Jump to content

WOS mic & watch holder.


p2n

Recommended Posts

To save filling up the WOS thread, I though I'd post my design of the watch/mic holder.

This is the mark 2 design. There is still plenty to improve, but it works well.

The parts:

  1. 12mm MDF board
  2. Sticky back foam http://amzn.eu/2OJAKGS
  3. Lapel Mic http://amzn.eu/5vx2I9Y
  4. Camera ball mount http://amzn.eu/cXi9TWv  (Borrowed from my camera)
  5. Small spring http://amzn.eu/fmMyvMi
  6. Screws & Glue etc

All that's missing is a mechanism to hold the watch in place. At the moment I'm using an elastic band.

Parts.thumb.jpg.68ba2ecf307fcbc2b0ee4177890f5503.jpg

The Mic: all I did was remove the cap so that the watch back contacts the transducer instead of the case. The hole drilled in the MDF is big enough to allow the mic to move easily. There is a small spring threaded on the cable.

P4080009.thumb.JPG.b29c6e19fdb8d4ac2fc23632324403c1.JPG

On the back there is a small block with a slot and small hole to stop the spring going right through.

P4080011.thumb.JPG.d385ee43899046983872f5de9e4e7087.JPG

 

The whole lot is fixed to a simple stand using the camera ball joint. This particular one allows a full range of movement so the watch can be held in any position.

P4080010.thumb.JPG.ebc269fc3284884f8258b664906a90a3.JPG

 

As mentioned before there is no pre-amp but the sound and volume seem reasonable. Mind you this is a pocket watch so it may not work as well with a wrist watch.

Here is a recording of my Ingersol-trenton.wav

 

  • Like 1
Link to comment
Share on other sites

4 hours ago, p2n said:

To save filling up the WOS thread, I though I'd post my design of the watch/mic holder.

This is the mark 2 design. There is still plenty to improve, but it works well.

The parts:

  1. 12mm MDF board
  2. Sticky back foam http://amzn.eu/2OJAKGS
  3. Lapel Mic http://amzn.eu/5vx2I9Y
  4. Camera ball mount http://amzn.eu/cXi9TWv  (Borrowed from my camera)
  5. Small spring http://amzn.eu/fmMyvMi
  6. Screws & Glue etc

All that's missing is a mechanism to hold the watch in place. At the moment I'm using an elastic band.

Parts.thumb.jpg.68ba2ecf307fcbc2b0ee4177890f5503.jpg

The Mic: all I did was remove the cap so that the watch back contacts the transducer instead of the case. The hole drilled in the MDF is big enough to allow the mic to move easily. There is a small spring threaded on the cable.

P4080009.thumb.JPG.b29c6e19fdb8d4ac2fc23632324403c1.JPG

On the back there is a small block with a slot and small hole to stop the spring going right through.

P4080011.thumb.JPG.d385ee43899046983872f5de9e4e7087.JPG

 

The whole lot is fixed to a simple stand using the camera ball joint. This particular one allows a full range of movement so the watch can be held in any position.

P4080010.thumb.JPG.ebc269fc3284884f8258b664906a90a3.JPG

 

As mentioned before there is no pre-amp but the sound and volume seem reasonable. Mind you this is a pocket watch so it may not work as well with a wrist watch.

Here is a recording of my Ingersol-trenton.wav

 

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.

Link to comment
Share on other sites

The website for the microphone a really is not adequate it all. But it does give a clue of noise canceling and it has to plug into the computer input it can't work on a normal microphone input. That tells us it is a electret type of microphone which already has built-in circuitry usually a special FET.

 

https://www.speedlink.com/en/oxid/PC-Accessories/Microphones/SPES-CLIP-ON-MICROPHONE-BLACK.html#HIGHLIGHTS

Link to comment
Share on other sites

We have an interesting challenge listening to the watch? This microphone is going to try to pick up the sound versus the traditional trying to get a vibration. What would really be nice to have is what's found at the link below. This way we can verify the microphone is really picking up what it's supposed to and correctly displaying the results.

https://www.witschi.com/en/group-of-devices/measurement-of-mechanical-watches/microsignalgenerator/microsignalgenerator.html

 

Link to comment
Share on other sites

15 minutes ago, JohnR725 said:

We have an interesting challenge listening to the watch? This microphone is going to try to pick up the sound versus the traditional trying to get a vibration. What would really be nice to have is what's found at the link below. This way we can verify the microphone is really picking up what it's supposed to and correctly displaying the results.

https://www.witschi.com/en/group-of-devices/measurement-of-mechanical-watches/microsignalgenerator/microsignalgenerator.html

 

Here is another document that can help us understand what we are trying to register.

http://www.iaeng.org/publication/WCE2007/WCE2007_pp624-629.pdf

  • Like 1
Link to comment
Share on other sites

Think about the physics for a second - sound is vibration, so what we're really discussing is the method of transmission from the watch to the transducer.

