WatchUSeek Watch Forums banner

181 - 200 of 451 Posts

·
Registered
Joined
·
21 Posts
Inconsistency in data beat error.

Dear Contrate_Wheel. I have had many hours of doing tests with the program you do not a have done. I still admired the performance.

However I have found on several occasions some inconsistency in the data delivery on the beat error.

It is possible that the algorithm is choosing the third sound of the entire event 3, (1: Unlock 2: Impulse and 3: Drop or blockage) However, the sound of falling or blockage that is the stroke of tooth wheel exhaust against locking face of the pallet, is not very stable, ie not as stable as coming from the true clock regulator that is the balance wheel.

Here also I have two screenshot made almost at the same instant. The first with tg and another with other software. I tested with two other software and the results are consistent with each other but not with tg.
beat error xx.JPG beat error xxx.PNG

Therefore asymmetry that is displaying the program is not the balance wheel but relationship escape wheel / pallets.

If the graph is derived from the time between an unlocking and the next, this screenshot is not corresponding.
Como calcula el beat error.JPG

The amplitude is not recorded. But if it is registered the beat error. If the amplitude is read between unlocking event to fall, How you are determined by the software here?




It may be that I'm doing something wrong. If it were so I need to make the necessary corrections.

Guido
 

·
Registered
Joined
·
87 Posts
Discussion Starter #182
It is possible that the algorithm is choosing the third sound of the entire event 3, (1: Unlock 2: Impulse and 3: Drop or blockage) However, the sound of falling or blockage that is the stroke of tooth wheel exhaust against locking face of the pallet, is not very stable, ie not as stable as coming from the true clock regulator that is the balance wheel.
Hi my friend Guido,

sorry for longly deserting this forum...
Thanks for your testing: you are exactly right about tg not using the first sound as its reference!

For the beat error, to be precise, tg uses the entire waveform of each tic as a reference. When the two waveforms do not match, as in your case, the third sound usually takes the lion's share.

When constructing this algorithm, I considered for some time if my choice might be an error. I was aware that most timing machines (and timing programs) use the first sound, but on the other hand that sound is difficult to detect in really crappy audio (remember that tg was designed specifically for really crappy audio). Unfortunately, no sound comes directly from the balance wheel: the first is determined by the impulse jewel impinging upon the lever, the third is determined by the locking of the escape wheel. So I got the impression (but I didn't experiment, therefore I don't know) that moving the banking pins would disturb the symmetry of the first sound, and altering the pallets' depths would affect the third sound.

I also reasoned that if there is a difference of several milliseconds in the two waveforms, whatever the reason, this fact must be as indicative of asymmetry in the escapement, as it is a several millisecond beat error. Trying to adjust a watch to 0.0ms indicated beat error when something else is off-balance by, say, 3ms, is, to my beginner's understanding, pointless. Timing machines that only give the beat error, might lead the user to believe that once the beat error is 0.0 then the escapement is in perfect symmetry. Often, looking at the graphs, I don't know the reason, but I know that it isn't.

Let me repeat that these are my beginner's impressions. If I am making any mistakes, I would be grateful for them to be pointed out.
 

·
Registered
Joined
·
21 Posts
Hi friend Contrate_wheel, thank you for answering my concerns about the software that you have made available to us. I mean what you have done with this software, it is fascinating. What explains that the program algorithm is designed to work with a very poor sound, it is true. And in fact I think that given the nature of the sound of a clock, it's amazing what the software manages to make with such a bad sign. However, that in itself is already an advantage. If you can achieve both a bad sign, sure you can get amazing results with a reasonably clear signal. My opinion is that as the software is now better focus our attention on finding a suitable microphone and preamplifier. A SUGGESTION THAT HERE IS WHAT I USE: reparacionderelojes.weebly.com/modificacioacuten-de-pyle-pp440.html


About the sound from the balance wheel, with all respect I would point out that sound the impulse jewel contact, YES it COME FROM THE BALANCE WHEEL. Please see this video that I have prepared: The Beat Noise Of The Swiss Lever Escapement.
The video REALIZE from this: Audemars Piguet escapement


Although I am not an expert on the subject, I still think it would be better to use the first sound waveform ticking for a more accurate beat error.


Thank you very much

Guido
Mi email. omegacronoARROBAgmailDOTcom
 

·
Registered
Joined
·
273 Posts
Has anyone tried this software with a piezo yet ? What were the results ?
I suggest at least 10% of the posts here are dealing with piezo-recordings. Just read through the postings and have a look for keywords like: piezo, guitar pickup, pickup clip, pick-up-clip, clip-on guitar pickup, conact microphone.

In short:
YES ;-)
Result: perfect signal (additional preamp recommended)
 

·
Registered
Joined
·
1,282 Posts
Timegrapher cf tg

Just a quick comparison.

The changing traces on tg are a result of my microphone not picking up the signals all the time.
Timegrapher is more stable but one can't complain about the value of tg. Fantastic.

FYI Watch was Parnis Big Pilot with a Seagull movement.

Timegrapher Parnis Small 20160830_121649.jpg
Parnis 20160830.JPG
 
  • Like
Reactions: contrate_wheel

·
Registered
Joined
·
3 Posts
Re: Timegrapher cf tg

Thank you so much to this community for the software, it is awesome. I have a watch that was losing about 20s a day and when I put it on this, the beat error also showed 1.8ms so I sent it out for service.

Quick question, while everything looks normal, it does not seem to record amplitude. It just shows two dashes ("-"). Is that normal?
 

·
Registered
Joined
·
87 Posts
Discussion Starter #191
Hi, my friends!

