Need Off the Shelf Beat Amplifier Watches or Clocks

[rfrazier]

Hi all. As documented in this thread:

https://mb.nawcc.org/threads/position-in-which-to-regulate-cheap-chinese-automatic.192922/

I'm engaged in a project to potentially regulate some cheap Chinese made Stuhrling watches. I don't wish to buy a full blown timegrapher or clock or watch analyzer. I want to use analysis programs like clock tuner, wild spectra, or tickoprint on my tablet or similar things on my PC. I need a high gain low noise beat amplifier to get the signal from the clock or watch into the tablet or PC. The tablet can pick up the beat of my Grandmother clock or wall clock with the built in mic. I can hear those across the room. It can pick up the beat from my travel alarm if I rubber band the mic from my ear buds to the clock. I can hear that clock 2 feet away. Watches are a problem though because they're so soft.

Before you mention it, I'm currently testing a Timetrax Model 60 beat amplifier.

Timetrax Model 60 Watch Clock Beat Amplifier

Timetrax Model 60 Watch Clock Beat Amplifier Diagnostic Tool w/ PICKUP Sensor | eBay

Find many great new & used options and get the best deals for Timetrax Model 60 Watch Clock Beat Amplifier Diagnostic Tool w/ PICKUP Sensor at the best online prices at eBay! Free shipping for many products!

www.ebay.com

It does pretty good on the travel alarm as an example, but does not pick up the wristwatch well. The output has a fair amount of hiss and 60 Hz hum. Check out the attached audio file with it attached to the travel alarm. I've put in a tech support request to the seller / maker. I have searched the forum for beat amplifier but haven't found anything very new. Same with Google and DuckDuck go.

Anybody have any thoughts. I could DIY something, but I'd rather just buy something that works, up to the price of the Model 60, which was $ 80 all said and done. If the maker can get the Model 60 to work with less noise, I'll probably keep that. All help is appreciated.

May your beat rate be stable and your beat error be small. Ron

 

[praezis]

Hi Ron,
making a DIY mic amplifier is more than easy. Here is a schematic from the manual of my free PC timing machine (see attached). The simple disc can be replaced by a cheap guitar clip mic.



Frank

 

[rfrazier]

praezis Thanks for the reply. I'll admit, that DOES look like a pretty simple circuit. I find it interesting that there's no power supply. Also, I looked up the transistor. It only has a gain of 45 so I'm surprised it works for this purpose. Will your circuit work for watches which have very soft sounds?

I'm in communication with the developer of the Timetrax Model 60 beat amplifier. We're troubleshooting to determine where the hiss in my amp is coming from. He seems very willing to work with his customers like me.

I think I may have eliminated the 60 cycle hum. It seems to have been coming from the tablet power adapter. Dooh! Tickoprint even warns you about that when you run the program. That's the only thing that made me think of that.

Hypothetically, how would someone get your software?

May your beat rate be stable and your beat error be small. Ron

 

[John Runciman]

Timetrax Model 60 beat amplifier

You already have a key part of what you need which is a pickup and the pickup for that should be pretty decent. At least the earlier style of them were very good so I'm assuming the current style which looks different is the same thing.

Then I see you fixed your noise problem and yes I was going to suggest the power supply because they're not necessarily always the cleanest especially when you're dealing with low audio signals that the real quiet power supplies

then for the software there's free software out there.

Here is a schematic from the manual of my free PC timing machine (see attached).

So free software don't see the link? I believe this both the free version as I think I saw it on my laptop the other day and there's a pay version which obviously probably gives enhancements of some sort.

then this because everybody needs a choice or a comparison. So the version of the link below the basic version runs for free and then for a small charge the pay version ads in a few features but initially I believe it runs in the full version so you can evaluate it

About Watch-O-Scope - Watch Timing System

@praezis Thanks for the reply. I'll admit, that DOES look like a pretty simple circuit. I find it interesting that there's no power supply. Also, I looked up the transistor. It only has a gain of 45 so I'm surprised it works for this purpose. Will your circuit work for watches which have very soft sounds?

