Jump to content

My homebrew Timegrapher ;)


rafau

Recommended Posts

This is my attempt of making a Timegrapher - like tool. I converted the mono cable from an old Vibrograf microphone stand and connected it to the laptop. I use the Watch-o-scope software and find the signal a bit too weak. It does work but I'm afraid it might be not accurate enough.

Does anyone know if vintage vibrografs had a sort of preamp built in the main machine? There's nothing apart from the microphone in my toy and I believe it used to be connected straight to the vibrograph.

 

90c37fdb05a5748ddd7ae76b7e4409ac.jpg

 

5bdb18de48da75731ac69202461f12cf.jpg

 

6e104a7343586b280a435b864979f3df.jpg

 

9524729f4fcfbe900708332029814da3.jpg

Link to comment
Share on other sites

Hi,

these microphones don't contain a preamp, however their piezo element is more sensitive than a common piezo disc. Depending on your soundcard (additonal amplification activated?) you may need one or not.

Your picture shows that one of both tic noises is too weak, and its 1st pulse was not detected (scattered line).

In the Info-pdf of my PCTM (Link) you can find the schematic for a simple DIY preamp. You can place it inside your mic body or even into the plug.

Frank

 

  • Like 1
Link to comment
Share on other sites

14 minutes ago, praezis said:

Your picture shows that one of both tic noises is too weak, and its 1st pulse was not detected (scattered line).

I have used Weishi timegraphers on hundreds of watches and occasionally scattered dots are shown. They indicate extra noise in the mov.t,, not a problem with the instrument. Sometime, but not always, a better reading can be obtained adjusting the gain on the machine. As mentioned above already, this app/mic results must be compared with a watch which is know to display real good - flat, horizontal line without dots out of place on a reference machine.

Link to comment
Share on other sites

33 minutes ago, jdm said:

I have used Weishi timegraphers on hundreds of watches and occasionally scattered dots are shown. They indicate extra noise in the mov.t,, not a problem with the instrument. Sometime, but not always, a better reading can be obtained adjusting the gain on the machine. As mentioned above already, this app/mic results must be compared with a watch which is know to display real good - flat, horizontal line without dots out of place on a reference machine.

Hi jdm,

we discussed this topic just recently.

Problem with Weishi, even with Witschi Watch Expert, is: you cannot say if the scattered line comes from extra noise or from neglecting the small 1st pulse of the noise. Mostly it comes from reason no.2 - in my experience. 

Connected with the mentioned problem is: the displayed amplitude value may be unreliable.

The user of a TM can find the real reason only with a scope that indicates the assumed tic start.

Frank

Link to comment
Share on other sites

1 minute ago, praezis said:

Problem with Weishi, even with Witschi Watch Expert, is: you cannot say if the scattered line comes from extra noise or from neglecting the small 1st pulse of the noise. Mostly it comes from reason no.2 - in my experience. 

Connected with the mentioned problem is: the displayed amplitude value may be unreliable.

You know that the issue (which can be minor and negligible) is with the mov.t, when a good mov.t of the same type a produces a straight line on the same machine. Sometime, but not always, one can reduce or eliminate spurious dots adjusting gain.

In my direct experience, these instruments are reliable when it comes to basic parameters and line pattern. But, as also mentioned in the recent thread, what they can't do is to precisely pinpoint (and even reproduce, Weishi does not play actual sound on speaker - bad) "strange sounds", as below. I had the pleasure to be there when Mark showed me that and produced the video.

 

Link to comment
Share on other sites

Hi,
to understand the issue that I am referring to, you have to really understand how a TM works and how its displays and dots are generated, how amplitude is measured. You have to understand the importance of detecting the 1st pulse of the noise correctly - else faulty display will show, with Weishi or any other brand.
Of course the movement is always origin of issues, if you want. Two identical movements can have very different noises, though ok, but one easy to evaluate, the other very hard to evaluate. Here the issue goes to the TM: One can evaluate a challenging noise still ok, the other will produce incorrect results. I am not talking about interfering noise somewhere between escapement noises, like in Mark's video.

I am in full agreement with Marc, when he underlined the importance of a scope (wave form) display. You are nearly blind without.

