Jump to content

Recommended Posts

28 minutes ago, LittleWatchShop said:

The low-noise aspect of the design seems overstated to me.

Perhaps what we're seeing is a confusion of where exactly do we need the low noise? I think of more importance than the op amps would be the pick up. Only picking up the noise the watches making not any extra low noise found nearby. As far as the electronics how quiet is that really have to be?

For instance I'm looking at two separate older witschi machines and that both using the tlc274 found at the first link. Then The Chinese like the tlc074 found the second link. Than currently the op amp I'm using is the tlc2272 found at the third link.

https://www.ti.com/product/TLC274

https://www.ti.com/product/TLC074

https://www.ti.com/product/TLC2272

 

 

  • Like 1
Link to comment
Share on other sites

1 hour ago, LittleWatchShop said:

Valid point.  My comments here are not to take away from the great work of Vorketter et al., who have contributed to this fun thread.

The additional opamps are needed for achieving the filter function.  The low-noise aspect of the design seems overstated to me.  There is about 70dB of gain at the peak (my simulation results).  If the sensor used produced an ouput of several volts p-p, let me just say that is 1Vrms to make this easy.  That means that the input signal level is on the order of 300uV.  That is a long way from the low-noise capability of these amps.  The larger concern in terms of noise, is unwanted vibration, and 60/50Hz leakage.  The circuit I posted is running here in my lab.  When I squeak my chair (as I lean back), I see the response on the scope...and I am 3-4 ft away from the sensor.  I am running closer to 50dB gain at the peak.

I suspect that replacing all of these amps with 741's would yield a similar result with the supply voltage increased (741 prefers a larger supply).

I like my circuit, but I would like more poles and zeros to get steeper skirts on the high and low sides, but nothing is for free.

Most of the noise is from the resistors, not the op amps. The buffer stage op amp contributes nothing (measurable) to the noise seen at the preamp output. I suspect the noise is not too important if our only concern is to measure the oscillation frequency. But listening to the output may be useful (to an experienced ear) and the noise definitely makes a difference in this case. And maybe future software may tell us more about what is wrong with a watch by listening to it.

As far as how much gain we need, this depends on the microphone. As I stated previously, I hope people continue to investigate improvements in this area. Picking up sounds in the room really does not seem like a good thing.

If you think the passband shape is important, and it may be, why not consider a digital filter?

I like your comment on the 741 op amp. How much difference has decades of progress made?

  • Like 1
Link to comment
Share on other sites

1 hour ago, JohnR725 said:

I think of more importance than the op amps would be the pick up.

First of all, thanking all the work of the forum members. I have learned a lot in the last few days.
I agree with what @JohnR725 think about where we eliminate the noise. It is better to pay attention to bring a clean signal to the preamplifier. Because everything that comes from there will be amplified, signal or noise.

  • Like 1
Link to comment
Share on other sites

35 minutes ago, wlysenko said:

why not consider a digital filter?

Precisely.  I started to go down this path in an earlier post but deleted it.  I suspect that @svorkoetterimplemented some digital filtering in his software algorithm (he wrote the tool, right?).  Apart from ambient noise (60Hz, etc) overloading the early stages, you could do without filtering at the analog level and do it all digitally.  What I found with my little circuit was that without some LP filtering (I have added some since the schematic was published) the second stage can get overloaded with some random switch-mode power supply filling the ether.

Regarding the "listening" aspect.  I do not have the trained ear to ferret out some malfunction by listening.  What I see on my oscilloscope shows a lot going on.  I cannot interpret all of what I see, but I believe the visual display is a treasure trove of useful information.  While I do not get printed metrics, I really like seeing the pattern.  The software you guys are using gives the same visual display plus the useful metrics--which is very nice.

  • Like 1
Link to comment
Share on other sites

5 hours ago, guidovelasquez said:

Thank you very much for everyone's contributions. And how great that the preamplifier that @CWRNH built is already working. I wish to make a request: I would like to hear the output of the amplifier with the sound of a clock. The fellow @CWRNH said that he is using a microphone from an old timegrapher. I am very interested in listening to a recording. So in advance I would appreciate a few seconds of recording.

Im using a Vibrograph b-200A microphone. 