Really what you just need is a impedance matching bow first certificate. In other words the piezo is very high impedance and the input circuitry of a computer a tablet is expecting a capacitive microphone. They're expecting something entirely different. Then as the computer input is a microphone input for the most part you probably have enough single C don't need for a much he is need the buffer things. Otherwise there's some simple op amp circuits that can be made then as I said he already have the microphone anyway just need to clip it on to something like the watch your movement holder holding the watch

 

[rfrazier]

Hi all. John Runciman Thanks for the reply. I will respond more thoroughly after supper. I just wanted to pass along a quick update. I did get rid of the 60 Hz hum from the tablet by removing the power adapter. Then, when I tried to record the attached file on Audacity on the laptop, the hum was back. Apparently it was an artifact of the PC this time. (dang it) So, since I don't believe the amp or the sensor is providing the hum, I artificially removed the 60 Hz hum from the attached sound file with Audacity. Other than that, the attached file is representative of the sound I'm getting from the sensor and amp when it's attached to a wristwatch. The problem I'm now trying to chase down is hiss, which is substantial. There's barely enough signal to do beat counting in Wild Spectra for example. More later.

May your beat rate be stable and your beat error be small. Ron

 

[rfrazier]

You already have a key part of what you need which is a pickup and the pickup for that should be pretty decent.

Yes, you're right. That pickup does seem to be very sensitive. Output all by itself is very low so I cannot just connect it to the tablet.

They're expecting something entirely different.

Another good point. The Model 60 does have a pc / phone / tablet output. Presumably it does impedance matching.

About Watch-O-Scope - Watch Timing System

Thanks for that tip. I tried it briefly. It does do the basics although it doesn't show the current beat rate. Right now I'm pretty fascinated with Wild Spectra Lite on the tablet since it allows me to tinker with levels and thresholds and fine tune the beat recognition. This watch has very soft sound. I can only hear it half an inch from my ear. I am actually able to read the watch with this system but I'd just like to eliminate the hiss. Clipping the sensor to the watch band lug seems to do pretty well.

May your beat rate be stable and your beat error be small. Ron

 

[praezis]

Ron,
the link for download is in the QR-code - and here.

Btw. this transistor's B is about 200. Power supply comes from the mic input of the PC. Just try it, you will be pleased.

Frank

 

[rfrazier]

the link for download is in the QR-code - and here.

praezis My laptop cannot scan QR codes.

So, thanks for the link. I suggest also putting an actual URL in your flyer.

Power supply comes from the mic input of the PC. Just try it, you will be pleased.

That's interesting. I hadn't thought about the PC supplying the power. Pretty cool.

May your beat rate be stable and your beat error be small. Ron

 

[John Runciman]

It does do the basics although it doesn't show the current beat rate

Can we see a picture of what it shows? + Curious about how you're connecting the watch to your set up a picture that might be helpful also

 

[John Runciman]

I artificially removed the 60 Hz hum from the attached sound file with Audacity. Other than that, the attached file is representative of the sound I'm getting from the sensor and amp when it's attached to a wristwatch. The problem I'm now trying to chase down is hiss, which is substantial. There's barely enough signal to do beat counting in Wild Spectra for example. More later.

Yes your audio file sounds interesting but not in a good way of interesting. If you're doing audio processing you can basically filter everything off below 1000hz which is what the timing machines will typically do anyway.

 

[praezis]

Ron,
just now I also listened to your audio file. What you have is no hum, it is just noise very highly amplified. The watch tics are barely higher than the noise. A filter as mentioned by John doesn't make the least difference here.
Something is generally wrong with your setup.

Your watch noise:


Noise picked up with a circuit as above:


Frank

 

[rfrazier]

HI all. I appreciate the replies. I'm going to post something quick here as I'm running to lunch then I may follow up with more later.

