D. I. Y. Watch Timing Machine.

[robmack]

Finished constructing my Watch-o-scope based on Stefan's design and suggestions from this thread.

I substituted a TL084 in a TSSOP14 package for the suggested TL071 in a 14-pin DIP.  With SMD components, this board fits into a case only slightly larger than a 9V battery (in the background).   I am still debugging the board as it stands.

Stefan, if your design doesn't evolve much from current state and involves mostly component value changes, then this board layout should accommodate your changes.

[guidovelasquez]

It is possible that this document may be useful in determining the filtering to be applied to the microphone pre-amplifier.

 

 

Signature Analysis of Mechanical Watch by.pdf

[guidovelasquez]

12 minutes ago, robmack said:

Finished constructing my Watch-o-scope based on Stefan's design and suggestions from this thread.

I substituted a TL084 in a TSSOP14 package for the suggested TL071 in a 14-pin DIP.  With SMD components, this board fits into a case only slightly larger than a 9V battery (in the background).   I am still debugging the board as it stands.

Stefan, if your design doesn't evolve much from current state and involves mostly component value changes, then this board layout should accommodate your changes.

It has a very nice appearance size too. I would escuhar a recording of the sound of a clock. I would also like to see a video of this prototype working.

Congratulations Robmack

[Endeavor]

@robmack Looks very nice ! If it works well, please sign me up for a PCB. I can do through-hole electronics and have many components, but for SMD I only have the hot air gun and solder (Both for repair work and if my solder isn't too old by now?). Perhaps you could sell as a DIY kit? Very interesting development indeed

Edited October 18, 2016 by Endeavor

[wlysenko]

My previous attempts to lower noise in the preamp were flawed because I changed the bandwidth in doing so. The following circuit fixes this. There is an added buffer stage that uses an LT1115 op amp to isolate the microphone from the first amplification stage. Now the high-pass filter on the first stage does not depend on the capacitance of the microphone. I have it set for a frequency of 800 Hz (according to Stefan's comment of 800 Hz to 8 kHz being a good passband). I did not change the other high-pass filters to 800 Hz, but maybe I should have. The low-pass filters are all at 11 kHz, just as in the original Watch-O-Scope design.

I built the circuit on my solderless breadboard. The measured noise was 6 mV, which should be compared to the 22 mV of the original circuit. I had a lot of trouble getting a buffer that was stable. I could not get the NE5534 op amp to work for this. It insisted on oscillating, even when I added the recommended 22 pF compensating capacitor across pins 5 and 8. My present circuit is still not perfect. It oscillates when the microphone (or signal generator) is unplugged from the input.

[robmack]

Excellent work on refining the design to reduce noise wlysenko!!  I'd like to integrate your research into v2 of my board some day.

I've completed my Watch-o-scope amplifier based upon Stefan's original design plus adding additional resistance and bypass on the negative inputs to the op-amps.  I finished the case and attached the watch vice.  The result is below:

 

The output is reasonably good as shown by this window capture of Watch-o-scope analyzing the Waltham pocket watch clamped into the vice.

As I mentioned earlier, I have nine raw PCBs which I can transform into amplifiers like the above if there is interest.

[matabog]

12 hours ago, wlysenko said:

My previous attempts to lower noise in the preamp were flawed because I changed the bandwidth in doing so. The following circuit fixes this. There is an added buffer stage that uses an LT1115 op amp to isolate the microphone from the first amplification stage. Now the high-pass filter on the first stage does not depend on the capacitance of the microphone. I have it set for a frequency of 800 Hz (according to Stefan's comment of 800 Hz to 8 kHz being a good passband). I did not change the other high-pass filters to 800 Hz, but maybe I should have. The low-pass filters are all at 11 kHz, just as in the original Watch-O-Scope design.

I built the circuit on my solderless breadboard. The measured noise was 6 mV, which should be compared to the 22 mV of the original circuit. I had a lot of trouble getting a buffer that was stable. I could not get the NE5534 op amp to work for this. It insisted on oscillating, even when I added the recommended 22 pF compensating capacitor across pins 5 and 8. My present circuit is still not perfect. It oscillates when the microphone (or signal generator) is unplugged from the input.

 

 

 

please see 

 

I posted a schema with a FET for a buffer. I also cut at 4kHz. I used NE5532 (dual amp).