All mics have a diaphram of some sort so by puttng that diaphram in contact with he watch, we are holding it still so that airborne vibrations would have little effect, but vibration transmitted mechanically would be picked up.

In my setup, the mic insert is in contact with the watch and is also enclosed by the MDF holder. That means I'm picking up the watch vibration and very little background noise. I did made a recording whilst chatting in the room and you can hear muffled voices in the background - but at a level that you expect if you were listening to someone in an ajoining room and not enough to effect the WOS. If the mic is an electret type with a built in pre-amp then thats good - saves building one!

Electrical noise is also not a problem in a normal room. However, I tried a reading in the evening and found the signal swamped by noise from the thyristor phase control based dimmer switch next to me!

You can split hairs on the type of mic etc but the main points are:

  1. It works and work well - perfect for a beginner or hobbyest
  2. These are all off-the-shelf items
  3. No electronic construction needed
  4. The whole lot can be built for under £50 (less if you skip the ball mount or use a cheaper one)

I used to work in the electronics induictry and make PCB's but that was 20 years ago and don't have kit now. I can't be the only one who doesn't want to mess about with ferric chloride or bread board!

 

3 hours ago, guidovelasquez said:

Here is another document that can help us understand what we are trying to register.

http://www.iaeng.org/publication/WCE2007/WCE2007_pp624-629.pdf

Interesting doc - a little advanced for me at the moment! I'm lucky I still have my oscilloscope in the attic, so if I need to see the waveform with that clarity I'll dig it out

 

3 hours ago, JohnR725 said:

What would really be nice to have is what's found at the link below.

Now that woudl be a real nice thing to own :) but if you had one of those you'd probably own one of these as well

https://www.witschi.com/en/group-of-devices/measurement-of-mechanical-watches/watch-expert/watch-expert.html

Edited by p2n
bloody typo
Link to comment
Share on other sites

What witschi appears to be doing is moving away from dedicated hardware with embedded processors to basically PCs running software. Then an example of this is microphone with software you supply the computer. Unfortunately I suspect it's probably still expensive but should be cheaper than a full machine.

https://www.witschi.com/en/group-of-devices/measurement-of-mechanical-watches/chronomaster/chronomaster.html

Link to comment
Share on other sites

Thanks for the contributions and the links. It is evident that there are solutions out of reach of our pocket. But the important thing is that they give us ideas. Thanks for the explanation of your p2n microphone. Now I dare to give an opinion with my limited knowledge. I think we should analyze the professional microphones of these time machines all microphophones or most of them that I have been able to see, are based on piezo electric microphones or called crystal microphones. (as the turntable pickup). The microphones for time machines are expensive. But it should be cheap because they do not have too many complications. A rotating arm an electric piezo sensor suitable shield and shielded cable.

See this microphone

LepsiWatch-5.jpg.571b3034ad2e1e2044cccc308e5adb60.jpg

Link to comment
Share on other sites

47 minutes ago, guidovelasquez said:

turntable pickup

Now that's an idea! You could pick up a cheap cartridge and modify the stylus quite simply and cheaply (under £3 fleabay). A pre-amp is approx £15. You'd set it up so that the needle just contacts the watch stem/back.

Link to comment
Share on other sites

One of the important things mentioned in this particular discussion is the only way to make use of the software is we have to have a decent pick up that easy to make, find or basically possess. I had shown the software to a group of watchmakers hobbyists and professionals who like the look but not having a pickup or having to make something the whole thing crashed and burned right there.

So the ideas presented in this discussion are interesting and the only way to see if It works or not is to make it. But transmission of the ticking through air I think is problematic. The example here works the watches in the case but I've seen cases that are heavily textured is not going get it could clean contact. Then what about the movements without the case? Universally anything resembling basically anything serious is some form of a vibration pickup. Then yes I recognize sound is a vibration but traditionally mechanical vibration not audio vibration through air.

Then one of the problems with the discussion of software is the number of pages and that we've discussed ideas in the past but they're buried way too far in the discussion. So for instance the link below and yes you get it in the UK I'm only giving the US source mainly So you can see the PDF data sheet. Notice the reference that it can be used as a photograph pickup. Plus we know it's already being used as a pickup for watches.