Can we see a picture of what it shows? + Curious about how you're connecting the watch to your set up a picture that might be helpful also

Posting a picture below. Warning you may not want to do this to an expensive watch. But, I'm clipping the alligator clip on the sensor to the watch band lug of the watch. I have no permanent workspace so I assemble and disassemble the setup when needed. The picture is literally in my lap, but you'll get the idea.

you can basically filter everything off below 1000hz which is what the timing machines will typically do anyway

That's interesting. I don't normally process the audio in audacity. But, since most or all the hum was coming from the tablet or the pc and not the amp and sensor, I edited it to give a more representative example of the output of the amp.

What you have is no hum

Something is generally wrong with your setup

There was hum before but I removed it because I didn't feel the amp and sensor was the source. You may be right. I'm starting to think something is wrong with the amp. It seems to be generating more noise than it should in relation to the ticks of the clock or watch. Also, when I touch the sensor, I hear not only the scuffing noises that I expect but also distortion that probably shouldn't be there.

Below is a picture of the basic operation of WatchOScope free. Note that I DON'T think this is an accurate representation of what my watch is doing. There's no way the watch is that far off. I think it's detecting fake ticks from the noise. With Wild Spectra Lite on the tablet, I can squeak out enough signal to read the watch but I don't think it's as good as it could be.

May your beat rate be stable and your beat error be small. Ron

 

[John Runciman]

just now I also listened to your audio file. What you have is no hum, it is just noise very highly amplified. The watch tics are barely higher than the noise. A filter as mentioned by John doesn't make the least difference here.
Something is generally wrong with your setup.

Thank you that explains what I was hearing. I was listening with my headphones and definitely didn't sound right.

Posting a picture below. Warning you may not want to do this to an expensive watch. But, I'm clipping the alligator clip on the sensor to the watch band lug of the watch. I have no permanent workspace so I assemble and disassemble the setup when needed. The picture is literally in my lap, but you'll get the idea.

I'm a bit confused here? I have the earlier versions that microphone and they work really well. I'm not using the same amplifier though I'm using something different. But I read the specifications of the amplifier he has and it's designed to do exactly what you're doing so I'm not sure why the signal sucks possibly?

Looking at your picture again I think I know what the problem is. The ticking sound of the watch is transmitted as of vibration to the sensor. The problem with a watch like this is it's a skeleton movement with a big ring that probably has a plastic movement ring underneath. In other words at the small movement in a big case and the plastic ring does not transmit sounds well at all. Try moving the sensor to the crown yes I know if it's a good crown you wouldn't do it but it's going to be the only way to get a good signal out of that case as a guess. Then yes this happens with even professional timing machines which is why the crown always faces towards the sensor if the watches case stop because that's the best way to get the signal out usually.

WatchOScope

Then this software you're using has a oscilloscope display give us a capture of that. It should be quite apparent that the software is having a problem figuring out where the signals are.

The other thing to try is I don't suppose you have a pocket watch? Pocket watches tend to be really loud and noisy In other words the perfect thing for testing microphones and software.

 

[rfrazier]

Hi all. Just wanted to post a quick note here. But it's definitely past my bed time here in the Eastern time zone. But, I am having some success learning what it takes to make this amp and sensor happy. I'm getting some usable signals even with the wristwatch and also with a travel alarm clock which is louder. I'll share my results tomorrow probably in the afternoon. Some of the tips here have been helpful so I appreciate that. More later.

May your beat rate be stable and your beat error be small. Ron

 

[rfrazier]

Hi all. OK. Finally made it back to the forum. This post is of necessity long. It's much easier to read on a PC than a phone or tablet.

I'm happy to announce that I've finally got my setup working to my satisfaction and I am able to read the wristwatch on my tablet. I'm not able to read it on my PC, but I don't think the problem is in the amplifier or the sensor. I think the PC sound system is the problem there.

