us know how you get on. Kind regards.


Sent from my iPhone using Tapatalk

[jdm]

15 minutes ago, Deggsie said:

 By the way I would adjust the movement first then worry about setting the rate. That is to say the beat angle needs to be closer to 0.0 ms than it is currently.

If the machine doesn't display rate, something is very wrong. like the dots are too much spaced even when the line is almost horizontal, as in the last picture. A 3ms beat error would not cause that.

I would record sound and post it here, so it can diagnosed by ear and software in "scope" mode.

 

Edited March 31, 2019 by jdm

[Deggsie]

If the machine doesn't display rate, something is very wrong. like the dots are too much spaced even when the line is almost horizontal, as in the last picture. A 3ms beat error would not cause that.
I would record sound and post it here, so it can diagnosed by ear and software in "scope" mode.
 

Back to basics... what is the accuracy over say 3 min period? I was always taught ‘if you hear hoofs, don’t jump to the conclusion it’s a zebra’




Sent from my iPhone using Tapatalk

[jdm]

19 minutes ago, Deggsie said:

Back to basics... what is the accuracy over say 3 min period? I was always taught ‘if you hear hoofs, don’t jump to the conclusion it’s a zebra’

3 mins is a base too short to manually measure rate. Suppose it's off by 5 mins a day, you would have about 1/3 of sec off in 3 mins, that's too little to accurately measure by hand.
Anyway, even if you measure the rate accurately, that will not tell you what is wrong to cause the weird display. To know that we will have to use our reasoning, timegrapher guides as published by Witschi, as well our ears as Mark showed me in a another case.

 

Edited March 31, 2019 by jdm

[Klassiker]

1 hour ago, jdm said:

If the machine doesn't display rate, something is very wrong. like the dots are too much spaced even when the line is almost horizontal, as in the last picture. A 3ms beat error would not cause that.

I would record sound and post it here, so it can diagnosed by ear and software in "scope" mode.

********************

3 mins is a base too short to manually measure rate. Suppose it's off by 5 mins a day, you would have about 1/3 of sec off in 3 mins, that's too little to accurately measure by hand.
Anyway, even if you measure the rate accurately, that will not tell you what is wrong to cause the weird display. To know that we will have to use our reasoning, timegrapher guides as published by Witschi, as well our ears as Mark showed me in a another case.

 

I think Deggsie's assumption that the rate is extremely high is correct. Like I posted above, I can get a similar looking trace by manually setting the rate in my W1900 to 17786 bph but analysing a 18000bph movement. I'm betting those wide-spaced dots in the second picture are not horizontal traces, but almost vertical. Only IFELL can tell us for sure. A check of the rate over 3 min. or even 10 min. won't hurt and will provide an important clue.

The Weishi 1900 doesn't output the true sound of the movement, unfortunately. No scope mode either. Graph type B will also show steep lines, if the rate is way out. I agree the sound would be useful. Maybe there is another way of recording it?

[Deggsie]

What I should have included in my response is that I have had the same problem on my weishi (same model). These are very reasonably priced instruments for us hobbyists, but are not infallible. Just check the rate fundamentally. If you don’t think 3 mins is enough, go to 5...... hell even 10 if you want, but check the fundamentals first. Once the rate and beat become within a certain band of tolerance the machine is very useful.

If we conclude the rate is not far out, then start delving deeper. As I said, I was always taught not to think zebras just because I hear hoofs.

Kind regards
Deggsie


Sent from my iPhone using Tapatalk

[jdm]

3 hours ago, Deggsie said:

What I should have included in my response is that I have had the same problem on my weishi (same model). These are very reasonably priced instruments for us hobbyists, but are not infallible. Just check the rate fundamentally. If you don’t think 3 mins is enough, go to 5...... hell even 10 if you want, but check the fundamentals first. Once the rate and beat become within a certain band of tolerance the machine is very useful.

Well, the max rate error is on these machines 999 s/day. With a bigger deviation than that just --- is shown. So, I think, the machine is doing its job. I don't doubt that the rate is crazy fast, the OP has to find out why. 

Edited March 31, 2019 by jdm

[IFELL]

Wow, what a deep discussion this developed into )
Well, upon checking the operation in different positions, I discovered that ‘crown up’ the balance wheel started to gradually slow down, I could even hear the roller touching the main plate.

When I ran down the mainspring, the balance wheel had a lot of side and up down movement. So the main suspect is here to blame, shortened of balance staff’s upper pivot.
That on it’s own wouldn’t make the weird reading, but I think as the roller was hitting the main plate, it produced those extra ticks.

I have completely disassembled the balance and ordered new staff from eBay (£8).
To be honest the hair spring has seen better days, but it is centred and doesn’t stick.

I made some pics so you can have a look, will post to this thread when the new part arrives.

Thank you all for your input!







Sent from my iPhone using Tapatalk

[JohnR725]

I snipped out a section of your balance wheel hairspring image. It appears to be somebody has re- pinned your hairspring? Re-pinning the hairspring where they have versus the end where it probably should have been will make your watch go a lot faster.

[Klassiker]

I just looked at your latest photos. The pivots look OK to me. I agree with John above. That is an odd place to pin the hairspring. Have you tried it pinned at the end?

[watchweasol]

Hi  I would personaly re pin the balance spring back to where it was origionly and  set the balance in beat again and re try on the timegrapher. Some one has at sometime re pinned it for some reason probably to speed it up because of an underlying and un diagnosed fault. return to the status quo and then begin at the beginning. the pivots look ok I agree with John and Klassiker