Amplitude numbers

[szbalogh]

That end curve seems total wrong to me. It should be concentric around the balance staff. 

Is it like this if put back in the movement? 

It seems to me that the HS end curve is pushing hardly on the outer regulating pin (green):

If so, then the HS cant be flat and centered and sticking to something. This can lead to low amplitude and noises that the timegrapher recognizes false and showing totally wrong rate even if the BPH is selected correctly. 

Great beat error does not lead to such high difference in rate. 

[nickpeh]

Took it off and redo manipulating of hair spring, installed and put on time grapher , now no loss gain of days again.

[JohnR725]

6 hours ago, szbalogh said:

That end curve seems total wrong to me. It should be concentric around the balance staff. 

Is it like this if put back in the movement? 

The problem with your image is we don't really care what it looks like out of the watch it's in the watch where it's important. So it appears to be the terminal curve is too far out. Then we can't tell if they hairspring is bent at the stud unless it's in the watch.

So first two links are talking about people vibrating a hairspring but then talk about forming a terminal curve. You need to restore the terminal curve plus verify your not bent at the stud. So ideally the hairspring should leave the stud form a beautiful curve that exactly corresponds to the curve the regulator pins follow. Ideally they hairspring should be dead center between the regulator pins. Then they hairspring has to be centered over the balance wheel. This is why seeing it out of the watch is meaningless because it's in the watch we need to see.

Images attached one is yours with what looks like a bent hairspring other from one of the links on forming terminal curve's to give you an idea of what you're looking for.

Third link YouTube video showing how to bend the hairspring. Most important thing when bending the hairspring is bending the correct place to fix the problem rather than generating a new problem. So I think you said you replaced the complete balance wheel? If so find the other balance wheel with hairspring practice on that hairspring. Hairspring work requires lots and lots of practice to get it right. Your bend appears to be relatively simple at least from casual looking so if you get carried away with bending it's going to be a nightmare to fix.

Then until you get parallel lines that we recognize all the rest the numbers on the timing machine are worthless. With so many random dots the timing machine is picking up incorrect information the numeric numbers are not necessarily correct. So until we get something for the timing machine to recognizes numbers just tell you have a problem..

https://watchmakingjourney.com/2015/02/14/update-hairspring-vibrated-also-where-ive-been/

http://raulhorology.com/2012/12/the-mark-of-true-watchmaker-hairsprings-part-3-forming-the-terminal-curve/

https://www.youtube.com/watch?v=idO5elKgFMA

ttp://www.ranfft.de/cgi-bin/bidfun-db.cgi?10&ranfft&&2uswk&Orient_46941

[nickpeh]

I suspect the watch being magnetized too, when a compass placed top of crystal , needle seems turn away from original direction. Will magnetism of watch caused such reading? who had try reading a magnetized watch on time grapher before?

[szbalogh]

I can imagine that magnetization can lead to wild patterns depending on the magnitude. On the other hand my experience says that the pattern is regular and flat, just the rate and amplitude is affected. I am working with a really strong magnet on my workplace and need to demagnetize my watch regularly.

[nickpeh]

1 hour ago, szbalogh said:

I can imagine that magnetization can lead to wild patterns depending on the magnitude. On the other hand my experience says that the pattern is regular and flat, just the rate and amplitude is affected. I am working with a really strong magnet on my workplace and need to demagnetize my watch regularly.

Hi， so u think the pattern of the dots on time grapher not affected much by magetism? I am trying to narrow down the causes..

[szbalogh]

Just now, nickpeh said:

Hi， so u think the pattern of the dots on time grapher not affected much by magetism? I am trying to narrow down the causes..

Yes exactly. But demagnetization cant do any harm and one can avoid running in circles with fault identification. As some others suggest one should start with a proper demagnetization. 

Still it would be nice to have a close-up picture or a video (even better) of the hairspring while the watch is in resting position and while working. 

[jdm]

On 2/28/2017 at 0:24 PM, nickpeh said:

hairspring with balance assembled to balance cock

I suggest that you avoid using that so-called tool, because the the HS is not distorted before, it will be later.

[JohnR725]

If this was an older watch with a steel hairspring then magnetism makes the watch go fast very fast. Even if the coils aren't visibly stuck together it screws up timekeeping. On the modern watches (Nonmagnetic hairspring) if the magnet is sitting on top of the watch for those that like to try experiments visibly you can see it will affect they hairspring and definitely screws up timekeeping. So yes magnetism can be a problem but a much bigger problem is if the hairspring isn't exactly where it's supposed to be. The timing machine picks up sounds if they hairspring is rubbing bumping hitting those are additional sounds. Then it's doing any of those things I just described it makes the watch run abnormally fast because you've basically shortened the hairspring.

So link below will answer a lot of questions. So he shows two examples of why a watch runs fast but he also shows more than that. It's important when learning watch repair to pay attention to what the balance wheel looks like and hairspring when it's functioning correctly. So at the beginning of the video and end of the video look at how they hairspring is opening and closing rhythmically as the balance wheel rotates.

So in his first example he talks about the effect of magnetic fields. It's not just on the hairspring if the steel parts are magnetized they affect the watch also depending upon how strongly their magnetized. It's very importance in the example of the first tool to slowly move the watch away exactly as he says otherwise you will magnetized the watch. The second tool simulates the moving away by electronically modulating the field simplistically. So for that kinda machine put the watch on push the button Usually watchmakers obsessed and will rotate the watch and push the button again why do it one time when you can have fun into a multiple of times these machines work really well unfortunately they're expensive.

Then he simulates coils stuck together with oil. Notice with the oil they hairspring does not open and close rhythmically the balance wheel does not oscillate quite right visually you have a problem. You can simulate this problem also by bending the hairspring and not having it where it's supposed to be.

https://www.youtube.com/watch?v=E6_KOcc4eD4

 

 

 

[nickpeh]

thx u guys for the input, after much reading, still narrow down to HS, magnetized movement will not show those dots pattern in pic. I will try again when I  can to see if I can do better HS manipulation.

[rodabod]

I'm not sure how you are manipulating your hairspring, but I suggest that you remove the hairspring from the balance and attach the hairspring to the balance cock, which I have suggested already. Make sure the hairspring sits within the index pins, and that the collet is centered on the upper jewel.Take photos.

[szbalogh]

Yes, as Rodabod  said. 

This is a bad shaped HS. The collar should be directly over the jewels but You should do this check.