I don't know if the TimeTrax Model 60 is performing to it's absolute peak potential as there is a little hiss in the signal. It seems worse with softer signals and less with stronger signals. But, I've got it doing what I need it to and I haven't been able to find anything that looks better in the market for a similar price.

I've spent much of the day compiling and editing the photos and audio tracks and typing information I'll share below. I want to let you all know how I got this working in case others may benefit from the data. I may take this later and make a how to use a software timing machine thread. Thank you all for your suggestions (and keep them coming) which have been useful. All have made me think.

Here's my connection to the watch. That's not my arm, but a chair arm. John Runciman 's suggestion to move the sensor to the crown made a difference. I learned the hard way that if the alligator clip squeezes the crown and pulls it, the watch can stop and that screws up measurements.

I think I've learned something important about the TimeTrax Model 60 amplifier. It has a volume or gain control which is a bidirectional switch. Power turns on with the gain or volume at a mid level. Holding the switch left for about 10 seconds from the mid point goes to minimum level. And holding the switch right for about 10 seconds from the mid point goes to maximum level. I think the mid or higher levels are good if you're listening to the built in speaker. But, the only way I was able to get a clean signal into my PC or tablet was to use the minimum level.

I never was able to get the program I was trying to use on the PC to work. See the audio file below of the sound I have coming into Windows from the watch, amplifier, and sensor.

input to windows from TimeTrax model 60 with wristwatch.mp3.

I feel that the problem is with the PC's sound system though and not the amplifier and sensor. Nevertheless, I'll show you the procedure I tried in case you want to try it.

Here are the microphone settings I have in Windows 7.

Here are the level settings. I have the microphone set at full level and full boost.

Here is the listen tab. By checking the check box, I can listen to the microphone signal on my headphones. You don't want to do this without headphones as you could get terrible feedback from the speakers playing the microphone input.

The programs I'm trying to use are free versions and may be a little dated. There may be newer versions and non free versions will no doubt do more. I'm running Windows 7 on the PC and Android on the tablet.

Here are the WatchOScope settings. I've changed the averaging period, vertical scale, and turned on the adjust threshold manually setting.

Here is the threshold adjustment screen. I tried to set it to get real ticks and discard false ticks. In the lower right you can see that the Windows microphone is getting a signal.

Here is WatchOScope running. It looks like it's working. Area 1 shows the current ticks being updated. The ticks jump vertically from time to time and I don't know why. Area 2 shows some kind of tick indicator that moves up and down and to the right while in operation. Area 3 shows the autodetected target beat rate, and preset lift angle, and averaging period.

I want to emphasize that NO timing machine will work if the TARGET BEAT RATE is wrong. This has been my observation. Auto beat rate detection may fail if the clock isn't close to being right or has a non standard beat rate. If the watch or clock doesn't have a target beat rate that is a multiple of 1800 or 3600 in the case of many timing machines, you may have to enter it manually.

I believe you CAN enter a manual beat rate target in WatchOScope.

I believe this number is right for this watch. HOWEVER, area 4 shows the watch losing 41 minutes / day. This is way off, and I don't know why. It's really losing about 1 minute per day. I therefore DO NOT TRUST this program on my PC. I don't know what the problem is.

I have had much better success with the tablet. So, I'll turn my attention to that. Here is a sound file from the tablet. This is hard to capture as I must run a jumper cable from the tablet to the PC. I increased the volume level of the file a bit, but that's all. The 60 cycle AC hum is coming from the PC and not the tablet. Note that this is the same setup with the watch, sensor, and amp. For whatever reason, it works way better on the tablet.

output from TimeTrax model 60 to tablet and from tablet to windows with wristwatch.mp3

I'm using Wild Spectra Lite for my main app. It gives lots of control over settings. If you use it, you need to go through the tutorial in settings, extras. There are things you'll NEVER discover by tinkering, including hidden buttons on the screen that change things.

You CAN enter a manual target beat rate in Wild Spectra.