A remark to your attached video:
No doubt that the balance was rubbing a screw head, resulting in bad performance.
However I doubt very much the conclusion, the visible extra noise came from there! In contrast it looks like a very typical ringing hairspring noise.
Why? Normal they occur in one half swing only, not in both - as in the video. A rubbing balance would generate noise in both half swings, one noise with every rubbing spoke, guess how many in a full swing, with 4 spokes? Removing and mounting  the balance can remove this (harmless) noise already.

Regards, Frank

Link to comment
Share on other sites

1 hour ago, praezis said:

to understand the issue that I am referring to, you have to really understand how a TM works and how its displays and dots are generated, how amplitude is measured. ...

Be reassured that I do understand that :biggrin:. I agree that here can be cases where a minor issue, like a slightly weaker 1st pulse cause the machine to produce an erroneous dot. However your reasoning contrasts clashes with an ineludibile fact: If two identical mov.t give good Vs bad results on the same machine, the fault is with the mov.t, not the machine. All the rest is idle speculation.
We can also wait for the OP to not just test a "good watch" with his machine, but also test the watch in picture on another TM. My bet is that it will show the same result.

Quote

No doubt that the balance was rubbing a screw head, resulting in bad performance.
However I doubt very much the conclusion, the visible extra noise came from there! 

So according to your theory, something could be silently rubbing? That would be a miracle in physics!

As mentioned above, I was there when Mark diagnosed and fixed the issue, so I have seen with my eyes marks on the balance spokes, then the problem going away after the pallet bridge was tightened. Attached the original sound capture, there are two parts, first dial down, and 2nd dial up, were the balance weight causes the rubbing, and sound to be much more pronounced.

PaulosSeiko.m4a

Link to comment
Share on other sites

 

Quote

Be reassured that I do understand that :biggrin:

Honestly, I was a bit in doubt :biggrin:.

Quote

If two identical mov.t give good Vs bad results on the same machine, the fault is with the mov.t, not the machine. All the rest is idle speculation.

Please read again my previous post.

Quote

We can also wait for the OP to not just test a "good watch" with his machine, but also test the watch in picture on another TM. My bet is that it will show the same result.

Bet accepted.

1 hour ago, jdm said:

So according to your theory, something could be silently rubbing? That would be a miracle in physics!

A miracle would be 4 spokes knocking twice per swing on a screw, giving just one single noise per swing :biggrin:. No theory at all, just looking on the facts and a bit thinking...

Thank you for the audio file! It proofs what I could only assume: a ringing hairspring, but no knocking (from 00:34).

Frank

  • Confused 1
Link to comment
Share on other sites

23 minutes ago, praezis said:

 

Honestly, I was a bit in doubt :biggrin:.

Please read again my previous post.

Bet accepted.

A miracle would be 4 spokes knocking twice per swing on a screw, giving just one single noise per swing :biggrin:. No theory at all, just looking on the facts and a bit thinking...

Thank you for the audio file! It proofs what I could only assume: a ringing hairspring, but no knocking (from 00:34).

Frank

Frank, can we not be argumentative please.

You have your opinion and we observed differently, and the watch was physically in our possession, the sound went away when the pallet cock screws were tightened - can we leave it at that and get back to the OP's issue?

  • Thanks 1
Link to comment
Share on other sites

Thanks for the responses, I didn't mean to cause arguments :) As I see now the whole thing about timegraphers is more complex than I thought.
Going back to my homemade device, here's an example of what the software shows in scope mode and a sample analysis it generates.
Question: is the signal from microphone strong enough looking at the first picture? Sorry for the poor quality, I'm still getting the laptop connected to the internet.

62adeb951e1a8c0e652afe69846f073f.jpg

2096ed4e4b029cf0a0a1ed7fb67f0ad7.jpg

Link to comment
Share on other sites

10 minutes ago, rafau said:

Question: is the signal from microphone strong enough looking at the first picture?

Are the two images above related to the same watch?

If yes, there is clearly a problem, because of the inconstant readings in scope mode, which seems to not be detecting a valid signal at all.

If not, please post matching pictures clarifying to which watch they are referring

Link to comment
Share on other sites

For the original question it's not valid in the context of this discussion. The original timing machines vacuum tube later on moving to transistors finally integrated circuits and up until relatively modern none of them had preamplifiers in the head unless there was something else going on sometimes. I've seen it in pickups that had stuff for electric watches some of the preamp might be in the base of the microphone. It's only really now were you see them putting the preamps in the head which is actually a really good thing. Otherwise the preamplifier was on the input of the timing machine circuit.

