• The online Bulletins and Mart and Highlights are currently unavailable due to a failure of a network piece of equipment. We are working to replace it and have the Online publications available as soon as possible. Thank you for your patience.

Timegrapher question

Alex Hamilton

NAWCC Member
Jan 29, 2021
22
3
3
61
Carolina Beach
Country
Good afternoon all,
My first question is, when using a Witschi timegrapher, does the top line represent the entry pallet stone and the lower line represent the exit pallet stone?
This is what I've always assumed, but I would like to hear from someone who knows for sure.

My second question is about the "period setting" of 2-60 seconds.

My understanding is that the "period setting" is the length of time that the amplitude information is gathered and then the average is calculated.
Also, is there an avantage in one "period setting" over another, or is it dependent on the possible faults in the trace lines.
Any thoughts would be appreciated.
 

Chris Radek

NAWCC Member
Apr 13, 2014
928
411
63
Lincoln, NE, USA
timeguy.com
Country
There are two lines instead of one if the watch isn't exactly in beat. This is because one pallet's sound happens slightly earlier than the other. The earlier one is on top. It can be either one. The timer doesn't know which sound is exit/entrance because they sound the same and the timer can only listen.

But you can look at the watch and tell. With a little practice you can see which pallet is making the top line and use this to determine which direction out of beat the balance is.
 

Alex Hamilton

NAWCC Member
Jan 29, 2021
22
3
3
61
Carolina Beach
Country
There are two lines instead of one if the watch isn't exactly in beat. This is because one pallet's sound happens slightly earlier than the other. The earlier one is on top. It can be either one. The timer doesn't know which sound is exit/entrance because they sound the same and the timer can only listen.

But you can look at the watch and tell. With a little practice you can see which pallet is making the top line and use this to determine which direction out of beat the balance is.
 

Alex Hamilton

NAWCC Member
Jan 29, 2021
22
3
3
61
Carolina Beach
Country
Hi Chris,
As always your input is much appreciated.
I am familiar with the relationship of beat error and the separation of the lines. My question, which may not have been framed very well, has more to do with me being able to identify which side of center a fault is coming from when one line is normal and the other line is showing a fault.

It is my understanding that the timegrapher registers 3 distinct sounds in the escapement. The first sound being the impulse jewel hitting the fork slot as it comes in. The second and lightest sound being a combination of the EW foot hitting the entry stone as it slides and the impulse jewel hitting the other side of the fork slot. The third sound being a combination of the EW tooth hitting the locking face of the exit stone and the fork lever hitting the banking pin, which is the loudest of the 3 sounds.

Logically, it seems that with a 3 distinct sound strengths that it would be able to distinguish which is which.
Is it just not written into the software?
 

gmorse

NAWCC Member
Jan 7, 2011
12,830
2,345
113
Breamore, Hampshire, UK
Country
Region
Hi Alex,

Putting it another way, the three major sounds are the unlock, then the impulse and finally the drop onto the next tooth. The timing machines usually start the analysis with the unlock, ignore the impulse and finally use the rather louder sound of the drop to establish the rate. Not sure if you're referring to a Weishi, (the Chinese ones), or a Witschi, but this document from the latter is useful.

Regards,

Graham
 

Attachments

Alex Hamilton

NAWCC Member
Jan 29, 2021
22
3
3
61
Carolina Beach
Country
Hi Graham,
I am familiar with the Witschi Training Course that you attached. Thank You. In the screen shot I have attached from page 15, it refers to the the fault on the top line as" Entry Pallet clips poorly", which is why I have assumed the top line represents the entry side of the pallet fork.
I have just not able to find any confirmation on this in any other documentation.

Top Line entry jewel.JPG
 

Chris Radek

NAWCC Member
Apr 13, 2014
928
411
63
Lincoln, NE, USA
timeguy.com
Country
No, your assumption is not correct. I tried to say that above, but sorry I didn't explain adequately. The top line is whichever pallet makes its noise earlier. (Worrying about the 3 sounds per screen dot is a distraction and has nothing to do with your main question.)

That explanation in the book is extremely bogus. What does "clips poorly or is smeared" even mean? And even the action to take is bogus. If one pallet is working right and the other isn't, why would you need a new escape wheel? That makes no sense at all.