To do low level measurements like this you must, repeat MUST, repeat *MUST* disconnect the power adapter from the tablet and run it and your amplifier on batteries. Otherwise, you're almost certain to get 60 Hz hum. You may get it anyway.

Here's a picture in Wild Spectra showing the tablet with the power supply connected. The colored bar at the bottom is called a waterfall in some audio processing apps. You might call it a water slide here since it goes sideways with time. I don't totally understand it but the colors indicate sound intensity. You can see the watch ticks. The comb like pattern at the bottom is 60 Hz hum.

Here's the version without the hum. You can see a change at the bottom. Here's what the areas of the screen do and the hidden controls. (More are mentioned in the tutorial.)

Tapping area 1 tightens the top and bottom threshold for beats.

Tapping area 2 widens the top and bottom threshold for beats.

The area between area 3 and 4 shows the beat signal. The purple arrows show detected beats. Tapping area 3 widens the (visual) spacing between the beats. Tapping area 4 compresses the (visual spacing between the beats.

The area between 5 and 6 is a substantially magnified representation of the incoming signal. Tapping area 5 expands the (visual) representation of the signal. Tapping area 6 compresses the (visual) representation of the signal.

The area between 7 and 8 is the water slide I mentioned. The area just above that is a representation of the spectrum. Tapping area 7 reduces the signal amplitude coming into the program. Tapping area 8 increases the signal amplitude coming into the program. It seems to work well if I adjust it so there is some blue between the beats.

Tapping area 9 on either side of the screen makes the numbers appear or disappear.

This shows the system before I've adjusted the beat detection threshold. Ignore ghosting in the image. That's a camera artifact. The lower circled area shows the upper and lower threshold. Note only a few beats are being detected shown by the purple arrows. The upper circle shows that the BPH rate is horribly wrong. It should be 21,600 for this watch.

This image shows the system after I've adjusted the beat detection threshold. The horizontal lines are close together and it's reading the center of the pulses. The BPH in the upper left is about where I'd expect and the drift in the upper right is about right. In the center, the 6.000 is the target beat rate per second. The blue pointer is the rate over 5 seconds I believe. The center green pointer is the rate over 20 seconds, and the red pointer is the rate over 60 seconds. I pay the most attention to that. The pie wedge width of the pointer is the variance within that time frame.

In this image I have scrunched up all the beats visually so they're very close together.

This image is a close up of the beats after scrunching. The horizontal purple lines are the upper and lower beat detection thresholds. Note that they're within the outer envelope of the signal, which varies a good bit. This configuration seems to give pretty reliable beat detection. The system controls the vertical position of this pair of lines and I cannot change that. It adapts as the signal fluctuates.

Here is a sign which pops up when you turn on the Wild Spectrta setting to display beats.

Here are 4 Wild Spectra setup screens.

The one that says Spectrum is set to Spectrum DB. The target beat rate can be set to autodetect at a resolution of .5 beats / second (multiples 1800 BPH). So it could choose 5.5 BPS, 6.0 BPS, 6.5 BPS, etc. Or it can be set to autodetect 1 beat / second (multiples of 3600 BPH). So, it could choose 5.0 BPS, 6.0 BPS, 7.0 BPS, etc. You can also choose from some standard beat rates or enter a custom number. I modified the minimum spectrogram height to make the colored bar shorter.

I modified the show larger buttons and thicker lines to make those bigger.

Now I turn my attention to Clock Tuner. I get decent results from it which seem to agree with Wild Spectra.

You CAN enter a manual target beat rate in Clock Tuner.

Here is Clock Tuner running. It doesn't give me control over pulses, but it does seem to be detecting them well. You want all real pulses to be detected and not missed. You don't want fake pulses to be detected.

Clock tuner has minimal settings.

I also tried TickOPrint. It has no settings screen that I can find. It doesn't give me control over levels, only a little green light that says it THINKS it's getting good ticks.

