D. I. Y. Watch Timing Machine.

[Forester]

isup.me say it is just me :)

I don't use a firewall, so maybe my ISP cause this problem

 

I try on my iphone using wifi, same problem (so no computer related problem), using data on the iPhone, no problem at all..

 

Still strange, never had any problems related to this with my internet provider.

[cdjswiss]

Today I have today been really impressed by Watch-O-Scope v1.1.

 

I connected my Chronografic mike via a Kemo preamp with only x100 amplification. Amplifier power was from a LI-ion battery. The whole lot was floating on the PC earth.

 

 

 

The signal to noise was fair, and after applying the Watch-O-Scope broad band filter it was excellent. Here is a sample measurement made on a home-built watch with an ETA 2824 movement. The PC clock was calibrated following the instructions in the user manual.

 

 

And here is the same watch, dial up, in my Timegrapher 1000. A good agreement for timing but proving my suspicion that the Timegrapher, although accurate for timing, is low on amplitude measurement.

 

 

 

 

[Geo]

What a nice clear readout Colin, I can see how you are impressed with it. Do you think your timegrapher has alway been wrong, or have the results changed over time?

[cdjswiss]

What a nice clear readout Colin, I can see how you are impressed with it. Do you think your timegrapher has alway been wrong, or have the results changed over time?

It has always been low on amplitude and I had assumed the need for more cleaning and oiling or replacement of mainspring (this last I have done twice). But my suspicions were aroused when I made a video of the balance swing and could fairly clearly see that the amplitude was larger than that reported by Timegrapher. Now Watch-O-Scope has confirmed this.

[Forester]

Nice setup Cdjswiss, I like the idea with the Kemo preamp. Building the suggested preamp would also be nice to do, maybe later if I can find the time.

Would the Kemo preamp perform with a piezo speaker also? What model Kemo do you use? (can't read it on the picture)

Great project svorkoetter, big thumbs-up...

[svorkoetter]

Thanks Forester. The Kemo preamp, from what I've been able to glean on-line, has a gain of only about 100x. That may not be enough for use with a piezo, depending on the sensitivity of your sound card's input.

[Forester]

I thought I had an easy way out...

On the other hand, buying a Chinese timer is also an easy way out, buying a new watch also, but that is not why we are on this (great) forum :)

 

Ok so with a Kemo (or any) preamp I would need a decent microphone, the piezo route would prefer you're amp.

I have no problems with soldering, it's quite some fun to do, it is just that I have no experience with making my own pcb.

Last winter I built my own (Lima) SDR radio, as it came as a complete kit.

Maybe I will try to make the pcb. I did read you're way using a sharpie, there is also a way using a (hot) iron and printed paper?

The projects are piling up.

 

Thanks again

[svorkoetter]

Just some random findings and thoughts...

 

• We recently changed the halogen light bulbs over my wife's desk to LEDs. Now, whenever those lights are on, I get spikes at a rate of 120Hz showing up in Watch-O-Scope. These spikes have fast enough rise times that they make it through the filter (i.e. sound-wise, it's a buzz, not a hum). Something in the driver circuits in the LED bulbs is creating this electrical noise. I used to have the same problem with the halogen lights, but only when they were dimmed. So, I have to turn those lights off when using Watch-O-Scope at my desk. Unfortunately, we're planning on switching my lights to LED too.
• I'm beginning to think that a large part of the AC hum problem that some people are seeing (despite all the filtering) is related to the use of a large piezo disk. All the commercially available watch microphones seem to have a very tiny piezo. I think the large disk literally acts as an antenna, and the hum is coming in on the audio ground line, not the signal line. I'm going to experiment with cutting a very small part out of one of those disks, and gluing that to a plastic disk.

[cdjswiss]

Interesting - I have 3 LED lights over my desk. One main 50 Hz Osram 220V 10 Watt LED bulb and a whole array of 12V DC LEDs in two desk lights (not that I should expect any 100 Hz noise from these) but I have not seen any 100 Hz noise from the main light when using my Chronografic mike with the x100 pre-amp.

 

I lowered the main Osram light (it's on an antique pulley system) too just 15 cm above the mike - still no 100 Hz

 

 

I guess that your noise source is, as you suspect, the large piezo disk.

