Beat error problem

[CaptCalvin]

When I change the watch from a flat position to a hanging position, I see the beat error gradually increase as the amplitude decrease, from 0 in the flat position to about 8ms in the hanging. That is extreme. What can this mean?

[saswatch88]

Slightly Bent pivot, cracked balance pivot jewel, too much side shake, hairspring coils sticking......It’s normal to see beat error and amp drop in the 4 positions other than DD and DU just because of gravity. But I would agree 8ms is a lot. Do you hear any galloping in the pallet or metal on metal sound when holding in that position. Problem could be in the escapement too. Hold it up to your ear and listen, move your head from side to upright and listen for a change in the tick. this how the old schoolers did it back in the day before all this fancy schmancy electrical equipment we use today. Lol I still do this even though I have a timegrapher.

[CaptCalvin]

If there's any extra sounds I would imagine the time grapher would display some rather dirty lines. But in my case there is one extremely clean line, and when I change over to a hanging position, the line separates into 2 very clean lines with the gap growing wider and wider until the beat error reads 8 or so.

[CaptCalvin]

Ok this was on me. The beat error in the flat position was not in fact 0 but actually somewhere around 5 or 6.  I hadn't bothered to check the actual number and adjusted the beat error just looking at the lines. I hadn't thought about the possibility that I can make it so the lines can go off screen, appear on the opposite side, converge and give the illusion of perfect beat. Watch is now behaving normally after correcting this simple mistake. I spent the entire afternoon re-servicing the movement just to find this out.

[saswatch88]

2 hours ago, CaptCalvin said:

Ok this was on me. The beat error in the flat position was not in fact 0 but actually somewhere around 5 or 6.  I hadn't bothered to check the actual number and adjusted the beat error just looking at the lines. I hadn't thought about the possibility that I can make it so the lines can go off screen, appear on the opposite side, converge and give the illusion of perfect beat. Watch is now behaving normally after correcting this simple mistake. I spent the entire afternoon re-servicing the movement just to find this out.

Lol it happens. I def thought that was weird to go from 0-8 I would believe 5-8 though. Glad you figured it out.

[JohnR725]

I don't suppose you want to give us a clue of how much amplitude you lost? Typically when amplitude goes superlow everything else flies out the window and is not a problem until you get the amplitude backup to at least 200°.

Then if you using the Chinese like the 1000 machine it has a rollover error so if you go past about 4 ms on the graphical display then graphically it's not going to look right you need to look at the numbers. We've discussed this somewhere else in the discussion group before several of us tried experiments. So this is why usually you want to go with the numbers they are usually right unless you go to an extreme and then they can rollover also although usually they just go to zeros

 

[oldhippy]

Check the endshake. Then go all though the above what other members have suggested. 

[CaptCalvin]

1 hour ago, JohnR725 said:

I don't suppose you want to give us a clue of how much amplitude you lost? Typically when amplitude goes superlow everything else flies out the window and is not a problem until you get the amplitude backup to at least 200°.

Then if you using the Chinese like the 1000 machine it has a rollover error so if you go past about 4 ms on the graphical display then graphically it's not going to look right you need to look at the numbers. We've discussed this somewhere else in the discussion group before several of us tried experiments. So this is why usually you want to go with the numbers they are usually right unless you go to an extreme and then they can rollover also although usually they just go to zeros

 

I don't remember what it was but I'm pretty sure it was in the 250-280 range. Nothing that would be alarming.

39 minutes ago, oldhippy said:

Check the endshake. Then go all though the above what other members have suggested. 

It turned out to be a very simple mistake on my part as I have explained in my last post. There was nothing wrong with the movement.

[mikepilk]

2 hours ago, JohnR725 said:

I don't suppose you want to give us a clue of how much amplitude you lost? Typically when amplitude goes superlow everything else flies out the window and is not a problem until you get the amplitude backup to at least 200°.

Then if you using the Chinese like the 1000 machine it has a rollover error so if you go past about 4 ms on the graphical display then graphically it's not going to look right you need to look at the numbers. We've discussed this somewhere else in the discussion group before several of us tried experiments. So this is why usually you want to go with the numbers they are usually right unless you go to an extreme and then they can rollover also although usually they just go to zeros

 

This has caught me out before - one line had rolled over and looked to converge.  I use a Weishi 1000 Timegrapher.

Now I always do a visual check on the alignment of the impulse jewel with banking pins and pallet/escapement jewels whilst assembling the movement. I can usually get it within 1 ms that way.

[JohnR725]

8 hours ago, mikepilk said:

Now I always do a visual check on the alignment of the impulse jewel with banking pins and pallet/escapement jewels whilst assembling the movement. I can usually get it within 1 ms that way.

If everybody would do that they would save a lot of time on the timing machine. Seeing as how the modern watch with the floating stud if you only looking at the timing machine you can easily go right past where you're supposed to be a then ponder why the watch never goes in beat.

1000 timing machine is a great machine you just have to pay attention that the numbers agree with the display and visually agree with the watch.

[jdm]

17 hours ago, CaptCalvin said:

If there's any extra sounds I would imagine the time grapher would display some rather dirty lines. 

Not really. I had a case of a pallet bridge slight rubbing and lines were quite straight but timing was unstable. Only listening and looking at the waveform helped in diagnosing the problem. The importance of doing that was pointed out to me by Master Mark.