Jump to content

Recommended Posts

Guido, as I explained shortly after your original post, you can't use a .wav file to calibrate Watch-O-Scope. It will either give zero error (if the sound card's recording and playback circuits use the same internal clock), or a random error (if they use different clocks, or you are playing it from a different computer). It is like using a photocopy of a ruler to measure how accurate the ruler is.

Link to comment
Share on other sites

Regarding the wav file, not recorded, was created a machine capable of saying, "runs a rate 0.0 s / d, with 262 degrees of amplitude and beat error 0.0". It is difficult to find a watch that works well.

 

It means that when you run the file to a computer sound card which runs "perfect", must have an error rate close to zero.

 

The source from which I obtained the wav file is very, very reliable.

 

If I say for example that a penny on the dollar measure 19.05mm, that way I know if it is true or not? Option One: to say "No, because that is not a measuring instrument." Option 2: Say: Maybe, but there is no way of knowing. Or Option 3, take your best instrument and measure it.

 

With all due respect I can ask this question: Have we tried?

 

Guido

Link to comment
Share on other sites

Guido, I will try your file. I believe you that the wav file is perfect (it's fairly easy to create such a perfect wav file), but the device you're playing it on (my sound card) is not perfect.

 

Regarding the penny, the only way to know if it is exactly 19.05 mm is to measure it with a known accurate ruler or caliper. What you are proposing is like measuring it using another penny.

Link to comment
Share on other sites

I just tried it on my home computer (the one I normally use Watch-O-Scope on). With the correction initially set to 0, your file gives me an error of -2.7 seconds per day. With the correction set so my quartz reads accurately, your file gives me +12 seconds per day.

 

The quartz watch is known to run at +0.3 seconds per day (it has gained 7.5 seconds since November 1st).

 

I have verified (by looking at the file in Audacity), that your file is indeed perfect. There is one tick beginning every 8820 samples, corresponding to exactly 5 ticks per second. So, what that +12 seconds per day reading is telling me is that there is a +12 seconds per day error in the playback circuitry of my computer. Put another way, instead of playing back exactly 44100 samples per second, my computer is playing back 44106.3 samples per second.

 

I will try the file also on my computer at work when I get to the office later this morning.

 

post-140-0-23435600-1448629769_thumb.png

 

 

Link to comment
Share on other sites

Dear Stefan, thank you very much work to do for me. First, I do not have a quartz watch as a standard. Also I have no instruments or knowledge that you have. Thank you. The data you put in your previous post are very useful to me. I've tried it with only two computers. I will be very grateful to resuldados you get in your office.

 

Sincerely,

 

Guido.

Link to comment
Share on other sites

I just tried it on my laptop this morning. When my laptop is reading the quartz watch correctly, it reads your sound file at -2.7 seconds per day.

 

I have learned something very interesting though. On both computers I've tested so far, even with the rate correction in Watch-O-Scope set to zero, your file does not read 0 seconds per day. This means that both sound cards' outputs have rate clocks that are separate from their input rate clocks.

Link to comment
Share on other sites

Now I have more questions than before.

 

I'll thank me much to explain their findings. In my case presents the same. In my two computers the error rate of the sound card to run the audio wav file, (I run it in Cool Edit Pro set to loop), the result is -2.7 sec / day. ???

 

Its hard for me to understand. But delanalisis sure you do, I will know more.

 

Guido

Edited by guidovelasquez
Link to comment
Share on other sites

Oh, that is strange. You're right, both of mine were -2.7 s/d, and both of yours are too?

 

I looked at the raw data in the wav file again, and it turns out that it is not as perfect as I first thought (I only looked at the first 10 seconds last time). The first tick starts at exactly 0.000s in the file, and the the 800th tick, which should occur at 159.800 seconds actually occurs at 159.805 seconds. That's an error of 0.005 seconds late (slow) per 159.800 seconds, or 2.703 seconds per 86400 seconds, which is -2.7s/d.

 

I think we can conclude that the sound card input and output run from the same internal clock, and that if anyone uses your wav file with Watch-O-Scope set to a Rate Correction of 0, they will get a result of -2.7s/d. Unfortunately, that also means that even if your wav file were perfect (i.e. 0s/d), it would not be useful for calibrating Watch-O-Scope, since, as I've put it before, you're measuring the accuracy of your ruler with the same ruler.

