Jump to content

Recommended Posts

16 hours ago, LittleWatchShop said:

Frankly, the reference device is an atomic clock.  So most, if not all, of this thread is OCD.  Interesting...fun...and useful in a vacuum, but at the end of the day, a mechanical movement pales in comparison to digital electronics. So...why bother? 

I don't get your point. If you make or maintain a measurement tool, of any kind, you need a reference for calibration or just validation. It can be more or less accurate but it has to be there, that's Metrology 101.

You can have all the atomic clocks and Network Time Protocol to disseminate time, the issue is how to bring that accuracy to a standalone device and know if it is reliable. 

If not liking this kind of stuff one can get into watercolor painting or other interests that do not require that kind of approach at all. 

  • Like 2
Link to post
Share on other sites
  • Replies 1.2k
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

I'm a watch DIYer, having serviced two watches myself, partially using tools of my own making. One thing I've been working on over the last few months is my own PC based timing machine. I've finally g

So I played a bit in Solidworks, and this is what I got. A friend of mine helped me with the 3D printing.   sup_2.STL mic 2.STL pesa.STL sup_1.STL   the

Dear svorkoetter, These days with my failing eyes and ears I have given up working on watches and confine myself to pocket watches and clocks. I find that I can deal with the larger components easier

Posted Images

3 hours ago, LittleWatchShop said:

Frankly, the reference device is an atomic clock.  So most, if not all, of this thread is OCD.  Interesting...fun...and useful in a vacuum, but at the end of the day, a mechanical movement pales in comparison to digital electronics. So...why bother? 

Fine, I am OK with doing your best, but polishing a turd has no real benefit.  Accept that mechanical watches are not accurate and move on.  Value the beauty of their architecture and reset them once per week.  Rent or download the movie "Castaway" and face reality!

I'm confused with this answer? although it's probably because I didn't state what I wanted clearly that's the problem it's my fault maybe I should apologize sometimes I'm stupid because I can't explain things.

Okay to be honest I want something that doesn't exist?

let's start off with what does exist and to some degree why I don't care? For frequency calibration we just need GPS we need a precision GPS so I'm snipping something out of the witschi calibration manual of their GPS receiver we need a specific kind of GPS receiver. But even in the calibration kit they don't have what I want? You would think at the bargain price of £1600 they would have what John would like to have although they don't.


8 hours ago, guidovelasquez said:

Well, you're right. We would need a very reliable standard to do our tests.
Here I have a proposal. It is not created by me. It comes with Graham Baxter's eTimer software. It is a waveform in wav format.  Is 2min and 40 seg, the sound of a 18000 BPH or 5BPS clock. Executed properly and with any of our timing machine programs, it should give the following results. -1 second of disface per day. 0.0 beat error and 262 degrees of amplitude when the lift angle is set at 52 degrees. Not much but it has helped me to learn.

what we need is a delivery method and that's the problem? We know that all the timing devices are supposed to pick up the vibration of the watch not the audio sound.

what I mean by delivery system is we need something small that would output ideally a vibration with little audio. Like I have zero idea would an earbud work? Especially if you plug the hole.

The reason why complain about delivery is for instance if I had a phone App which is quite common. Then I go to a friends house and asked him to play the wave file with his 15 inch subwoofers massively huge's speakers if we crank the volume up I'm sure the phone app would work just line outside of his house that's made out of concrete block. Pack the phone app I work fine five blocks away. But I'm looking for something really tiny that more closely simulates a watch that preferably be relatively inexpensive easy to get or basically I want the impossible.



GPS calibration of timing machine types.JPG

Link to post
Share on other sites

I don't mean to start a flame war--if I offended anyone, I am sorry.

Mechanical watches are beautiful.  They are deep in history (everyone should read The Longitude Problem) and thus connect us to our past.  They are steeped in philosophy--what is time and why does it move in only one direction?  For me, they are personal because my Dad fixed watches and sent me to college with his income.

But, they are (generally) not so accurate when compared to my phone or my computer.  I wear them...I love them...I work on them (still learning)... I want to get them close to the correct timing, but I don't sweat about the last nanosecond of error.  I have a Vibrograf--needs a new roll of paper--it helps.  It, itself, probably needs calibration.