You CANNOT enter a custom non standard beat rate as far as I can tell. Maybe the pro version can do so. I don't know.

At first it looked like it was running fine.

Then it went insane. Therefore, I don't trust it. I doubt seriously if my watch actually and immediately went from losing 30 seconds / day to gaining 30 seconds / day.

WOW. This has been quite an adventure. I can now measure my watches with the TimeTrax Model 60 sensor and amplifier and have them work well enough for my needs. I'm keeping the amp and sensor. If I had needed to, the developer said he would exchange or refund the purchase. My travel alarm clock ticks louder than the watch and my pendulum clocks tick louder still. I don't have any doubt that I could read them with this equipment. The initial gain or volume on the amp is probably good for listening to things on its speaker, as is the high gain or volume. In my experience, the low gain or volume is best for attaching the amp to a PC or tablet or phone.

Thank you all for the effort you put in responding to my thread. I've put in a good bit of effort to document this and share it. I hope it is helpful to those who wish to do similar things. If you have questions, comments, suggestions, or experiences with such things, please share with us.

May your beat rate be stable and your beat error be small. Ron

 

[John Runciman]

setup working to my satisfaction and I am able to read the wristwatch on my tablet. I'm not able to read it on my PC, but I don't think the problem is in the amplifier or the sensor. I think the PC sound system is the problem there

I was thinking about something today and that is we forgot to mention something. Computer microphone inputs aren't always the best quality and were never designed to do what were doing. Some of them will work just perfect others will not. Sometimes if you might have a line input and I think your amplifier has a higher output or a different output for something that has a different input so in other words you don't plug into the microphone you plug into the line input if it exists that might give you better signal. Another way is to totally bypassed the computer's audio and go to a USB input device.

Was trying to find a suitable USB sound device is looking at another discussion group for they were using it to make test equipment. They were commenting it was hard to find really good ones because you can't see what's inside the package. I had that happened by two identical ones from the same dealer at different shows one year apart one was super quiet and one was super noisy and worthless but they look the same. But one of the ways around the computer is input is to get a USB sound device.

Then I'm attaching a variety of things down below. A PDF explains where the signals come from and which ones are really important and if you look at your oscilloscope you'll see that you're missing one of those. You basically still have way too much noise. An image out of the witschi manual showing it's on oscilloscope notice how the signal is really really quiet and you can very clearly see the watch ticking sound without a lot of background noise. I also attached a watch o scope Filing a notice how it looks different than yours?

Then my drawing isn't the best but I snipped out one of your images and added some additional lines. In order to measure amplitude and basically everything the keys sound is the roller jewel hitting the fork which is a super super quiet sounds. You can do timing off the really allowed sound which is the locking but you can't do amplitude because you need both the roller jewel hitting and the locking if you want to get proper amplitude.

 

[praezis]

Ron,

please forgive me that I am talking directly (as we are used to in my area).
Your audio signals have no similarity to a clear watch ticking noise. They are fuzzy, distorted and much too noisy. I had to convert them to .wav but that should not be an issue. My suspicion is faulty microphone, amplifier or connections.

Computer microphone inputs aren't always the best quality and were never designed to do what were doing.
...
But one of the ways around the computer is input is to get a USB sound device.

My experience was, no computer internal soundcard is so bad that you couldn't use it for watch testing - if you present it an adequate input signal, not just noise.

However now available low cost USB soundcards became rubbish! Amongst some being extremely noisy, they got extremely inaccurate because they no more use a quartz and will give wildly varying rates.

Frank

 

[rfrazier]

Hi all. This is a quick post before lunch. More later. Also, I'll have to take a break from the forum over the weekend while I prepare for a winter storm.

I love learning all this cool info. As a newbie with cheap watches, some might be over my head, but it's very interesting.

Just for kicks, I recorded a sound file of my grandmother clock. I can hear the sound of that all the way across the room. So, it will give a strong signal. I'm not focusing on that clock, but it's an example of something that's louder. I can only hear my wristwatch 1" from my ear, so it's a VERY soft signal.