Here is a recording of a small watch still inside its heavy case.  Recording in noisy but not loud room.

image.thumb.png.4d1995ccfb4526119d345fa4ff6bbe3a.png

SMALL WATCH HIGH BEAT IN LARGE CASE.m4a

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

5 hours ago, guidovelasquez said:

Thank you very much for everyone's contributions. And how great that the preamplifier that @CWRNH built is already working. I wish to make a request: I would like to hear the output of the amplifier with the sound of a clock. The fellow @CWRNH said that he is using a microphone from an old timegrapher. I am very interested in listening to a recording. So in advance I would appreciate a few seconds of recording.

Here is a Recording of a 2892 ETA inside its case.

image.thumb.png.5176bf9e71301b51527bf0da61a1dffb.png

2892.m4a

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

5 hours ago, guidovelasquez said:

Thank you very much for everyone's contributions. And how great that the preamplifier that @CWRNH built is already working. I wish to make a request: I would like to hear the output of the amplifier with the sound of a clock. The fellow @CWRNH said that he is using a microphone from an old timegrapher. I am very interested in listening to a recording. So in advance I would appreciate a few seconds of recording.

And one more,  An 18 size Walthamimage.thumb.png.65bf40d2648a237aaa4c2127aadabda6.png

18s Waltham.m4a

  • Like 1
Link to comment
Share on other sites

1 minute ago, guidovelasquez said:

 

There is some noise, it gets quieter when I put the board in the metal case.  The VibroGraph microphone is far quieter than the regular Piezo mic.

Here are some snapshots with nothing on the micimage.thumb.png.05bb52599c1e80946158491820a2e63a.png

image.thumb.png.f22b47a830cd027bb25d91e1ffd19918.png

Link to comment
Share on other sites

27 minutes ago, guidovelasquez said:

 

On the top is a image from the Waltham watch, on a cheap piezo mic.  Most of the noise is probably background noise I supose.  But there is a HUGE difference.   One of my other Piezo mics is a little better but still picks up every little sound in the room.  The Vibrograph mic works far better.image.thumb.png.f2d1c28b1c77337cf0ef35dae8d31ef8.png

Next image is a recording of the same watch on the vibrograph mic.  Same settings as above, no digital filtering.

image.thumb.png.dd4fd9f4c6b9029de3149623e53c18c0.png

 

BIG DIFFERENCE.  Not sure why but the vibrograph mic works much better.

Link to comment
Share on other sites

7 minutes ago, LittleWatchShop said:

I have a vibrograph...maybe I will try it with my circuit

You should,  I took mine apart to be sure the element wasn't falling apart.   I have some other mics I am going to try.  A few are for a microset which have passive components and circuitry built in.  They are powered with 5v, I haven't figured out how they work yet.  I can see a very weak signal on my scope but can't get them to work with this amp.

Edited by CWRNH
Link to comment
Share on other sites

On 3/16/2021 at 6:47 PM, wlysenko said:

I was recovering from my second covid shot and did not participate in the recent activity on this thread.


I am sorry about CWRNH's problems with the preamp he built, which are now resolved it appears. Thanks to those who helped to solve the problems.  CWRNH's circuit is based on my circuit, which is based on svorkoetter's original design. My circuit is described in the preamp noise article  at wlysenko.blogspot.com (click on link at the bottom of the page).


The schematic in my article is for my version 15 of the preamp. The latest version (no. 16) is the same except that C3 and C5 are now 0.1 uF instead of 0.47 uF and C7 is now 0.03 uF instead of 0.1 uF. These changes improve the frequency response and lower the noise a bit.


Question for CWRNH: Why are you using an LT1115 for the buffer stage? The data sheet says this part  has the lowest voltage noise, which is good, of course, but this device is not recommended for use as a voltage follower.


LittleWatchShop questioned the need for the 680 ohm resistor in the buffer stage (the data sheet uses 600 ohm in their example). It seems the AD797 opamp, which is what I used, likes to run at a high current. At high current, the voltage noise is lower, which is what we want, at the expense of some current noise.


I hope people continue to look at improving the microphone. My microphone is nearly identical to the original svorkoetter design. I found that it picks up sounds in the room  all too well. If I listen to the output with headphones, I can hear my own breathing! I think we want something that is sensitive to watch vibrations but not to sounds in the air.

