Jump to content

Recommended Posts

4 hours ago, p2n said:

I put more info on the stand and holder on this thread:

Excellent work. I just have one question. The microphone is electret or other type. The signal level noise is very good. I can not hear the net humm.

 

Share this post


Link to post
Share on other sites

The movement of the clocks generates three readings that help the watchmakers to determine the condition or state of a watch. These are the Readings: Rate, Amplitude, and Beat Error. Here we see the graph produced the "Timing Machime", We will focus on the amplitude. The amplitude refers to the rotation made by the balance wheel from its resting point (theoretically stopped and without wind) to any of its extremes. Be levógiro or dextrógiro. This must be between 300 degrees. The position of the watch, the cleaning, the quality of its lubrication and condition of the machine, will affect this result. In the case of the graph here, made with Watch-O-Scope, it indicates that they are 202 degrees. To obtain this result it is necessary to enter a value to the time machine. It is called elevation angle (lift Angle) value that is found in tables prepared by manufacturers or made by people from different documents. There you can see the Lift Angle value in the graph 41 deg.  After watching the video in normal playback is difficult to see how accurately rotates the balance wheel. By the way, I made a point of reference aligned with the ruby ellipse or impulse pin. And in the reproduction in slow motion, then yes you can see that the balance wheel turns a little more than half a turn that would be 180 degrees. And it is what the time machine tells us. 202 degrees

 

 

Share this post


Link to post
Share on other sites

I've already posted this on another popular watch forum, but I think it better fits here. I'm sorry if this post get too lengthy due to all the images.

Hey everyone! I would like to share a project that I've been working on for a while. After a few different variations, I think that this is my final design for the time being.

Project background: For many in this hobby, applications like Toolwatch are great because they allow you to determine daily rate, but it doesn't provide anything beyond that. Those of us who enjoy tinkering with watches should have a way to also check the amplitude and beat error. Sure, I could have purchased a timegrapher from China but I thought it would be a fun project to attempt constructing one myself (with the help of several resources).

I used these microphone stand plans for reference and made some changes.
There are also plans for a DIY pre-amp on the Watch-O-Scope website, but I followed the instructions from my good friend Guido (also posting on this thread) and modified a PYLE pre-amp by removing a capacitor from one channel and replacing the electrolytic capacitor with a ceramic capacitor on the same channel.

The first iteration gave promising results but I wasn't entirely pleased with it. It seemed like the audio levels were somewhat inconsistent and I would get a ringing or echoing noise depending on the amount of pressure against the piezo disk (the contact microphone element).

1.jpg.19387b2d315de58ea661d91711fb349c.jpg

After some discussion with a friend, we came to the conclusion that many professional timing stands never have direct pressure against the piezo element - instead, the metal clamp that holds the watch is coupled with the piezo using a different piece of metal that transfers the vibrations. We came up with these sketches for a better design. 

2.jpg.b5bd9ed286fca5057516dc78c306661b.jpg 3.jpg.4c78dc0f5527b12dc8f5f698fce3a06c.jpg

Completed and polished pin before attaching it to the piezo.

4.jpg.0b2e1f3995a049bb7544256aebede8cf.jpg

Pin installed. Soldered some thin cables from broken earbuds to a 3.5mm connector mounted in the stand.

 5.jpg.7104b3e8cf20b28d2aea581d587554b5.jpg 6.jpg.bbecffc8f8f2b43fb6291d740d29b2c2.jpg 7.jpg.bcc5578962e9d357977d9a24eb379f81.jpg

The results are good. Now it's time to add a copper plate to the back to reduce electromagnetic interference.
Another good idea is to add some felt so I don't scratch any crystals when testing the dial down position.

8.jpg.6dc3bf363a8349fd1c37f7c5ba5ed298.jpg 9.jpg.67e3767cb08b3aaaafd5dac6d3dabd4e.jpg 10.jpg.19dfc2a39f0218fa7c55ce782e83604f.jpg

Finally, the build is complete! I'll admit that this might not be the most beautiful creation, but it's definitely functional and it does exactly what I need it to do.
I have ideas for further improvements, but they are not necessary and are really just a challenge to get the cleanest audio signal (there's still quite a bit of static).

Waveform from Watch-O-Scope (highly recommended software, made by a fellow WUS user).
This is amazing software that I pair this microphone stand with. An alternative is TG Timer, but Watch-O-Scope definitely is the better option.

12.png.606151c1bc3fd8da445e0a419231df9b.png

Quick audio recording I made: Test Recording #1 - Mega.nz

Share this post


Link to post
Share on other sites
49 minutes ago, guidovelasquez said:

Excellent document. Although you comment that it does not seem to be aesthetically good, for me it is. It's the best I've seen.