I use the amp in very noisy environments and it works. 

 

[pipi]

hello
my quick and dirty amp doesn't work as expected. well it works but has limited gain somehow. reads finger taps on piezo but cannot hear even the loudest watches i got. so much for a crude prototype without sleep. there are no shorts, bat is at 9v and turned pot all the way up and down
can you guys pls check my build and suggest something before i start over and lose or break parts

thanks in advance

 

[svorkoetter]

I've taken your two photos, corrected for perspective, flipped one, and then traced the connections, to come up with this:

I don't have a lot of time to look at it right now, but the first thing I see is that you've got pin 6 of the first stage coupled to pin 9 of the second stage via C3 and R3. It should be pin 7 to pin 9. Anyway, with the above in hand, you can probably easily compare against the schematic to find any other errors. It seems like you might also be missing the ground connection on R7.

[pipi]

thank you!
i was just about to start over made some easier pics for that as well here i share

 

[pipi]

progress now my amp works with high gain for 5 seconds then fades away lol

i really wanna break this and start over but i dont have spare parts so ill bug you guys again

sorry

 

[pipi]

yay found the error it works now

thanks guys

[svorkoetter]

6 minutes ago, pipi said:

yay found the error it works now

thanks guys

What did the problem(s) turn out to be?

Just a piece of advice: When I design and build a circuit, I start with the schematic of course, and then I design the circuit layout, with a printed copy of the schematic by my side. I use a highlighter to trace each connection on the schematic when I make the corresponding connection on the layout. Then, when I build the circuit, I have a printed copy of the layout by my side, where I highlight each connection when I make that connection on the real board.

Finally, and most importantly, I go over the completed circuit board one more time, with a copy of the schematic by my side, highlighting each connection as I inspect the corresponding connection on the board. This double-checks both errors that occurred when designing the layout from the schematic, and errors when constructing the board from the layout.

The step in bold above can be done by anyone building a circuit, even if they are not the designer. I have caught and correct many construction errors this way. It always takes less time than giving up and starting over, and also tends to save money spent on blown parts.

[pipi]

i was gonna do the same but could not access a printer easily lol

error in red forgot a trace

 

[pipi]

ok now time to use proper shielded cables adding sockets and metal casing after cleaning flux and cutting the board to size. then the mic jig. shall i ground the casing or just isolate it i dont want the cover acting as antenna. also like many of you my desk pc injects insane amounts of noise to amp sadly. i had a cheap usb sound card with mic in ill try with that and report back if i can find it.

[robmack]

5 hours ago, pipi said:

ok now time to use proper shielded cables adding sockets and metal casing after cleaning flux and cutting the board to size. then the mic jig. shall i ground the casing or just isolate it i dont want the cover acting as antenna. 

In my implementation, I have a single ground point, which is at the same potential as the -ve terminal of the battery.  The microphone is unshielded but the exposed length of cable is extremely short before it enters the metal case so there is very little EMF pickup.  The metal case is connected through a screw and a metal standoff to the single ground point. The output cable is 2-lead STP cable.  The shield of the cable is connected to the single ground point only at one end.  At the other end it abruptly terminates within the 3.5mm plug and is not connected to anything.  One of the wire pairs is connected to Tip and the other to Sleeve on the 3.5mm audio plug.

[pipi]

done with amp

good thing i dont smoke for years found a zippo case. made without switch first; used both stereo jacks switches in series then it felt dull decided it needs a switch to push with ur finger for more fun hehe.added a dc jack as well bypassing batt when inserted. turned out nice. whole case is grounded cables are very short twisted pairs will add a led later on when i got time.

thanks guys

 

 

[guidovelasquez]

Congratulations. Good job. I would like you to share a sample of the sound recording in MP3 or WAV format. A few seconds. Also in screen capture "scope" mode. To display the signal / noise ratio.

I am making some corrections in my microphone but I still regard being noise network of 60 Hz.

Here I have this prototype. I used a microphone for guitar.

Here we can see what is inside it.

 

I made some modifications. How to change the cable for one with better shield. Recycle one moden only as support. There sticking with super-glue the microphone.

Exposing the sensor surface with the help of a stainless steel ring.