Link to comment
Share on other sites

I have a question: How can watch-O-Scope determine who truly wav file has an error of 2.7 second per day, while the "Rate Correction to" 0 "regardless of the computer on which the test is performed? If the WOS is a rule and depends on the computer clock, why the same results are obtained on different computers if we have to assume that every sound card is slightly different, ie that has a slight imprecision?

 

I made one last test and I think it will be the last thing I do.

 

I ran my file * .wav files from my desktop. The analog sound got it from the headphone. What brought you to the microphone input of my laptop. And guess what happens, the "Rate Correction" being "0" ...? the rate is 5.8 !!!

 

post-670-0-34495900-1448848790_thumb.jpg    post-670-0-89056200-1448848812_thumb.png

 

Guido

Link to comment
Share on other sites

The sound card input and output use the same internal clock, so Watch-O-Scope will show exactly the error in the .wav file, regardless of what the speed of that internal clock is. Suppose the sound card clock was wrong by -200s/d (that is, it plays too slowly). That means that the playback of your file (which already has a -2.7s/d error in it) will come out at -202.7s/d. Now, you feed this into the sound card input, which is also wrong by -200s/d (it records too slowly). That input speed error will cancel the output speed error, and Watch-O-Scope will just see -2.7s/d.

 

To use the photocopy analogy again, suppose you have two rulers, and there is a 1% difference between them. Now suppose you put them side-by-side on the photocopier, and set the photocopier to make an 80% sized copy (i.e. a 20% reduction). On your copy, your two rulers will now be 80% of the size they were originally. But, there will still be exactly 1% difference between them.

 

Regarding your experiment with the laptop and desktop computers, you are playing a file with a known -2.7s/d error, through a sound card in your desktop with an unknown error, into another sound card in your laptop with a different unknown error. The result could be anything, and thus does not give you any useful information.

Link to comment
Share on other sites

Thanks for the explanation Stefan, as I mentioned above this would be my last attempt to understand a "mystery". Well, in science there are no mysteries. Simply unprepared minds to understand some phenomena.

 

Leaving aside the calibrated with a * .wav. I have a problem. my quartz watch with which to gauge desire WOS, I did a test. for a month. The resusltado is to come forward 10 seconds in a month. Which means that only 0.3 seconds a day ahead.

 

However when analyzed with WOS, this is the result.  The "Rate Correction" was set to "0".

 

post-670-0-18441700-1448903763_thumb.png

 

 

Think WOS analysis, it is affected by the inhibition.

 

Sincerely,

 

Guido

Edited by guidovelasquez
Link to comment
Share on other sites

Hi Guido:

 

The result indicates that you must set the rate correction to -4.1, because 4.4 - 4.1 = 0.3.

 

Since you ran your test for about 11 hours, inhibition will not affect it. The inhibition interval of most watches is very short.

 

Those large spikes at just after 6:00 and 7:00 were probably caused by a loud noise, either in the room, or in the movement (for example, the date changing). But as you can see, they had a very tiny effect on the average.

Link to comment
Share on other sites

The model number of the sensor is BM15015-06HC

 

I am going to order some of these to experiment with. Allied gave me a lead time of 2-3 weeks, so that's pretty reasonable.

 

I was looking at the datasheet for this sensor, and one of the applications they list is "phono cartridges". That makes me wonder if a phono cartridge might be usable as a watch sensor. I'm going to try placing a watch on my old turntable (not turning of course), and place the needle on the crown to see if it will "hear" the watch.

Link to comment
Share on other sites

On the sensor. I tried the turntable about 25 years ago ..... did not work and I think there are two types. magnetic pickup and the crystal .... I remember is the crystal. But they still need a pre-amplifier with very high gain.

 

Vibrations in "needle" the turntable, are stronger than the "tick" of a clock. I did it with a Grundig brand Radiola my father .... always looking for answers in my inquisitive mind, I wondered: "How could capture the sound of a clock, like what makes a" timingmachine "" (my first contact with a "timingmachine" was about 35 years ago in watchmaking my uncle. he had one. in fact the microphone is currently I have it.)

 

Guido.

Link to comment
Share on other sites

Regarding irregularities in the graph, I do not think they have to do with some acoustic disturbance. Well, I used an electromagnetic sensor for the magnetic pulse of the clock. It may be due to some electrical interference, switch or similar.