Then today for making the things like the physical mounting. Even if everyone doesn't possess a 3-D printer on the kitchen table there are companies out there professionally print stuff allowing us to come up with some very complex shapes and designs that would make construction of a microphone a lot easier. That is once we come up with a design for something

https://www.alliedelec.com/rs-pro-285784/70637796/

 

 

Link to comment
Share on other sites

34 minutes ago, JohnR725 said:

Then what about the movements without the case?

Been thinking about that. I'm going to try with a pocket watch movement and report back. If I get a wrist watch movement that works I'll try that also.

I've got a few other ideas for the pickup I'll try and also post if anything works.

My objective is to find something the average person could build on the kitchen table with nothing more than common tools and bits from amazon/ebay.

This is only the Mk2....

Link to comment
Share on other sites

Righto - had a bit more of a play and noticed:

Repeatability issues: Rate ok, Beat ok, Amplitude - either it's my watch, WOS or the setup, but it's all over the place

Out of case - not working at all.

Levels - a lot of variation. I was joking about the pizza base before, but it turns out it does actually give the best signal. :$

Stay tuned for V3

Link to comment
Share on other sites

You should be posting the scope diagram that will give us a clue.

Problem is we looking at a complex waveforms it's made up of five separate sounds which typically get idealized into three sounds that look really nice pure and clean. For instance the PDF below and the image I'm posting below. Notice how nice and clean it looks your scope will not look like this.

So the absolute most important sound is the roller jewel hitting the fork everything uses this. This requires very quiet background and a really nice initial sound just like it shows a diagram down below. So rate, beat an amplitude all rely on this sound. Amplitude relies on by the diagram below the third sound which is locking. It's really easy to trigger off  because it's a really loud sound. The basically you're going to trigger off the most quiet sound and the most loud sound. Which the diagram below is super easy to do.

But the diagram below is not the way they look. I'll take a flash drive to work today and see if I can capture some witschi oscilloscope images so I can show you that even the witschi has issues with where it picks up things.

So reality versus the fantasy below. What if we don't trigger off the roller jewel what if we have a non-technical term a mushy signal. So if we can't get a clean initial triggering the graphical display may show some way Variations will still get a rate will still get a beat and If you get the amplitude because you moved to triggering part farther in the amplitude numerically will be much higher. IIf you don't pick the initial up at all and you move much farther in like all the way to locking you'll get a graphical display with rate and beat zero amplitude.

So let's see what the scope looks like we'll get an idea of what the software is trying to interpret.

https://www.witschi.com/assets/files/sheets/Test and measuring technology mechanical watches.pdf

ideal waveform amplitude.JPG

  • Like 1
Link to comment
Share on other sites

That's why it's important to get a clean signal. I used the sound of the p2n clock. I applied a high pass filter to "clean" the signal. and then we have the three important sounds in the analysis. In the second image the first arrow from the left is the unlock, clear or release. The second indicates the impulse. and the third that indicates the highest peak, the blockage, the most intense sound, of the escape wheel caught by the ruby of the pallets of the anchor.

waveform.JPG.146c631fc326f65bf4ad04518a03276e.JPG      5acd6aa5e5257_waveformarrows.jpg.06f623d5e8836ed4f007d03d5bfd139b.jpg

I like the statement that says: "Knowledge is power". For that we share information.
 

 

 

Link to comment
Share on other sites

Been poking about on the interwebs and found two other software watch timing programs. Here is the output from the same watch in the same position with the same mic on the same laptop and (almost) the same time of day.

Interesting variation ...

The site favourite: Watch-O-Scope:

wos.PNG.2dc158ee2f4f82ebb2376d5581e02471.PNG

 

PC Timing Machine - written by a site member:

PCTM.thumb.JPG.f375b42c4751d342eccf0559c0b77a97.JPG

 

Open source TG

TG.thumb.JPG.ccf2f97ee0aa4797e440b35b518fb79d.JPG

 

It seems the WOS has signal issues - it doesn't cope as well with a poor signal and high SN ratio as the other two. I also noticed that the other two are more immune to other noises - if I move the mic/holder/watch etc they don't seem to be as affected as WOS.

 

 

On another point - I have arriving soon a piezo vibration sensor + signal pre-amp board. It had analogue (and digital) output so will see if I can use that to get a better signal-to-noise ratio that the lapel mic.

 

Link to comment
Share on other sites

The problem with all of these is we really should find in our oscilloscope program for the sound input. Or one of the normal sound input programs to see if we can see with the waveform actually looks like.

Notice the amplitude 300° is wrong one of the programs doesn't know what the amplitude is at all that's probably right and the other one says it's 200 and something this is what happens if you can't pick up the signal clearly.

Then if we weren't testing software/Hardware your first display indicates a watch having a problem. As soon as you get problem displays the graphical display is not smooth then the numbers start to just fly out the window. Oh something else to try with his software put it in the raw display mode and let's see what that looks like. Ill give us an idea some of the noise in the signals strength and stuff.