And that these are the top 5 positions in which the microphone works.

Here a screenshot of the microphone signal.

I used filtering digital Wide  (160Hz- 16kHz.

 

The Gear:

 

 

Edited November 2, 2016 by guidovelasquez
Multiple imagen error.

[SSTEEL]

Amazing job my friend, thats pretty small too, so you could some how fix it to an axis so you can test a watch movement in six positions too.

 

EDIT > I see you already have the positions figured out, well done  

 

Just noticed my Patreon badge isn't displaying, I had a problem recently with my PayPal, so hope that will be sorted asap.

Edited November 2, 2016 by SSTEEL

[pipi]

nice

i will make a mic and stand solution soon when everything is ready i will share my results

i work on it when i can just for fun since im a home amateur project goes slow no rush for me

for now i do my tests with 27mm piezo soldered to an old crocodile clip

[guidovelasquez]

Perfect. Let's wait for the results. Thank you for sharing your experiences.

[jguitron]

Hello,

I want to share my experience so far. I was blown away by the awesome work put into WOS and wanted to give it a shot.

This may be valuable to those of you that, like me, have NO experience in electronic work at all, so that putting together a mic and an amp seems insurmountable despite the great tutorial on the WOS website.

 

I got the recommended mic from amazon:

https://www.amazon.com/Korg-CM-200-WHBK-Clip-On-Microphone-White/dp/B00CPG5Z3Y/ref=sr_1_3?ie=UTF8&qid=1479917818&sr=8-3&keywords=korg+contact+mic

 

 

I found an amp for the PC/laptopt that appeared to be very similar to the one constructed on the website:

https://www.amazon.com/SmartLife-Compact-Channel-Microphone-Amplifier/dp/B01J0W1YG2/ref=sr_1_1?ie=UTF8&qid=1479917874&sr=8-1&keywords=pc+mic+amplifier

 

...but I didn't get it from amazon due to extra long shipping time and price. I found the exact same one here:

http://www.gearbest.com/speakers/pp_152359.html?wid=21

 

Significantly cheaper and I received it within the week! I was impressed with that since it came from China.

 

Here a few pics of the device:

The back of it shows the AC power plug which is 5V instead of the 9V used for the amp on the WOS site. This may have implications on performance as I'll point out later on... (Reverb apparently used to sing karaoke into the PC) 

 

I have a laptop with a combo plug with mic and headphones together, so I also got the recommended adapter to separate mic and earphones. It's the white cable plugged into the computer. The USB is the power source of the amp. 

 

 

So, once all was connected I fired up WOS to set things up.

My first concern was the volume. Despite the amp being turned all the way to the max and having set up the pic on windows to 100% in addition to a 12db boost, the oscilloscope was weak.

That's the most I could get out of the system. I see the automatic threshold line there and my first question is of the quality of the reading is affected by such weak signal, or it's the same as long as WOS recognizes it....

 

 I went ahead and CALIBRATED my soundboard with a quartz watch and I had to add 1.0s as compensation.

So I analysed a relatively new ETA 6897:

 

I decided not to regulate it until I got some more input from more experienced users.

 

My main questions are regarding the level of amplification. I was expecting much more from the amp I got and instead it's tuned all the way up. I also chose a filter at WOS because the manual says it add a boost to the signal as well, although I didn't notice any difference before and after on the sound waves.

Should I look on for another amp?

Is the quality of the reading compromised by the low levels?

Are there any suggestions as to what to improve on?

 

I appreciate any input. I'm thrilled to be able to have the traces above without having to shell out a few grands for a Wischi.

 

Thanks again!

JG

[szbalogh]

15 minutes ago, jguitron said:

Are there any suggestions as to what to improve on?

Try mount the watch with the stem pointing to the microphone first.

There is also a software for frequency filtration. Basicly makes the soundcard to a preamp. Check Your soundcard settings! Increase input volume or switch boost function on if available.

 

Edited November 23, 2016 by szbalogh

[Watchtime]

wow, beyond my hat...amazing skills shown by some members here, I love it.

[svorkoetter]

I think that the problem you're seeing is that the automatic level determination can't distinguish the parts of the ticks, and probably adjusting manually won't help either, because the signal magnitude is not much more than the noise. What is the gain of that amplifier? It must be quite low.