 

Colin.

[guidovelasquez]

Indeed, the piezoelectric same disc captures the signal from the network of 60 Hz. But so does the cable. I improved a lot my microphones using a two-wire plus shield. I leave the shield connected to the system chassis the other two carry the signal and ground.

 

Please can you tell me how to include images in my post as you do?

 

Thank you.

[svorkoetter]

Guido, just click "More Reply Options" when writing your reply, and you will be taken to a full editor where you can insert pictures too.

[matabog]

Hello!

 

I want to share with you the following:

 

The test board plan of Stefan's design with a FET at the entrance:

 

The test board plan for JohnR's design (only the second stage amp filtering is different, the FET at the entrance and the first amp are the same):

 

The schema for JohnR's design:

 

Please pardon the hand-writing effect, but I don't have the time to put them in specialized editor, but maybe they help somebody.

I implemented both amps on the same test board with switches. Stefan's design is filtered at 400Hz high-pass and JohnR's at 4kHz. I find the later more environment-noise-friendly.

 

Bogdan

[guidovelasquez]

Dear Friends,

 

I enjoy visiting this forum. I have learned many things. Now that it is addressing the issue of software Watch-O-Scope, I also want to share my experience with this application.

 

 

Here I include some reading I've done with microphones that I built. I also have one of an old machine control clocks.

 

 

I tested the performance running on Windows 7 and Windows 10 and is excellent. Here Running on Windows 10.

 

 

Regards.

 

[guidovelasquez]

Here are some photos of the microphones I use. Even that is a professional unit "timing machine", I had to change the microphone cable, by one of three two signal wires and shield.
    
 
This is a microphone built from a piezo electric drive something big. I shield with aluminum foil. It is the hallmark of a tin of powdered milk. Cut to create the shield.

 

 
Here the softwere Watch-O-Scope, running on Windows 10.

 

Regards.

[matabog]

please share a photo with the waveform, to see how the amps behave.

 

Thank you,

Bogdan

[guidovelasquez]

Here is the configuration of the sound card. It can be seen that the gain is at the lowest level. With little volume on the microphone.

 

Configuration of the sound card

 

 

 

 

 

 

In this picture is the waveform sound settings above. The only caveat is that the preamplifier PYLE PP444 was modified.

Please see the following link to see the change.

http://reparacionderelojes.weebly.com/modificacioacuten-de-pyle-pp440.html

 

"We believe" that by making the amendment changed the response of removing the RIAA equalization curve. Any comments will be well coming.

 

 

Regard

[guidovelasquez]

Here another clock and sound settings used. Image clock waveform catches and MP3 audio file.

 

 

 

Sicura_EB8021_Synt0917-120315.mp3

 

 

 

[matabog]

That is a clean sound!

[guidovelasquez]

Hello friends. I want to share with you this method of calibration for the sound card of the computer to reliably use the Watch-O-Scope software.

 

Calibrating Your Sound Card Clock

To use Watch-O-Scope software.

 

Setting Up the Hardware To Calibrate.

 

 

1.       Beat rate:  Auto.

2.       Lift angle: 52

3.       Averaging Period: 20 Seg

4.       Setting Time: 5 seconds

5.       Vertical Scale: 1 dots/milliseconds

6.       Scope Sweep Time: 1 second / sweep

7.       Rate Correccion : 0  Zero

8.       Apply Digital filtering to audio input:  Ok.

 

Sound settings in Windows.

Disconnect any microphone.

Go to the sound in the Windows taskbar click the right mouse button and choose recording device.

 

Enable stereo mixer or similar.

 

Click Properties.

 

Enable the LISTEN AND THEN OK.

 

Run the audio file Calibration.  

 

Download the audio file in wav format. Calibration.

Cal.2min 40seg.wav

 

 

Using the audio file in Windows Media music player or similar. Set in continuous function. If all goes well, you should be able to hear and display it in Watch-O-Scope

 

 

In the Scope mode we can see the intensity and appearance of the waveform. You can make adjustments to the volume control.

 

When configuring sound sensitivity thresholds should take at this level:

 

At this level. They climb to where they are the valleys between about peaks.