I should point out that, when I share sounds with you all, there are numerous processing artifacts that come into play. If the sound is on the tablet, I either have to connect a jumper cable to the PC to record in audacity, or I literally set the tablet on the PC and play the sound directly into the PC's built in microphone. That may actually introduce less 60 Hz hum believe it or not. I'm sampling at 8000 Hz in Audacity (somewhat low) to save file space on the forum. Then, I convert to 128 Kbps MP3, which may introduce compression artifacts. So, what you hear is not exactly what I hear. But you definitely have a point about noisy signals.

Here's the recording of the grandmother clock. It's a LOT cleaner. I think you'll find it more PALLETable. (Yes I know that's mispelled.) There's lots more blue and black in the water slide. Interestingly, I cannot regulate this clock this way. It makes a strange double tick sound on one half of the pendulum swing. Those count as double beats and throw the calculations off. The actual correct beat rate for this clock is about 3900 BPH, which is also not a multiple of 1800 nor 3600. I'll just have to use the tweak it and watch it for a week method. Again, I'm not primarily focusing on this clock right now.

See the MP3 file below.

grandmother clock TimeTrax model 60 from tablet into windows.mp3

Gotta run for now.

May your beat rate be stable and your beat error be small. Ron

 

[praezis]

Also a quick post:
as you are running Win7, it comes with the "Audiorecorder" app as part of the OS. It makes recording and store to a file very easy. Don't need extra programs and other contorted maneuvers ;-)

Frank

 

[John Runciman]

But you definitely have a point about noisy signals

One of the problems with bad signals or bad setups is people trying to help people that are posting look what my watches doing and it really isn't becomes a problem. A lot of times people are using apps with the intradermal capacitive microphone and that doesn't always work the best. So basically on watch repair discussion groups if you say you using it It's not like we won't help you but we don't always believe whatever were seeing because it may not actually be it's going on.

Then I'd be curious of your email the manufacturer as to what they perceive the problem is? A been in business for a very very long time and as I believe I set up above I use the older version of the microphone it works really well. I just don't use their amplifier circuit so maybe that's the problem but seeing as how it specifically made to do exactly what you're trying to do and it's not doing that there's something that isn't right here that were missing.

My experience was, no computer internal soundcard is so bad that you couldn't use it for watch testing - if you present it an adequate input signal, not just noise.

However now available low cost USB soundcards became rubbish! Amongst some being extremely noisy, they got extremely inaccurate because they no more use a quartz and will give wildly varying rates.

I haven't done a lot of testing with the computers microphone input other then over the years with microphone for dictation software. Which meant that I did buy a lot of nicer sound card systems versus what was originally on the computers if anything. But I do have a little more experience with the USB for the watch software application. Which I do agree some of them are very worthless. But what I like with the USB is that I can steal the 5 V power and not have to have a battery to power the preamp.

 

[rfrazier]

Hi all. Finally got back here and am able to respond a bit more. As I said in an earlier post, I may be away for a few days dealing with this upcoming winter storm and Christmas. I've also done more tests as I'll show below. I am now again considering exchanging this amplifier and sensor as I'll describe below.

Another way is to totally bypassed the computer's audio and go to a USB input device.

Good advice. Just what I was thinking.

A PDF explains where the signals come from

Thanks! I Look forward to reading it.

You basically still have way too much noise.

Agreed. See below.

can't do amplitude because you need both the roller jewel hitting and the locking if you want to get proper amplitude.

It won't kill me if I can't do that, but it would be nice to get the full capabilities of whatever software is running.

please forgive me that I am talking directly

No problem. I am a truth seeker.

They are fuzzy, distorted and much too noisy.

I agree. See below.

Win7, it comes with the "Audiorecorder"

I think it's called Sound Recorder on my system. I've never used it but it's something to think about. Thanks.