 

post-670-0-88027100-1448992689_thumb.jpg

post-670-0-86813600-1448992587_thumb.jpg

post-670-0-27936300-1448992723_thumb.jpg

It is stuck with a similar Rodico paste. (Tak Pritt)

 

Guido

 

Link to comment
Share on other sites

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 than watches now.  One thing that has always been very time consuming is the regulation and having come across your site a few weeks ago I sat down and constructed your pre amplifier and microphone. I used RCA jacks instead of the 3.5mm audio jacks as good quality RCA cables are plentiful in my house. I get a very clean signal and all I can say is Wow. The accuracy with which I can regulate a railway grade pocket watch is just phenominal. A Railway pocket is supposed to be good for 5 seconds a day and I have now been able to set one at under 2 seconds per day using your program. I do not post on forums and I have specifically joined this one in order to thank you sincerely for an amazing job that you have done. Not just producing the program but for making it available to anyone who wishes to use it. My tool of choice prior to the advent of Watchoscope was a stethoscope. I would count the ticks over a 15 second period, multiply by 2 and by 4 and then by 60 to get BPH I would then compare that with what the BPH should be, get it as close as I could by ear and then do the fine adjustments over the following 2 or 3 days. Now I can regulate a clock to perfection in under a half hour. Truly amazing so thank you again. I particularly find the Raw field incredibly useful, by observing the variation in the timekeeping ability of the watch, I have been able to diagnose several different problems with multiple watches so I know before I even strip them precisely what to look for. Again, truly amazing. Because I have grown up using BPH as a yarstick for regulation, If I hace a wish list for your program ,it would be to have a field showing real time BPH on the scope page. If you ever make this addition, please post it on this sie so that I can download the upgrade to the program. I see from your profile picture that you are into flying, light aircraft, Even at my old age, I still fly my Titan Tornado at least 4 times a month, It keeps me young. Thanks again.

Best regards

Parrick

  • Like 5
Link to comment
Share on other sites

Parrick, thanks for your kind comments on Watch-O-Scope. I'm glad that you're finding it useful. Be sure to post some pictures of your pocket watches, and your Watch-O-Scope set-up. Your stethoscope method must have yielded a pretty inaccurate result on the initial adjustment, since even an error of 1 tick in 15 seconds on an 18000bph watch would represent an error of 20 minutes per day.

 

Jamez361, I just received notification from Allied that my vibration sensors have shipped! Hopefully I'll have some time over the holidays to incorporate one into a new design for a microphone (and perhaps low-gain amplifier).

Link to comment
Share on other sites

svorkoetter. Thank you for taking the time to reply to my post. I certainly am finding it useful. I am not a computer person so I have asked my daughter to come over this weekend to teach me how to post pictures. You would be amazed how accurately you can count with a few years of practice. The stethoscope was a later addition to supplement my ageing hearing. By ear I am still able to set a clock/pocket watch to within a minute a day. I have amazed many of my fellow clockmakers with that trick but thanks to you I no longer need to continue with it. The reason that I asked for a real time display of BPH is so that I can compare it with what my ears are telling me. I seldom use the auto function on watch o scopes BPH setting field as I count the beats and then tell the program what the BPH should be. Also when I do the count I can usually tell how fast or slow the watch is and it amuses me when watch o scope confirms my count. Thanks again. Regards Parrick

Link to comment
Share on other sites

@Parrick If you want a real-time BPH and you are listening to loud watches or clocks :) you could try installing Wildspectra on an Android phone. It doesn't come close to the accuracy and user-friendliness of Watch-O-Scope nor does it show you the amplitude but it shows you the BPH. I found the application quite useful when I didn't have WoS but now I find it more then less accurate (I believe it's because of the Android OS that doesn't provide a quality clock to the application).

Link to comment
Share on other sites

  • 2 weeks later...

I am going to order some of these to experiment with. Allied gave me a lead time of 2-3 weeks, so that's pretty reasonable.

 

I was looking at the datasheet for this sensor, and one of the applications they list is "phono cartridges". That makes me wonder if a phono cartridge might be usable as a watch sensor. I'm going to try placing a watch on my old turntable (not turning of course), and place the needle on the crown to see if it will "hear" the watch.

I managed to get my hands on a bimorph sensors and I just put it in the old circuit (the one with the piezo sensor).
What I can say is that I was expecting more. There is a slight improvement but I didn't have the time to play with it.
 
Stefan, did you get to try them?
 
Thank you,
Bogdan
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...