This thread is full of fun stuff--great technology--good programming--clever electronics...all good.  I built my own piezo sensor circuit as a result of reading this thread.  I do not use the software...just look at the waveforms on my scope.  Maybe later I will invest in the software, but for now I need to concentrate on getting better at working on watches so that later this year I will feel confident to tackle my valjoux chronograph (a little scary).


  • Like 2
Link to post
Share on other sites
1 hour ago, LittleWatchShop said:

They are steeped in philosophy-

Thanks. I understand your point. I know perfectly well that no mechanical watch comes close to a medium quality quartz watch. But I agree on the importance and magic that mechanical watches have along with their philosophy and mystery of time.
I share with you an excellent book that I have read together with "the problem of length." I wish the movie was in Spanish.
Enjoy the book as many as you like.

Sundial to Atomic Clock.pdf
And a link to Witschi's clock sound generator.

Micro Signal Generator MSG Witschi.


  • Like 1
Link to post
Share on other sites
1 hour ago, guidovelasquez said:

And a link to Witschi's clock sound generator.

I would say, save your money, Guido ?.

If I read their data sheet right, you can get all that for free on your PC (signal generator software, soundcard output, piezo disc). 
What was so dearly wished here, a calibrated sound level reference, they do not offer.


  • Like 1
Link to post
Share on other sites
16 hours ago, LittleWatchShop said:

Frankly, the reference device is an atomic clock.  So most, if not all, of this thread is OCD.  Interesting...fun...and useful in a vacuum, but at the end of the day, a mechanical movement pales in comparison to digital electronics. So...why bother? 

The point would be to know if the timegrapher software, sound card, microphone, etc. work correctly or not.  Perhaps it says amplitude is 300° when in reality it is 190°?  I would not say that is merely an OCD level of difference.  It is an easy deficiency in the timegrapher algorithm when faced with a lower quality signal to make just this error.

So one does need to be trying to chase less than 0.1 s/d errors to care.  One needs a standard to be able to make any evaluation of any change in microphones, amplifiers and software algorithms.

Link to post
Share on other sites
12 hours ago, JohnR725 said:

But even in the calibration kit they don't have what I want? You would think at the bargain price of £1600 they would have what John would like to have although they don't.

It is of course possible to beat that bargain price by a significant margin, https://www.researchgate.net/publication/332543537_BUILDING_A_RASPBERRY_PI_ZERO-W_GPS_NETWORK_TIME_SERVER_FOR_UNDER_50


Link to post
Share on other sites
5 hours ago, LittleWatchShop said:

But, they are (generally) not so accurate when compared to my phone or my computer. 

when I was in Switzerland somewhere my journeys I acquired this little tiny pamphlet. it's basically a friendly book could give to the tourists talking about watches. To give you a clue of how all that is it mentions tuning fork watches. But one of the most interesting things is the very back I've scanned it in for us number 4? Then there's the other little minor problem the word up above was accurate they use the word precision and then they use accuracy and by the way precision and accuracy is not the same meaning there's a very definite difference in the meaning.

So we need to find out who wrote this book and correct them because it looks like the cell phone in the computer are more? There is a minor problem though cell phones connect wirelessly to something that has a GPS receiver built in and some were is a atomic clock. The computer is connected to a network that somewhere else is a atomic clock and everything is tied together so everything is keeping a very very precise time then there's that crappy mechanical timepiece that really isn't that crappy now is it at least for timekeeping.

then we have watches most of them are not connected wirelessly something traceable to an atomic standard. True they do make watches that do do this but mechanical watches typically do not. Mechanical watches typically a running all by themselves all day long hopefully. After all this is a watch repair discussion group we know that sometimes they don't run all day long but they're supposed to. And look what they do all day long according to silly book not bad really. They do a really beautiful job of averaging out all those imperfections and they do seem to keep time quite well.


Swiss watch interesting information.JPG

  • Like 2
Link to post
Share on other sites
4 minutes ago, guidovelasquez said:

I like Jack Forster's definition of accuracy.
Not even atomic clocks are accurate. Read this interesting article.

From Hodinkee on precision

Interesting.  Well...why must we correlate time to the earth and its goings on??  I think we offend the Martians when we do that!!!  Probably a hold host of others as well.

  • Haha 1
Link to post
Share on other sites

LOL, after my "rant" (for lack of a better term), I am monitoring the carriage clock I recovered and cleaned last weekend. Damn!  I don't think it has lost a second in the last four days!!!  Maybe I am the new reference.

Pin lever escapement...no jewels.  Oh my.