Here is something you can try: turn on the speaker so you can hear the clicks. Watch the dots appearing on the screen. When one appears on top, say UP and when one appears on the bottom say DOWN. So along with the clicks say UP DOWN UP DOWN UP DOWN etc. Get the rhythm in your head. Now while you are continuing to say that along with the clicks, look at the watch. See which pallet is in action when you say UP and which is in action when you say DOWN. (don't try this on a 28800 watch the first time, but once you practice the method you can do it just fine.)

If you have trouble seeing the pallets, just look at the hairspring. Is it expanding or contracting when you say UP? See how this gives you the same information? Now you know which line is which pallet.

You also know everything you need to deduce which way the balance needs to be rotated to decrease the beat error. Then after you turn it, you can do the same test again, and tell if it's now the other way because you overshot. No guesswork.
 
  • Like
Reactions: gmorse

Alex Hamilton

NAWCC Member
Jan 29, 2021
22
3
3
61
Carolina Beach
Country
No, your assumption is not correct. I tried to say that above, but sorry I didn't explain adequately. The top line is whichever pallet makes its noise earlier. (Worrying about the 3 sounds per screen dot is a distraction and has nothing to do with your main question.)

That explanation in the book is extremely bogus. What does "clips poorly or is smeared" even mean? And even the action to take is bogus. If one pallet is working right and the other isn't, why would you need a new escape wheel? That makes no sense at all.

Here is something you can try: turn on the speaker so you can hear the clicks. Watch the dots appearing on the screen. When one appears on top, say UP and when one appears on the bottom say DOWN. So along with the clicks say UP DOWN UP DOWN UP DOWN etc. Get the rhythm in your head. Now while you are continuing to say that along with the clicks, look at the watch. See which pallet is in action when you say UP and which is in action when you say DOWN. (don't try this on a 28800 watch the first time, but once you practice the method you can do it just fine.)

If you have trouble seeing the pallets, just look at the hairspring. Is it expanding or contracting when you say UP? See how this gives you the same information? Now you know which line is which pallet.

You also know everything you need to deduce which way the balance needs to be rotated to decrease the beat error. Then after you turn it, you can do the same test again, and tell if it's now the other way because you overshot. No guesswork.
Thanks for taking to time to follow up.
What you say makes sense. I'll try the UP/Down method and see if I can make that work.
 

John Runciman

NAWCC Fellow
NAWCC Member
Aug 13, 2003
512
85
28
Seattle, WA
Country
Region
There are several unfortunate problems with using a timing machine. Which line is on top as stated above is just whichever line is on top in other words it was the first sound the timing machine received that are processed so it could be either one. Conceivably every time you take the watch off and put it back on again they could be just the opposite it's random basically.

The beat on the timing machine is interesting for people that are not paying attention. Ideally you visually want to make sure the Watch is reasonably close in beat before playing with the beat adjustment if the watch has one. The problem is beat his just a numeric number doesn't have a plus or minus. It's not like you're adjusting the time we read or slower faster your perfect you don't know which side you're on and if you're not paying attention you go right past zero and be hopelessly confused.

Technically there are five sounds and the timing machine only needs to do with these to get the numbers it would like to display forests. To understand this I have a link to a video ignore the purpose the video until you get to about 1:13 where you see the most important sound of all for all timing machines the initial sound of the roller jewel hitting the fork. But you'll notice the first and second sound almost overlap but not quite. Then the next signal we need is the locking Which occurs at about 1:16 it also notice there is an overlap there. In the ideal perfect world the initial sound which they label is one is very very quiet then the second sound that we need is number four which is nice and loud. But if you have access to a timing machine that displays the oscilloscope it's not always that simple and that creates a problem.

The rather expensive timing machine I get the use at work has a nice on oscilloscope so I can see what the waveform looks like and an added bonus the machine will tell me what it's trying to use for the second signal. Typically on very low amplitudes especially with the Chinese machine it will get confused with anything in the middle and think that's the signal because it's occurring approximately at the right time and that will mean that visual your watch looks like it's not running very well at all and the timing machine will give you a very happy amplitude.