Thanks :)

Actully it's OK, I just think the back looks weird with three screws AND three rubber feet. I should have used some of those rubber feet with integrated screws.

Share this post


Link to post
Share on other sites

Dear all,

this thread helped me a lot. Plenty of people shared their experiences. Let me say: "Thank you all, folks!":Bravo:

So is time to slightly return what I got and share my fresh experiences.

I am newbie in watch repair but not in electronics. I can read plenty of stories here about sensors and amplifiers.
I can read "simple" or "easy" or "cheap" or "work for me" solutions. No one worked for me as I expected :phew:.

1. sensor - piezo :unsure:. All piezos have very low sensitivity. Does not matter of the size - (resonant frequency depends on size).
Most of available circuits are based on a "knock" detector. For watch sensing have very low sensitivity. Also sensor frame construction takes account a lot - because of energy absorption. 
Holding piezo in fingers and trying to detect a sound - no way. So finally I decided to use a microphone.

2. sensor - electret microphone (common type):rolleyes:. Described problem with a  sensor frame construction is much lower and sensitivity is much higher. 
What is bad the mic can hear any other noise not just watches. So pay attention to any fans, wind, building works or a colleague typing very hard next you.

3. amplifiers :cigar: - the sensor supports some signal. Input value extremely depends on where sensor touch your watches. Still is important signal-noise ratio (high sensor signal in very noisy background can be unusable!) not just absolute value of the signal.
Just FYI the ratio between mic sensor attached to closed watches and opened (touched to a watch frame) should be 100times!!!

Signal coming out from the sensor is very low and must pre-amplified about 90dB+ (partially in you external pre-amplifier and partially in you PC or phone).

What is bad and not very suitable for the electronic enthusiasts - high gain amplifier can get easy oscillation. Oscillating amplifier is definitely not the amplifier. Without deep experiences what going on with your amplifier and an oscilloscope tool your are not able solve that problem.  :thumbsd:

Another problem is any bread board solution hardly can be copied with success (still we are talking about high gain amplifier).

:hot:

Uffff, too many bad information. Yes I know, sorry. Just sharing mine.

Is there any repeatable solution today?

Yes, exists. I know it from today because finally this solution works for me and is usable without any special tools.:thumbsu:

You need:

1. microphone pre-amplifier from ebay based on max9814.
https://www.ebay.com/itm/MAX9814-Electret-Microphone-Amplifier-Board-Module-AGC-Auto-Gain-ASS/263928751428?hash=item3d73612544:g:~BoAAOSwD5ZZxMVm

2. quitar sensor
https://www.ebay.com/itm/Guitar-Acoustic-Clip-line-Pick-up-Pickup-Built-in-Vibration-Sensor-With-1m-Cable/282574713836?hash=item41cac3f3ec:g:EKIAAOSwf~ZZbI6k

3. some shielded cable, 3.5mm connector to your PC or mobile (I used PC because my Xiaomi phone heavy loaded pre-amplifier output. Actually don't know why), 5V power supply - prefer linear and external not USB or 4.5V battery, hot glue, solder station and basic electronic experiences or ask friend.

Job description
1. carefully desolder the microphone from pre-amplifier
2. open sensor clip and remove piezo. Drill 9.6mm hole inside (hole depends on removed microphone).
3. cut off the connector from piezo sensor cable and solder on microphone (keep an eye of shielding) and other side connect to pins on pre-amplifier board
4. put microphone into the hole and fix position and cable with hot glue
5. solder on cable connector for your PC or mobile (mobile has 4pin and different connecting than PC - use google.
6. Connect 5V power supply and try out to setup your PC mic input as was previously described in this thread.

How to test entire solution:
Clip on microphone sensor to your opened watches (getting the strongest signal available). Eanable "listening" option in microphone tab. From your speakers you have to hear sound of your watch beating. Preset gains to level below you hear oscillating.

If everything works fine just turn on a timing SW.

Good luck and apologize my bad English

:angel:

PS: I was trying out to check accuracy colleague of mine OMEGA. Sure unopened. And I was not able to find  proper gain level before oscillating. FYI.

PS: Why my phone heavy load pre-amplifier output ... maybe next time. 

PS: You can see resistor 392 (3K9) in place of original (2K2). I lost 2k2 during testing.  

IMG_20180912_143134a.thumb.jpg.5977052b26f46698f05c32422d96eb58.jpg

IMG_20180912_143115a.jpg

IMG_20180912_143151a.jpg

IMG_20180912_143305a.jpg

IMG_20180912_143334a.jpg

IMG_20180912_150425a.jpg

Share this post


Link to post
Share on other sites
2 hours ago, B52 said:

Dear all,

this thread helped me a lot. Plenty of people shared their experiences. Let me say: "Thank you all, folks!":Bravo:

So is time to slightly return what I got and share my fresh experiences.