Link to post
Share on other sites
6 hours ago, LittleWatchShop said:

This is an issue that I have wondered about but was afraid to ask.

a unfortunate problem when measuring amplitude is, is it right? You can visually look at it looks pathetic and the machine says it looks wonderful unfortunately visual observation beats out the machine. If the machine is a decent our oscilloscope and it shows you where it's triggering like witschi does with some of their machines you can see what the problem might be. So I've seen it works sometimes witschi will alternate between the high and low number. Looking at the oscilloscope I can see that sometimes the triggers off the locking and sometimes it triggers off something in the middle. But at least it gives you a clue as to why it has those numbers.

It's one of the amusing problems between witschi and the Chinese machines. Usually witschi if things are too far out of alignment will give you a cryptic error message. Sometimes you can change the operating parameters of the machine like turn off amplitude and beat and only look at the rate. Sometimes are some other settings you can change because of its displaying an error message it refuses to show anything. On the other hand the Chinese machines will happily give you numbers that conceivably have no resemblance at all to what the signal  Is.

  • Like 1
Link to post
Share on other sites
On 5/4/2021 at 12:48 PM, Bonzer said:

What amp design have you used to achieve this result?

Also, is this still using the clip-on piezo mic?

I've tried to look back at this huge thread, but not really got the full information.



I use a modified VibroGraf pickup.  I could not get good results from the regular piezo.  I got ideas from several amp designs and came up with my own twist.  Basically the type of capacitors, values, and layout.  Using a breadboard was not as efficient as making a purpose built board.  I kept all the capacitors very close to the IC pins also.  I use a 9.7v battery which outputs around 10v when charged.  I went overkill on the cables which made a HUGE difference.  I did not use sockets on my board.  I kept all the traces that carry the signal separate and away from the power traces.  I will share more when I get more time.  I am building a whole new one which should be even better and more universal.   Somewhere in the past threads I think I posted the mic element I used, it is the only one I got these results from.  Unfortunately I do not remember where I got it or what the model number was, thankfully I bought several at the time.  The factory vibrograf mic was almost as good but the signal is a little weaker.   I am not an electrical engineer I am a watchmaker, so everything is trial and error for me at this point.  I am not exactly sure why it works so much better than the first one I did but it does so I am happy.  I have spent hundreds of hours messing around with this so be warned it is not for the faint at heart... LOL!  The amp circuit itself is fairly common, there are hundreds of examples throughout various datasheets and amp design literature.  The WOS amp and all others I have seen follow the same basic ideas and principals of filtering and biasing etc, just the components and values change.  Some use more opamps, some less, some have voltage regulators, some do not, some have more stages, some less, etc.  The simple transistor amps have not had good results for me, simple piezo discs also did not have good results.  I found that buying high quality components with the right values and type made a difference..  My first amp had bad STN ratio with the same basic component values but I used all cheap components from Amazon.  The new version is using all good quality components with accurate values, I tested every component before I installed them.  I found that the cheaper resistors and capacitors were not accurately rated nor were they consistent.  Also the enclosure is good quality and all aluminum.  The mic enclosure itself and the way the piezo element is attached is important.  As I said before the Vibrograf stand and the mic element I used gave the best results, I tried several.  I have other commercially made Timing machines with various mics, I tried them all.  Check out wlysenko posts, his values and design give good results.   I found that eliminating background noise and interference was most challenging.  I did not use ceramic capacitors in the signal circuit, they seem to act as mics.  Different types of capacitors react differently to temp which causes problems, and different types of capacitors have benefits/drawbacks depending on where they are used in the circuit.  Internal resistance of the components and the resistance of the traces/connections themselves have effect on the output quality. I found that a combination of different types of capacitors to achieve the desired final capacitance made a difference.  There is an article I read on the web that describes the how, where, and why of using multiple capacitors of different types and values together to arrive at the desired value.  If you want to PM me I can try to answer specific questions. 