@Guido Velasquez: Thanks for your suggestion about the amplifier. Tg has now all basic basic features, so I really would like to focus on the amplifier design (and writing an operating instructions manual), I only need to find the time for these tings... I understand what you mean about sounds coming from the balance wheel, in fact, I have been inaccurate in my statement. I meant to say that no sound comes exclusively from the balance wheel, so there is always at least a second part involved.

@avihai: Thanks for shaing tg on instagram!

@cleef16: I understand that this thread is now rather long. I have collected the links and relevant install instructions on the github homepage of tg: https://github.com/vacaboja/tg The program is really straightforward to use (once you are familiar with the workings of lever escapements), but unfortunately I didn't find the time to write a manual. So, for operating instructions, you still have to read this thread.

@1afc: Nice! Was tg calibrated with the quartz watch procedure?

@jhermit: Yes, it is normal. The dash sign only tells you that tg is unsure about the amplitude, this happens usually because the sound quality is not good enough. Nevertheless, the rate indication of a calibrated tg is, in my opinion, very reliable.

As I said, my priority now, regarding tg, is to experiment a little with amplifiers and write a little user manual (also building a Debian package and fiddling with the amplitude algorithm would be nice...). Hope to do these things before Christmas or so. Thanks again for all your contributions to this thread!
 

·
Registered
Joined
·
10 Posts
Thanks for providing this software. Works great. Congratulations.
I took an old SKX007, unscrewed the back, and used my Logitech cam for sound input. I was able to regulate it to cca. +1 sec/day dial down. While the movement has significant positional variation it averages to +2 sec/day after 24 hours on hand.

 

·
Registered
Joined
·
1,282 Posts
Hi, my friends!


@1afc: Nice! Was tg calibrated with the quartz watch procedure?
Yes it was. I was surprised to see it so far off but I understand there is no need for atomic clock performance from a soundcard.
 

·
Registered
Joined
·
21 Posts
Hello contrate_wheel. Thank you very much for your work with the software. If I could have your email maybe you could start making a quick guide to using the software. I think I'd do it. In addition I have some documents that you might be useful. You could send the. Greetings.
 

·
Registered
Joined
·
2 Posts
Thanks contrate_wheel! tg is really cool and I got my SKX009 down from consistently -15 down to around +-1 -- it's only been a day or two of unusual activity, so hard to say confidently, but the initial signs are encouraging. I do have a question. When I have my watch being "scoped", I do see the s/d vary quite a lot when it's just sitting there not being moved (from +4 to +7 or -11 to -23!), and I don't know if it's an artifact of my laptop, or if my watch is just screwed up -- it does have significant positional variation (I've seen -29 [12 up] to about +17 [face up]). Unfortunately, I don't have a second watch, nor a hardware timegrapher and so am curious if anyone else has seen this. Or do I just need to find a local watchmaker who's willing to put it on their timegrapher to see? I also didn't have a good rig before I regulated the watch, so I don't know if it was this way before I fiddled with it.
 

·
Registered
Joined
·
87 Posts
Discussion Starter #196
@1afc: I was surprised too that audio cards can be so much off, but apparently it's common (there was even some more or less scientific study on the subject, that I might have linked in one of the 20 pages of this thread, it was motivated by issues of audio/video synchronization, or something like that). I was lucky that the two machines on which I developed tg were almost spot on, so initially I made the incorrect assumption that this was the norm...

@guido: Thanks, I will send you my e-mail address by PM.

@drewcsillag: It's very hard to say without a recording of the watch, or at least a screenshot of tg's trace. Could you share one of these?
 

·
Registered
Joined
·
1,282 Posts
I've read a lot of the posts but I may have missed something so I'll ask anyway.

tg is a fantastic piece of software and I enjoy using it.

However, my main microphone is a an old USB voip phone that I have skinned down so the mic is not be obstructed by buttons etc.

It works ok but is already many years old and I'm looking for a new one to take it's place because it will eventually break.

So the questions are
1. please let me know what USB mic you use and where you got it?
2. is it a piezo mic?
3. I have a soldering iron but that's about the extent of my electronics capabilities so has anyone found the need for an amp/filter and are there any workable modules that can be used without fabrication?
4. Has anyone devised a good stand for positional checking?

TIA
 

·
Registered
Joined
·
3 Posts
FYI, I am also experiencing the "inaccurate beat error" reading @Guido described. Maybe it is because of sound quality issues, but TG was reading a 1.8 BE and a Timegrapher was reading .1.
 

·
Registered
Joined
·
87 Posts
Discussion Starter #199
FYI, I am also experiencing the "inaccurate beat error" reading @Guido described. Maybe it is because of sound quality issues, but TG was reading a 1.8 BE and a Timegrapher was reading .1.
Thanks, this thing has been discussed over and over again. Well, I guess that I should do what the users want, therefore I will try to change tg so that it reads the same beat error as most timegraphers.

A recording of your watch demonstrating the bug would help me a lot in this. Could you send me one?
 

·
Registered
Joined
·
121 Posts
Thanks for the software!

I have an old MBP15 with Mac OS Sierra on it. Installation went smooth and look like it works fine. I'm using a mic of my Meizu headset right now and it doesn't seem very reliable. I know my watch is within 20 seconds and TG shows from -80 to +50. I guess I need a better microphone right?

All those cheap contact (guitar) microphones have a 1/4" plug (either male or female) and I'm not sure how to pair it with the MBP. Is some kind of 1/4 to 1/8 adapter sufficient? Please share your setup if you have a MacBook or something similar.
 
181 - 200 of 451 Posts
Top