WOW, So I replaced the capacitors as you indicated and it made quite a difference.  Here is a screenshot without digital filtering.image.thumb.png.0d6c3888284525c1297c4e1d099636f1.png

image.thumb.png.d34cebceb13ae533b7111c0b132214a5.png

 

AND With filtering

image.thumb.png.01f1a4d4a892673c4788ad00bc13059d.png

 

All of this noise is gone, you can see where the averages were before.

image.thumb.png.218c88c607664356d656d0cc19737fe8.png

image.thumb.png.c103360e3505fd0f341cbbe2563cedb6.png

Here it is without anything on the micimage.thumb.png.5b804b122dcb91136e20e3c2ef97441f.png

And for comparison here it is with the Amp turned off.image.thumb.png.dbfcbc23f7bc89f2cd64c4088d05a994.png

And also attached is a recording of the watch.

 

Waltham recording after replacing Capacitors.m4a

  • Like 1
Link to comment
Share on other sites

58 minutes ago, CWRNH said:

And also attached is a recording of the watch.

Listening to the recording, I notice a lot of noise. After analyzing it with Spectra, it becomes evident that below 3kH there is a lot of noisy signal.
Review these options:
1.- Make sure that the sensor inside the pick up of the vibrograph microphone is in good condition.
2.- Use good quality shielded cable to prevent electromagnetic waves traveling in all directions in the room from contaminating. Produced by battery chargers, energy saving lamps, led bulbs, etc.
3.- Noise in itself is not a problem. If the signal exceeds the noise by about 40 dB, good results can be obtained.
4. On the e_timer screen, the signal may be clean. But it is because the program applies a VERY EFFICIENT filter. I prefer to use the WOS screen. The waveform display. AND WITHOUT APPLYING ANY FILTER IN WOS.
In the last Waltham sound, I don't understand what the adjustment was made but it improved the sound noticeably. Enough already to be analyzed by timing machine. Finally I have the idea that the cable that carries the signal from the microphone to the preamplifier needs a better shield.
And I'm sure that if the whole circuit is inside a metallic enclosure, the signal will be clean.

Noise is present, but the signal strength is much higher.

Noise is present, but the signal strength is much higher.

  • Like 1
Link to comment
Share on other sites

I can confirm Guido's results.

I ran the audio files on PCTM. 
There is much low frequency noise, much too much for this elaborate amplifier.
Waltham was loud (PW?) compared to the noise, with software filter 'on', noise disappeared totally.

The other files showed a bit poor signal/noise ratio.

Frank

2892 filter'on':

2892.thumb.png.1fb1d09af6eaf8a3abd32368b1db955c.png

SmallWatch... filter 'off':

Smallwatch.thumb.png.a958224d3cb8f813c7e8ffe9840a517e.png

Waltham before..., filter 'off':

1200169126_Walthambefore.thumb.png.dc5202bb9881f82221fb3fd482eab76c.png

Edited by praezis
added picture
  • Like 2
Link to comment
Share on other sites

11 hours ago, guidovelasquez said:

Listening to the recording, I notice a lot of noise. After analyzing it with Spectra, it becomes evident that below 3kH there is a lot of noisy signal.
Review these options:
1.- Make sure that the sensor inside the pick up of the vibrograph microphone is in good condition.
2.- Use good quality shielded cable to prevent electromagnetic waves traveling in all directions in the room from contaminating. Produced by battery chargers, energy saving lamps, led bulbs, etc.
3.- Noise in itself is not a problem. If the signal exceeds the noise by about 40 dB, good results can be obtained.
4. On the e_timer screen, the signal may be clean. But it is because the program applies a VERY EFFICIENT filter. I prefer to use the WOS screen. The waveform display. AND WITHOUT APPLYING ANY FILTER IN WOS.
In the last Waltham sound, I don't understand what the adjustment was made but it improved the sound noticeably. Enough already to be analyzed by timing machine. Finally I have the idea that the cable that carries the signal from the microphone to the preamplifier needs a better shield.
And I'm sure that if the whole circuit is inside a metallic enclosure, the signal will be clean.

Noise is present, but the signal strength is much higher.

Noise is present, but the signal strength is much higher.