  • Like 1
  • Thanks 1
Link to post
Share on other sites
16 hours ago, JohnR725 said:

when measuring amplitude

About amplitude and timing machine programs.
I will make an apology.
I had my first contacts with timing machines, in 1975. I was 10 years old. In my very small town, my uncle's workshop was the only one that had a timing machine. In fact I can dare to say in the country, Guatemala no more than 5 watchmakers had a machine like that at that time. My curiosity led me to approach her, although access to these devices was "restricted", and even more so to a 10-year-old boy. I was fascinated by its function and characteristic sound.
Years passed and I had access to the "owner's manual", from which I got my first knowledge of the escapement mechanism of a watch. The device could only read rate and beat error (No numerical indication) The indication of beat error was graphical, point separation or approximation. There was no way to know the amplitude, but it does not take a good watchmaker to visually determine that a watch has low or high amplitude.
Modern timing machines have the amplitude register, which I will not go into in detail to explain how they determine it and I can simply say that it is physics and mathematics.
Here is the problem. Machines can do a better job than humans. But they always need the help of humans. If the machine is misconfigured, or there is a problem in its execution protocol, we cannot blame the machine for not doing its job well.
Timing machines need a decent enough sound signal to produce correct results. The Witschi or Chinese timing machines will try to simplify things, up to a certain limit, but when there is a deficiency in some link in the chain, it collapses. It's a machine.
I want to make a statement. In order to be successful, there are no shortcuts, you need:
Reasonable knowledge of the software or timing machine. (Read the manual)
Reasonable knowledge of physics and mechanics.
Reasonable knowledge of electronics.
Reasonable knowledge of acoustics and audio.
Solid knowledge of watchmaking and watch escapement. (Only if you want to locate defects and correct them)
And Creativity. What is creativity? Imagination. The ability to see a graph or number or listen to an audio and form mental images that allow us to understand what is happening inside.
So we can ask ourselves, where do I need reinforcement?
And of course, a watch movement that is within the acceptable ranges for analysis.

Timing machines are not perfect, but in the worst case, they avoid waiting 24 hours to find out how much ahead or behind a clock, measured against a medium quality quartz watch.
I have faith in machines. I have faith in the timing machines


Image 1. Set the period for the clock, that is, the Scope Sweep Time. That should be half the BPS.


Image 2 Inspect the waveform to get an idea of the amplitude. I use headphones to listen to the sound, which also gives us an idea of the amplitude.


Image 3, configure the trigger threshold that must be low at the peak, unlocking but not affected by the "floor" of noise. This is where the importance of a good signal-to-noise ratio is. (Low noise and strong signal)

Image 4 Amplitude at full horizontal wind


Image 5 Amplitude is affected by the position of the watch. The greatest amplitude is in horizontal clock. See the separation in the waveform on the tg screen between the first sound and the last loudest.

Image 6 Side By side Watch-O-Scope tg See similar results.

Image 7. I trust machines.

  • Like 1
Link to post
Share on other sites
2 hours ago, Bonzer said:

Thanks very much for this overview

Was this the one you got good results from?

The one I used was not the green one.  I have a bunch of the green colored ones here if you want one though.  It works but is not as good, it is weak and not as sensitive.  Still better than the disc ones.

Link to post
Share on other sites
8 hours ago, CWRNH said:

Nope not particularly but that one does work, Looks like that one but has different wires, came from here in the states.

where did your's come from? I know that the green one from UK is available in the US is a place where you can order them even if they're not in stock and then they get them from UK and they ship to you. But if yours is different I'm curious about that?

Link to post
Share on other sites

Placed an order for the bimorphs, piezos and transistors in February, arrived yesterday. Now I can carry on with my own experiment. The bimorphs seem to be yellow not green.



  • Like 1
  • Thanks 1
Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Similar Content

    • By Hattori
      Ok, I am not new here, I just created a new ID since I cannot find the password for my old profile "Tourbillon" after the email address I used to create that profile was cancelled by the operator/domain owner "telia" years ago. (I tried to contact Admin about that problem, but I got no response)
    • By guidovelasquez
      Good afternoon everyone.
      I have a great project, however my skills and knowledge on the subject are limited. Let's say that I inherited from a relative of mine, my father's brother, a watchmaker, this beautiful machine from about 1974.

      Although it is very well cared for, it does not work. I have inspected and I think the problem is with a zener diode (BZY92C10). Or a pair of transistors that activate the solenoid that hits the cylinder with the helical track that prints the succession of points.

      PCB1.  The side tracks is rotate to match

      PCB 2 Is rotate o mirror to match
      The motor turns and the paper strip advances. But the characteristic sound of the impression tapping is not heard.
      In fact I was able with the help of a tip to check current of 12-24 volts, I could see that at the output to the printer / solenoid, a pulse corresponding to the beat of the clock is generated. But it seems that the current is not enough to drive the solenoid coil.
      The solenoid coil works if I activate it with a 9 volt battery. In other words, you can hear the “click”.
      If any of you wish to join me in this project, I would greatly appreciate his help.
  • Recently Browsing

    No registered users viewing this page.