I am newbie in watch repair but not in electronics. I can read plenty of stories here about sensors and amplifiers.
I can read "simple" or "easy" or "cheap" or "work for me" solutions. No one worked for me as I expected :phew:.

How to test entire solution:
Clip on microphone sensor to your opened watches (getting the strongest signal available). Eanable "listening" option in microphone tab. From your speakers you have to hear sound of your watch beating. Preset gains to level below you hear oscillating.

If everything works fine just turn on a timing SW.

Good luck and apologize my bad English

:angel:

Good work! I'm curious how it sounds; do you have a recording in WAV or MP3 format that you can upload? :)

Share this post


Link to post
Share on other sites

Hi B52,  Welcome to the forum.  Your solution is very interesting, from the perspective of using an electret element as the sensor and the repeatability of the design using a well documented amplifier.  The technical details for the amp are on the Adafruit site:

https://learn.adafruit.com/adafruit-agc-electret-microphone-amplifier-max9814/downloads

Edited by robmack

Share this post


Link to post
Share on other sites

Hello B52. Thank you very much for sharing your experiences. We will have to test your project. However, in our limited knowledge and experience we have seen that what has worked best is the piezo electric disk. We know that it has its weaknesses, but even so it can be efficient. One of the reasons why we see that it is so, is that most time machines, use the principle of the vibration sensor instead of the microphone. For the same reasons that you have exposed: The background noise. We would like you to put us a piece of recording in wav or mp3 to check your results. Congratulations. And thanks for getting in touch with other elements that could be used. I would like to know if there is a pre-amplifier with a similar presentation to which you show us.

Share this post


Link to post
Share on other sites

Dear guidovelasquez,

yes you are right. The piezo is the most common solution. The mic solution is more suitable for who is not very familiar with electronics. You always have to thing that making properly working pair (sensor + amplifier) covers two good jobs - mechanic and electronic. Booth are important. Bad mechanic job you can ruin your perfect amplifier and vice versa.

Actually I have been working how to connect this sensor with  SW. My very fresh experiences till now is not optimistic again. Very first sharing experiences with tg-timer SW:

1. Miyoto some movement - working well, detection ok. Tg SW works as I expect.:thumbsu:

2. Junghans Kal75 - working movement but not properly - SW says no detection. My oscilloscope displays me quite usable signal but SW is not able to recognize that. User cannot recognise why. If is there a problem with sensor or somewhere else.:thumbsd:

3. 90 yrs old movement - also proper signal but different shape. The same situation as previously - no detection

https://www.ebay.com/itm/40pc-Vintage-Gold-Rectangular-Oval-Lady-MOVEMENT-Parts-Watch-AS-IS/132771982027?hash=item1ee9d3aacb:g:te8AAOSwCCVZtzA-

My movement is below 912 number:thumbsd:tg-timer Sw is nice piece of work and but actually have some limitation. And I would like to share my experiences. So my target is to explain usable solution (sensor + PC +SW) for most people. Is never-ending truth if you are beginner you should not be able to recognize where is the problem. 

Improving previously described solution is always welcome! :Bravo:

Anyway, feel free to use MAX4914 pre-amp with piezo. You have to add one stage - impedance converter and remove bias from MAX9814. Something like this:

http://sound.whsites.net/articles/high-z.html

I was not successful and patient enough this way .:angel:

cheers

Share this post


Link to post
Share on other sites

Thank you very much for your answer B52. And for the details that enrich our trials. I agree with you, it is necessary that you possess skills in the field of electronics and mechanics in order to obtain acceptable results. I have a question about the Junghas Cal 75 movement. Is it 18,000 BPH? Does it run within the acceptable ranges, that is to say, its error to march does not exceed more than 60 seconds? About the 90-year vintage movement, is it a bph that can recognize TG?

regards

Guido.

Share this post


Link to post
Share on other sites

Thank you very much for your answer B52. And for the details that enrich our trials. I agree with you, it is necessary that you possess skills in the field of electronics and mechanics in order to obtain acceptable results. I have a question about the Junghas Cal 75 movement. Is it 18,000 BPH? Does it run within the acceptable ranges, that is to say, its error to march does not exceed more than 60 seconds? About the 90-year vintage movement, is it a bph that can recognize TG?

regards

Guido.

Here example sund. Omega Seamaster De Ville  +18seg aprox.

https://mega.nz/#!uCIGzSTZ!OosG2VcTsj7uA8JDCczusvRjFTxIzMk5IVHAzD3UdCk

Share this post


Link to post
Share on other sites

hello guido,

well KAL75 seems to beat at 21600. But not sure actually. Beating is not stable due something and TG cannot recognize the sound wave. So firstly I have to repair that movement. It will take some time because I have to finish several already promised projects and find spares.

:cool:

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×