The shielded cable from the mic to the amp is the factory vibrograph cable.  The cable from the output to the PC is cheap non-shielded.  Compared to my professional factory made machines this signal is far cleaner except the microset is less susceptible to background sounds from room.  The Microset uses a piezo type mic but has noise canceling inside the mic circuitry that cleans up the signal before it even gets to the preamp I believe. 

The recordings I did were taken in a fairly noisy environment, florescent lighting, close to pc, wife was running a dremel, cars were driving back and forth, daughter was listening to Netflix, fan on my laptop is VERY loud.  So the signal may be cleaner under more controlled circumstances.

"4. On the e_timer screen, the signal may be clean. But it is because the program applies a VERY EFFICIENT filter. I prefer to use the WOS screen. The waveform display. AND WITHOUT APPLYING ANY FILTER IN WOS.
In the last Waltham sound, I don't understand what the adjustment was made but it improved the sound noticeably."

I turned off all filtering in E-timer I believe.  The last recording is after I changed the c3,c5, and c7 caps.  The audio recording was unfiltered in the last recording of the Waltham. 

I have the gain on the board (R7) turned up pretty far, I may back it off a bit.

Also board is dirty, I need to lean off the residue from the flux.  I am going to omit the IC sockets in the next build, and add ground pours.  I am also cleaning up the Traces, and laying them out differently to try to lessen noise.

Link to comment
Share on other sites

1 hour ago, praezis said:

I can confirm Guido's results.

I ran the audio files on PCTM. 
There is much low frequency noise, much too much for this elaborate amplifier.
Waltham was loud (PW?) compared to the noise, with software filter 'on', noise disappeared totally.

The other files showed a bit poor signal/noise ratio.

Frank

2892 filter'on':

2892.thumb.png.1fb1d09af6eaf8a3abd32368b1db955c.png

SmallWatch... filter 'off':

Smallwatch.thumb.png.a958224d3cb8f813c7e8ffe9840a517e.png

Waltham before..., filter 'off':

1200169126_Walthambefore.thumb.png.dc5202bb9881f82221fb3fd482eab76c.png

Do you know of a less elaborate amp with better results?  Please share.

Link to comment
Share on other sites

On 3/16/2021 at 6:47 PM, wlysenko said:

I was recovering from my second covid shot and did not participate in the recent activity on this thread.


I am sorry about CWRNH's problems with the preamp he built, which are now resolved it appears. Thanks to those who helped to solve the problems.  CWRNH's circuit is based on my circuit, which is based on svorkoetter's original design. My circuit is described in the preamp noise article  at wlysenko.blogspot.com (click on link at the bottom of the page).


The schematic in my article is for my version 15 of the preamp. The latest version (no. 16) is the same except that C3 and C5 are now 0.1 uF instead of 0.47 uF and C7 is now 0.03 uF instead of 0.1 uF. These changes improve the frequency response and lower the noise a bit.


Question for CWRNH: Why are you using an LT1115 for the buffer stage? The data sheet says this part  has the lowest voltage noise, which is good, of course, but this device is not recommended for use as a voltage follower.


LittleWatchShop questioned the need for the 680 ohm resistor in the buffer stage (the data sheet uses 600 ohm in their example). It seems the AD797 opamp, which is what I used, likes to run at a high current. At high current, the voltage noise is lower, which is what we want, at the expense of some current noise.


I hope people continue to look at improving the microphone. My microphone is nearly identical to the original svorkoetter design. I found that it picks up sounds in the room  all too well. If I listen to the output with headphones, I can hear my own breathing! I think we want something that is sensitive to watch vibrations but not to sounds in the air.

I have a couple questions.

Should I have used ceramic capacitors for instead of film capacitors?  (I used the film type because they were what I had on hand, I had to combine 2 C1s and 3 C7s to get the correct values).  I read that sometimes the ceramic capacitors can act like mics so I thought the film type may be better.

Would I 9.6v power source be okay to use with this circuit?  I want to use a rechargeable higher capacity battery.

What would be the best shielded cable and connectors to use? 

Would ground planes for the grounds and v- lessen or worsen the noise?

Here is my idea for the new board so far.

Bottom copper..

image.png.f3b4e6b8d1030e91293990cd57f50a18.png

 

Top copper...

image.png.393de55de092a8f5b6735b3abaa0552a.png

image.png

Edited by CWRNH
  • Like 1
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...