So we still get stuck with a problem basically where do we stand? Were not sure about the software's ability to process the signals except we know that the software all works. At least lots of people indicated it's worked in I've seen the software working. But if we can't get a good clean signal in then all the software becomes worthless. If the watch were trying to use as a test watch is having issues then evaluating the software flies out the window. It be really nice if we could come up with a universal test watch or maybe our own synthetic test device providing we could make that consistence.

 

Link to comment
Share on other sites

12 minutes ago, JohnR725 said:

we really should find in our oscilloscope program for the sound input. Or one of the normal sound input programs to see if we can see with the waveform actually looks like.

Audacity:

scoped.thumb.JPG.015dbf2fd6a799b13ca17d3dfc51374f.JPG

Note the mains hum - WOS and the other use DSP to remove this.

13 minutes ago, JohnR725 said:

your first display indicates a watch having a problem.

It shouldn't matter - the software should be able to cope with all, after all that's what it's for. The problem in the first image with WOS is that the software has poor signal filtering and isn't triggering correctly - as a result the data is corrupted.

WOS scope mode and raw plot:

WOS-scope.thumb.JPG.49c018bf7a6c0629b396132629e7e56a.JPGwos-raw.PNG.47e55f8b4351b20cb22501d94326828e.PNG

 

My summary would be that WOS needs a strong signal - it's has the weakest signal processing. The other two worked fine with a low signal and high SN ratio. WOS GUI seems a little dates as does PCTM - neither scale. TG has a nice UI but limited more limited options.

 

Link to comment
Share on other sites

I think it's an unfair prosecution of WOS. It's a bad microphone. A poorly preamplified signal, The clock is far outside the Rate. I have my opinion about WOS. In the programs that I know and that are missing 2 in the list that the partner placed, Biburo and eTimer, and the others, I place WOS in second place. I think that to make an appropriate qualification to the program it is necessary to have a pre-amplifier and a good microphone. (reasonably.)
In the image below I also think that you can choose the value that I enclose in a yellow circle and the others I indicate in red. Regulate the clock so that it has an offset of - + 20 seconds, or less. With all the respect that friends deserve.

Configu.thumb.JPG.7ffa29feaf009cce9e083384c3b49cc0.JPG

Link to comment
Share on other sites

Here is an example of what WOS accomplishes. It's the audio track that you hung on your Ingersol. What other programs do is that they apply filters or algorithms that choose the best of the waveform. and they present it to you. It's what I did here. To correct the rate I modified the WOS calibration factor to a value of 242 seconds +. That explains why the clock is now marching near the right, which is false of course.

Ingersol.PNG.9b2397bcf765af42ce2f70cde873c422.PNG

Do not forget the filtering factors that you can apply in WOS

Edited by guidovelasquez
Language error
Link to comment
Share on other sites

So if I go back to look at everything the scope mode and the raw signal just looks not quite right. If you don't get a good clean signal you're not good get good results. I think about every timing machine I've ever seen except the apps for phones and tablets where they have no other microphone everything else is picking up the vibration. The idea of a cheap microphone is a great idea but it just isn't getting a nice clean signal. Then even the other software examples PC Timing Machine The graphical display looks rough. Which could be the way it's supposed to look or it could be an interpretation of the waveform is not correct. Then for Open source TG There seems to be zero graphical display? Audacity The signal looks soft which may not be a technical term I'm not seeing a clear signal.

It's where we were testing something like this we need a standard watch that we know what it's doing or some other timing machine to compare results to. Otherwise when testing with a watch that may or may not have issues that could be entering into the picture

So how should an oscilloscope look like? I'm attaching some diagrams this watch has seen better days. It's a small seven jewel Elgin that has issues. So we'll see from the oscilloscope nice clean signal. Then the graphical display yes it's all over the place it gets worse when it goes into a pendant position. What I find interesting with the oscilloscope mode is I've never seen it look like the graphical representation of what it's supposed to look like. Then I zoomed in the one section you'll notice a red dashed line the machine is telling us where it thinks locking is occurring. So on the top it's where it should be sort of notice how it's not parallel to the signal it's in the signal a little bit?  Then the lower one notice it's not where it's supposed to be. Basically it has a window where it's looking for a lock to occur if it finds a signal it assumes it's the correct signal. It's not apparently looking for the loudest Signal. Then yes the lift angle for this particular watch is 62°.

 

 

s-3.JPG

s-4.JPG

s-2.JPG

s-1.JPG

  • Like 1
Link to comment
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.

Guest
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.



×
×
  • Create New...