Large swings in amplitude

[AP1875]

I serviced this Asian 7750 movement in the video below. I’m seeing some large fluctuations in amplitude on the timegrapher, largest swing is around 300 - 250 then back up to around 280 and then to 300.

I found this on the witschi site

 

Im seeing some variables in rate on the timegrapher but measuring the time against a digital clock dial up its +2 seconds max over 24hr. The pattern is like a sine wave but not so extreme.

 

Heres the video. It’s best to speed it up by 1.5x I think. As the camera changes angle that’s 15 hours later.

Any help understanding this pattern appreciated.

 

 

 

Edited June 25, 2020 by AP1875

[rodabod]

Is the sinusoidal cycle over a period of 60s? 4th wheel?

[AP1875]

 

1 minute ago, rodabod said:

Is the sinusoidal cycle over a period of 60s? 4th wheel?

Not 100% sure what you’re referring to but the timegrapher test period is set to 12s. Thanks 

[JohnR725]

if you look at the PDF which I've attached number eight and number 14 show variations like you have. Unfortunately you do not have a paper tape machine like the diagram shows. In other words in the old days you let a lot of paper tape spit out the machine and then measure between the peaks and see when they occur. If it's a repeatable pattern as rodabod is suggesting above you can trace it to a particular part. Also as that's a chronograph do not run the chronograph when you're doing the timing that would just add to the complexity of the problem.

try changing the sample time to something shorter because it's doing a running average it's hard to tell at least looking at the numbers they jump too much because it's averaging. It definitely looks like a powertrain fluctuation something is possibly binding up and then releasing but it's hard to tell.

Timing-Machine-Charts.PDF

[Nucejoe]

Presumably you are seeing the fluctuation repeat itself with a consistant rate, if the rate corresponds to period of any wheel in the train, you should revise the suspected wheel. 

Five test periods go by in 60 seconds, 5(12 seconds)= 60 seconds.

I think that is what rodabod talking about.

 

[rodabod]

Yes, I’m referring to the period of the cycle. And I’m not going to calculate that for someone else. 

[nickelsilver]

It does look like a 60 second cycle, likely the 4th wheel is out of round or damaged pivots etc.

 

[Nucejoe]

48 minutes ago, rodabod said:

Yes, I’m referring to the period of the cycle. And I’m not going to calculate that for someone else. 

I saw it as chance to show off the great mathematician in me. LOL

[praezis]

On 6/25/2020 at 12:34 PM, JohnR725 said:

Unfortunately you do not have a paper tape machine like the diagram shows. In other words in the old days you let a lot of paper tape spit out the machine and then measure between the peaks and see when they occur. 

John, 
paper tape machines of the good old days are not necessarily needed to detect wave periods of any length.

I use this in my workshop:

I found that more often teeth are guilty than wheels out of round.

Frank

[JohnR725]

16 minutes ago, praezis said:

paper tape machines of the good old days are not necessarily needed to detect wave periods of any length.

I use this in my workshop:

Personally what I really like is this from work much nicer than paper tape. The problem is pointing out it's really hard to look at the liquid crystal screen over multiple screens and see trends. Even sometimes in the old paper tape days was hard to see things is the timing machines are really meant to be doing that you need a different function. Like the image attached that's a really nice function unfortunately it's on a really expensive timing machine.