Unfortunately that's not limited to Chinese machines which is why it's nice witschi has the On oscilloscope. So sometimes when I see my numbers fluctuating too much I switched on oscilloscope mode and I can see the problem. The timing machine is triggering off something in the middle it's giving a higher amplitude and then randomly going back to where it's supposed to be say can see the numbers randomly changing sometimes changing the amplification of the machine solves that problem other times it's a problem watch


My second question is about the "period setting" of 2-60 seconds.

My understanding is that the "period setting" is the length of time that the amplitude information is gathered and then the average is calculated.
Also, is there an avantage in one "period setting" over another, or is it dependent on the possible faults in the trace lines.
Any thoughts would be appreciated.
Sometimes the Chinese machines manuals are sparse and rather cryptic. the purpose this number is for the averaging of the numeric display. In order to keep numeric display readable you do not want to look at the instantaneous of every single measurement. So what they do is they average the numbers. The Chinese machines do the averaging different than witschi does them but they still do an averaging.

To test this out look at a watch look carefully at the graphical display notice how the dots are not perfect they will vary a little bit depending upon the watch in a variety of things. Set your averaging time to two seconds and notice how the numbers change. Then set it to 60 seconds notice they're nice and stable in every 60 seconds they change. The solicitor usually recommended 20 seconds when you're measuring watches music gives a stable display.

The only minor problem with the 22nd rule is the witschi machines which they're usually referring to average differently than the Chinese machines. Which means conceivably if you were running a test with one watch the Chinese machine in a witschi machine you'll see possibly slightly different numbers because of how they average and if you of more than one machine on a watch there also averaging over different parts of the waveform but overall if you increase the averaging number you get a more stable reading.
 

Chris Radek

NAWCC Member
Apr 13, 2014
928
411
63
Lincoln, NE, USA
timeguy.com
Country
There are several unfortunate problems with using a timing machine. Which line is on top as stated above is just whichever line is on top in other words it was the first sound the timing machine received that are processed so it could be either one. Conceivably every time you take the watch off and put it back on again they could be just the opposite it's random basically.
This is wrong, and isn't what I meant to say. Sorry, I am still not being clear. Let me try to explain a third way: Have you seen a mechanical timer like an old Vibrograf, with a spinning drum with a helix on it? It causes a dot to sweep in a consistent direction at a consistent speed matching the expected period of the watch's ticks. When the watch ticks the print bar smacks it to print a dot on the paper at that spot. So the earlier things are to one side, the later things to the other side. That's all there is to it, that's how the entire machine works. It's so brilliantly simple that they were made before computers and even transistors. And we can deduce so many things from the printout!

The modern machines emulate that because those are the readouts we all learned to read decades ago. Now they are still "sweeping" from top to bottom and making a dot when they hear a noise. So earlier things are up, later things are down. If the measured period since the last dot was printed is less than nominal, this new dot will be higher than the last. If it's more than nominal, this dot will be lower. That's why the trace moves upward if the watch is gaining and downward if it's losing!

Try the experiment I described above with your machine. If you have two lines it's because there's a beat error, and the upper one will be the pallet following the slightly shorter period, and the lower one will be the pallet following the slightly longer period. You can take the watch off and put it back on and it will be the same.
 

John Runciman

NAWCC Fellow
NAWCC Member
Aug 13, 2003
512
85
28
Seattle, WA
Country
Region
This is wrong, and isn't what I meant to say. Sorry, I am still not being clear.
I am confused? Which one of us is wrong you me both of us neither of us?

Now I think were both on the same page but I am confused by why am wrong? The timing machine doesn't know the entry from exit sound it only knows a sound. It receives a sound and it puts a mark on the display. Then it goes forward a little bit and the next sound forms the next line Then this alternates back and forth as it goes down the screen. With the separation representing how much out of beat you are.

This means that which ever signal it receives first it puts on top but that doesn't mean that the top is always the entry or exit stone because the machine cannot tell that at all as there is no difference only whichever signal it received first.
 

Chris Radek

NAWCC Member
Apr 13, 2014
928
411
63
Lincoln, NE, USA
timeguy.com
Country
I am not sure how else to explain it. I'm afraid my explanations are too long and saying more will not help. Have you carefully read everything I wrote and then tried the experiment I suggested? If not, please do! If you think a certain point is unclear maybe I can help.