Then making universal assumptions is bad. Typical computer inputs are expecting capacitive microphones the Piezoelectricity Inputs are very different Impedance than what the computers expecting. The simplistic of this means depending upon a variety of things the microphone is either will work not work sometimes work and it just depends on the microphone and the particular computer. Then even with preamplifiers I've seen dramatic differences in the volume level required for various laptops.

Excuse that normally I'd read the entire thread but the discussion is too strange for me right at this moment of time. So other things that come into play as you've mentioned above you have to get a good clean signal. I've seen at work with some rather expensive Swiss timing machines we sometimes can't get a good clean signal because of a very heavy stainless steel case.  Then we just switch to another machine but even then some watches are just hard to pick to up if not impossible.

Then even if you get a good clean signal is it really a good clean signal? This is a bit of a play on a wording as to what exactly does a clean signal mean anyway? Timing machine is attempting to interpret what you're giving it and mechanical watches are not perfect which can lead to interesting interpretations of what the machine thinks it's getting. For instance I typically work on pocket watches I have a really nice expensive Swiss machine at work. It does multiposition timing with automatic microphone displays nifty numbers should be a pretty good machine shouldn't it? Yet on things like pocket watches for a variety of reasons they can have inconsistent power through the gear train. Inconsistent power results in inconsistent timing results but the normal timing machine that's not always apparent. I can run a time plot of a watch immediately run it again and again and the results will be different every single time. This is where other ways of looking at the problem are required.

Time plots are really nice for inconsistence power through the gear train or other assorted problems. The Swiss are interesting they don't always put all the features in one machine. So I have to go another machine for a time plot 16 minutes rate amplitude beat. A modern watch doing a time plot like this it's absolutely no variation but I watches that I have timing issues usually all see all sorts of bizarre fluctuations not always consistent in those fluctuations. Fortunately mechanical watches usually average out all of this timing machines don't which is why you get weird readings.

Then the software described in this discussion has another feature raw mode. You can see a description of the below some ways much nicer than also a scope in that are not restricted to a tiny bit of the waveform you can see quite a bit more see all sorts of things that oscilloscope doesn't necessarily show you.

http://www.watchoscope.com/manual.html#raw
 

Link to comment
Share on other sites

Hi Marc,

I apologize if I was too argumentative. But also would have preferred response on arguments that I gave in my posts. 

rafau,

can you post an audio file picked up with your equipment? Preferred as .wav. I will examine it and tell you if ok or too weak.

Frank

Link to comment
Share on other sites

Thanks for taking part in the discussion John. Looks like you know a lot more than me about the timing machines. I don't know If I need a preamp that's why I starter this doscussion. Please, have a look at these pics. Scope mode and the analysis mode. The same watch, quite an old Atlantic Worldmaster.

bb008beb990170905f33322acbbe710a.jpg

73986a1e8eb6ea300b854994dac016d8.jpg

Link to comment
Share on other sites

2 hours ago, rafau said:

Please, have a look at these pics. Scope mode and the analysis mode. The same watch, quite an old Atlantic Worldmaster.

Seems to me this software has a serious inconsistency, because in pattern mode picks up the beat and give meaningful results, but in scope mode doesn't. Mode should be indifferent for parameters to show, does the documentation says otherwise?

Link to comment
Share on other sites

To understand what you're looking at let's look at an absolute perfect idealized view and see were the sounds come from. There's a link below it's a great training video for the lever escapement that's not what was designed for. So at about one minute in it shows the lever escapement it lasts just under a minute and all the rest of the video is totally irrelevant.

https://www.youtube.com/watch?v=g5c5RK4WFV8

So Study the video you can see what things are supposed to look like. The timing machine only cares about two parts the waveform. The first the roller jewel crashing into the fork this a very quiet sound this is why often times you need amplification to get that up to a usable signal. Then locking and crashing into the banking pin but technically it's locking everything else irrelevant for the timing machine.

Then before we look at the waveforms we have a problem and I have a saying. My saying is garbage in garbage out if the waveforms Are not right because you don't have the right amplification you don't have a preamplifier you don't have a good clean signal then all the interpretations are a total waste of time. Basically total waste of time to try to interpret garbage and if the garbage is legit which I doubt this watch has some very serious problems. So for an experiment find another watch preferably a large pocket watch pocket watches are really noisy let's see if anything looks different.