 

Obtaining difference

The audio file lasts 2 minutes and 40 seconds. You only need one minute to determine the error in the sound card on your computer. The result will be something like this:

 

This image is one result: -2.8 sec. This is uncalibrated. It was the result of my computer. Each computer will be different.

 

Application of difference.

The correction is reversed. If the program indicates that there is a difference of -4 seconds. You must place a positive number. 4. But if the test program indicates say +3 seconds, the setting will be -3. Set this value in the "Rate Correction" in Settings box.

 

Once calibrated the card the result will be something like this.

 

The ideal result should be:

DAILY RATE: 0.0 s / d.

AMPLITUDE. 263 or 264 degrees.

BEAT ERROR: 0.0.

 

Do not forget after calibration set the sound with the microphone as the default.

 

 

 

 

[svorkoetter]

Thanks for this Guido, but I'm afraid it can't possibly work.

 

The calibration signal is coming from the sound card, which may or may not use the same internal oscillator as the sound input. If it is using the same oscillator, the error will always appear to be zero. If it's using a different oscillator, then there is no way to know what the error is, since you don't know what the playback rate error is.

 

This is the same reason you can't use a recording of a watch played on an MP3 player to calibrate Watch-O-Scope, since you don't know what the playback rate error of the MP3 player is.

[guidovelasquez]

Dear Stefan, thank you for your explanation. I certainly recognize that you are the experts. And honestly I admire and respect the work they do. You make. In my very finite mind, and with the limited training that I have, (not college I studied) there are things that are beyond my understanding.

 

My idea is this. For a machine that measures the time (clock) accuracy has to be expected. All clock to the most accurate, fail, ie it goes ahead or fall behind. The machine that calibrates or otherwise determines if a clock must be much more accurate than the clock that is being scrutinized.

 

The machine that determines the precision of a watch must have a pattern. Fortunately the "clock" of a sound card is vague, but stable. That is, if five seconds ahead always will.

 

What we need is to know how fast or slow the "clock" of our soundcard. I need a "pattern" or standard.

 

This is my idea. If I have a pattern that is precisely how fast or slow. (This is relative because even fail atomic clocks one second every eighty million years.) I know if my system is good, fast or slow. If reproduce that pattern (the ticking of the clock) from another device, say an mp3 player or the like, of course I have no idea what is the accuracy of playback device.

 

 But if I do the same player that uses the clock of my system (ie: I use the player for my cumputadora) and I am sure that my boss is right, I know if there is a difference in my system or more exactly on the card sound of my computer.

 

The question is: Do I lose something try?

 

Guido

[svorkoetter]

Everything you say is correct Guido, except for one part: "But if I do the same player that uses the clock of my system (ie: I use the player for my cumputadora) and I am sure that my boss is right, I know if there is a difference in my system or more exactly on the card sound of my computer."

 

The problem here is that the perfect recording that you are playing will have an error added to it by the computer's player. Usually, that player uses the same internal clock as the computer's recorder. The two errors will cancel out, and the sound card will appear to be 100% accurate, even though it could be very inaccurate.

 

Let me make an analogy. Suppose you have something you want to measure the length of, but you don't have a measuring stick. So, you take a wooden stick, and make evenly spaced marks on it, to make your own measuring stick. But you don't know how accurate it is, so you take another piece of wood, and copy the marks from the first stick onto the second stick. Now you use that second stick to measure the first stick and conclude that the first stick is 100% accurate. But clearly you haven't proven anything at all.

[matabog]

I want to show you my six positions "microphone". It's quite bulky, but it does it's job:

 

 

 

Besides some nuts and bolts and a lamp base from Ikea (Hemma) I used a cheap case holder:

 

[guidovelasquez]

Impressive creativity for construction is this microphone. I will try to build one like that. I'm using a professional, but not mine. The owner will ask me how soon, and I'll have to return.

 

I'm thinking of using something like this.

 

 

With the idea that you have raised here think mine build.

[JohnR725]

Another option that we have for making a microphone is 3-D printers are becoming readily available. So making something out of plastic isn't going to be as expensive as it used to be and with a lot of freedom of design.