The timing machine doesn't know the entry from exit sound it only knows a sound.
This is true!

This means that which ever signal it receives first it puts on top
This is false! It's not whichever sound it hears first.

doesn't mean that the top is always the entry or exit stone because the machine cannot tell that at all
This is true!

there is no difference only whichever signal it received first.
This is false!

When your watch is out of beat, there will be a tick, then a slightly shorter period, then a tick, then a slightly longer period. If it's bad enough you can even hear it yourself, we're all familiar with that. So the two ticks are different!

Because of how all these machines work, the tick following the shorter period (whichever it is, entry or exit) will be on top. The one following the longer period (the other one, exit or entry) will be on the bottom. You can use this fact to even determine which way the watch is out of beat, using the technique I described previously. You can also determine which pallet isn't working right if one of the lines shows a problem (and this was the original question in the thread.)
 

John Runciman

NAWCC Fellow
NAWCC Member
Aug 13, 2003
512
85
28
Seattle, WA
Country
Region
I'm being hopelessly confused by this I'm not really very bright about these things it appears?

When your watch is out of beat, there will be a tick, then a slightly shorter period, then a tick, then a slightly longer period. If it's bad enough you can even hear it yourself, we're all familiar with that. So the two ticks are different!
I guess here is what my problem is the ticking sound is what I perceive the sound to be the absence of sound in between I regard as the absence of sound in between not part of the tick. I'm attaching some images that I swiped from witschi. So my problem is that both texts are identical and there only different because of the amplitude but they both be equally different as the amplitude changes. The spacing or the absence of sound that's controlled by the beat.

Because of how all these machines work, the tick following the shorter period (whichever it is, entry or exit) will be on top. The one following the longer period (the other one, exit or entry) will be on the bottom. You can use this fact to even determine which way the watch is out of beat, using the technique I described previously. You can also determine which pallet isn't working right if one of the lines shows a problem (and this was the original question in the thread.)
More confusion on my part. Okay we use your method we figure out the top line as they exit stone but this only works if the watches out of beat. If it's in beats you have two lines that exactly overlap and you can't tell one line from the other. Then you were out of beat you know the top lines the exit you move your stud lines overlap and The lines are now separated but this would mean that the top line has to switch over to the bottom position and then you'll be confused as a witch's line is the exit stone or not?

I am familiar with the relationship of beat error and the separation of the lines. My question, which may not have been framed very well, has more to do with me being able to identify which side of center a fault is coming from when one line is normal and the other line is showing a fault.
Thank You. In the screen shot I have attached from page 15, it refers to the the fault on the top line as" Entry Pallet clips poorly", which is why I have assumed the top line represents the entry side of the pallet fork.
I have just not able to find any confirmation on this in any other documentation.
Oh it's not that simple? For instance the image indicates it's silly description but could it be anything else? Did you know if your roller jewel is looser or has a defect that will show up in the lines

Basically one line shows one side of the escapement has a problem and one side does not have a problem. The escape wheel is common to both lines and it shouldn't be a problem here so their diagnosis is not quite correct. On the other hand witschi is Swiss in Swiss service centers if they have a problem they just replace a part. If they don't like anything they just replace it they have an infinite supply apart so they can replace everything. Plus from their own description is indicating a pallet fork problem shouldn't you change the pallet fork or fix the problem? You can even see discrepancies with the lines if the lubrication isn't even even if one of the stones is little more dry than the other that will show up on the timing machine affecting one of the lines.

My first question is, when using a Witschi timegrapher, does the top line represent the entry pallet stone and the lower line represent the exit pallet stone?
This is what I've always assumed, but I would like to hear from someone who knows for sure.
By the way was this just a theoretical question or do you have a watch that's having a problem?

I swiped something out of another timing machine manual notice they do not recommend changing the escape wheel If one of the lines is having a problem.

timing machine amplitude.JPG timing machine beat.JPG timing machine faulty escapement.JPG
 

Forum statistics

Threads
168,796
Messages
1,472,769
Members
48,587
Latest member
Helplessfanatic12
Encyclopedia Pages
1,060
Total wiki contributions
2,955
Last update
-