So how waste a little of my time look at the Two screens what dramatically jumps out as that's why isn't it? What is the model number of this watch what is the frequency that it runs  at? This is a magical watch it's running at both 18,000 and 21,600 makes it a rather interesting watch doesn't it? I would almost think you have two watches the numeric results of both screens are so dramatically different. But this is what would happen if you waveform sucks timing machine in this case software can't interpret things correctly. Your 400 seconds off the amplitude of 300° and the skip over the beat then in the other one things look more realistic two separate watches perhaps or garbage for signal.

Chinese machines have problems with very low amplitudes they misinterpret the middle part of the waveform as the locking signal then you get really nice high amplitudes which aren't there.So the waveforms could be right watch has a very serious problem but I think we need another watch to see if you're just not getting a clean signal it's a total waste of time to interpret the garbage were seeing. Even more interesting when the two screens are so dramatically different then oscilloscope should agree with the other one the numbers should not be this dramatically different. Just to something that's not right at all here.

 

Link to comment
Share on other sites

15 hours ago, rafau said:

I will need to have a deeper look into user manual then. The pattern mode looks fine as you say.

I've noticed now that on both scope mode pictures above the font for parameters is dimmed. Probably that means that the SW considers values or signal to be unreliable? Then in pattern mode the bad watch parameters are in red and the good one in green. That makes sense but still doesn't explain why the same watch shows differently in the two modes. 

Link to comment
Share on other sites

  • 3 weeks later...
On 1/3/2018 at 10:55 AM, rafau said:

Thanks for taking part in the discussion John. Looks like you know a lot more than me about the timing machines. I don't know If I need a preamp that's why I starter this doscussion. Please, have a look at these pics. Scope mode and the analysis mode. The same watch, quite an old Atlantic Worldmaster.

bb008beb990170905f33322acbbe710a.jpg

73986a1e8eb6ea300b854994dac016d8.jpg

Did anyone notice that the two different modes are actually detecting two different BPH? One is 21600, the other is 18000, there's definitely an issue somewhere, either with low signal or something else.....just my observation.....

Link to comment
Share on other sites

46 minutes ago, khunter said:

Did anyone notice that the two different modes are actually detecting two different BPH? One is 21600, the other is 18000, there's definitely an issue somewhere, either with low signal or something else.....just my observation.....

Actually  that has been mentioned above. When it shows 18,000 none of the other parameters is valid either.

OP has not reported further.

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.



  • Recently Browsing

    • No registered users viewing this page.
  • Topics

  • Posts

    • people be honest.... Swatch is evil for the watchmakers and repairers, BUT not everything in watches from Switzerland is from the Swatch-Group. As far as i know, Selitta got sacked by Swatch as a Movement-Assembler for them and they started to produce Movements in their own Name with slight Modifications. As far as i know, they sell Parts to the Market for their Movements. In most cases, if a ETA-Movement fails, it is a valid Option to replace it with a Selitta Movement, which i consider the Solution for this Mess with the Swatch-Group...... I have no Connection to anybody at Selitta, but being a Swiss-Guy, i still like to have Swiss-Made Watches, but not from the Swatch-Group.   ok ? regards, Ernst
    • Just one more greedy act by Swatch. They started a number of years ago here in the US..cutting off supplies to watchmakers that could build complications that many Swatch houses couldn't even touch. Old school masters who had gone through some of the most prestigious houses in the world. Otto Frei has some statements on their page about it. I tell all my customers to avoid new Swiss watches like the plague,..unless they just want an older one in their collection that still has some parts out on the market, or they have really deep pockets and don't mind waiting months and paying through the nose to get it back. Plenty of others to choose from..IE Seiko,..or other non-swiss brands Even a number of Chinese brands are catching up with the Swiss,..and I think that in time, their actions will be their downfall
    • Yes. If that's not what you are experiencing...start looking for something rubbing. A 1st guess is that one of the hands is rubbing against the hole in the center of the dial. Especially if you now have lower amplitude in face up/ face down positions.
    • Once a movement has the dial and hands put back and it is recased, would you expect the assembled watch to have the same amplitude as when the movement is in a movement holder and is without hands and dial? Thanks
    • C07641+ not sure what the "+" is for after the last digit.
×
×
  • Create New...