  • Topics

  • Posts

    • Nucejoe has made some good points about the calendar mechanism. If the watch physically stops at least the hands stop and the secondhand keeps going that means the watch train is running it means that it's a disconnect between the gear train and our minute hand in calendar mechanism. The Canon opinion which I'm going but I guess is the type that snaps into a drive wheel. The drive wheel is driven by the gear train the Canon pinion goes on the post there is no friction there. If the friction between the drive wheel in the Canon pinion disintegrates which he can then with the least amount of friction like a calendar mechanism it just quit striving. The hands come to a stop the watch keeps running the secondhand keeps moving because it's independence of all of this. It should have been obvious when the watch was going to gather and you check the setting mechanism before you put the calendar on you would have noticed zero friction as a guess. Even now when you go into hands setting you'll feel like there's no friction at all. Then the reason why the calendar mechanism works when you manually rotate the hands is because the setting wheel is driving the Canon pinion directly which is driving the calendar mechanism and that our wheel and all of that so all of that will run from that we just will not run from the gear train running the drive wheel that's connected to the Canon pinion. But that's just my wild guess and then we throw in Nucejoe's possible calendar mechanism increased friction than we need to really isolate all of this or we can continue to guess.  
    • I am not sure if I correctly understand you here. Only the seconds hand keeps running or minute and hour hands do move too, in case minute and hour hands move and show time right, then the fault is in date change train including date jumper mech, but if it doesn't show time correctly ( appear to loose time) then its loose canon pinion. You can tighten a loose canon pinion, use grease to lube it, not oil.
    • Thank you for your introduction and welcome to this friendly forum.
    • always confusing when it's two separate watches but I'm going to assume they're basically identical. What was their condition before you service them in other words did have a problem before and the problem came after you've serviced or was the problems there before? Then timing machine results you do have a timing machine don't you? In other words what's the running condition of the watch like of the watches barely running that would be an issue for a calendar change   then you fix the problem? so I'm guessing we only have to worry about the 2778 then? Yes this is what happens when you have multiple watches with too much going on it becomes confusing. then it would be helpful to have a picture of the dial side components because we didn't memorize every single calendar change mechanism. one of the places to look is the Canon pinion assembly in other words to visit have enough force to drive the calendar mechanism? I'm having to guess because I'm not finding a good tech sheet that shows the parts. If it's a kinda Canon pinion I think which is the Canon pinion that slips into her presses into with friction with the wheel if it no longer has and the friction then a cannot drive the calendar but if you manually set the watch that drives the Canon pinion directly and everything should function. then if you two of the parts list it would be listed as canon pinion with drive wheel. .But it would be helpful in the picture just to make sure http://www.ranfft.de/cgi-bin/bidfun-db.cgi?10&ranfft&0&2uswk&ETA_2778  
    • I think what the problem here is you're asking the wrong question. The question is what part number do I need so that I can end up with the correct balance? Or the correct balance staff for that matter? If you look at the second link it lists the parts for your watch. This is always where things get interesting? you'll notice for balance staff they list seven different ones. For balance complete they do show three but if you look at the part numbers there's only two. then I attached some images from bestfit online and of course additional problems perhaps. Notice you have lots of choices and probably only one that's right. then the physical bestfit book becomes interesting because is an indication that there are variations which is of course why there so many listed in the images. but fortunately it looks like you want the basic simple one which should be 100/66.  unfortunately for the third link it's out of stock. then providing this is the right staff number I snipped out a couple of more images of the bestfit book which is what the dimensions mean and the dimensions. One of the irritating things of the physical book is on the dimension chart if you're trying to find a particular staff you just have to be lucky to go through the list and find it it be really nice if they list of the sizes by the numeric staff also. http://www.ranfft.de/cgi-bin/bidfun-db.cgi?10&ranfft&0&2uswk&AS_1287 http://cgi.julesborel.com/cgi-bin/matcgi2?ref=AS_1287 http://cgi.julesborel.com/cgi-bin/matcgi2?ref=U\ZD]J http://www.julesborel.com/s.nl/it.A/id.151223/.f  
  • Create New...