Then I'd be curious of your email the manufacturer

I've been in communications with him. He's offered to refund or exchange the unit. He suggested that putting tinfoil or grounded tinfoil under the unit might reduce hum. We haven't been able to troubleshoot the hiss much yet. He also offered to call me personally. I haven't done that yet but I might. I just suggested to him that I could buy another unit and compare the two, then return one of them. I still want one, even if it cannot do better with an extremely weak signal like the wristwatch. I haven't found a better product in the market and it seems to work better with stronger signals.

older version of the microphone it works really well.

I don't have anything with high enough gain or the proper impedance to plug the microphone / sensor into directly. Here's a quote from one of the designer's emails:

"This is a high impedance, piezoelectric crystal input to a high
gain instrumentation amplifier"

As I said above, I've done some more testing. This discussion made me "remember" that I had some better audio equipment on another computer that I got for podcasting, which I never did. I connected the unit to another computer with a prosumer grade mixer and a prosumer grade USB audio adapter worth about $ 100 each. (Yes I know you can spend LOTS more on audio.) I adjusted the gains so the hissing I'm getting comes in at a decent volume in Audacity. Then I created the the audio file. It contains 5 seconds of hiss with the Model 60 on with no sensor attached, then 5 seconds of mostly quiet with the unit powered off, then 5 seconds with the unit on and the sensor attached, and generating hiss and 60 Hz hum. It's true that the hum is coming from my environment, but it's also true that the sensor and cable is letting it in. The only thing that changed between part 1 and 2 was that I turned the Model 60 off. The only thing that changed between part 2 and 3 is that I plugged in the sensor and turned the unit on. All the recording settings were completely the same. There was no clock or watch involved in this test.

Please see the attached sound file:

3 parts TT pwr on no sens - TT pwr off - TT pwr on w sens.mp3

We'll see what happens if I can get my hands on another amplifier and sensor.

May your beat rate be stable and your beat error be small. Ron

 

[John Runciman]

I don't have anything with high enough gain or the proper impedance to plug the microphone / sensor into directly. Here's a quote from one of the designer's emails:

"This is a high impedance, piezoelectric crystal input to a high
gain instrumentation amplifier"

Yes I'm already aware of that and yes it presents a challenge unless you have the proper circuit. I have a circuit board with an op amp goes into a filter which gets rid of everything below thousand cycles then goes into the USB sound device I have and that plugs into the computer. Plus for the USB I steal the 5 V off to run my op amp. In addition I've taken the microphone apart I know exactly what's in it.

So yes you can't just plug it into the computer microphone input it's expecting a capacitive microphone which has a little bit of circuitry between the microphone and the computer. But from everything I read the amplifier you have should work and supposedly will work with all the software which means something isn't right. We just need to figure out what the problem is.

 

[rfrazier]

But from everything I read the amplifier you have should work and supposedly will work with all the software which means something isn't right. We just need to figure out what the problem is.

Yes, I would think so. It's still possible the watch is just too weak a signal although I can get Wild Spectra and Clock Tuner on the tablet to read it, barely. But, hopefully I can get another iteration of the product and see if that makes a difference.

May your beat rate be stable and your beat error be small. Ron

 

[rfrazier]

Hi all. I just wanted to post an update. I've been in touch with the developer of the amp. It seems that the unit is primarily designed for clocks and that its ability to be used with wristwatches depends on the watch and the signal strength that comes from the sensor. I've admitted that these wristwatches I have are very quiet. Here are some quotes from our emails.

Dev: "I routinely use the product to regulate ETA 28800 wristwatches & 18000 American Pocket watches using Timegrapher on my iphone XS"

Dev: "Alligator clip on sensors are not "watch" friendly

Fine for clocks but will mar a nice watch

That said... the vise based watch pickups are even more susceptible
to noise as they are much more sensitive

But in the right environment, and on a soft foam pad, will produce
very good results"

Dev: "a clip on "clock" sensor which is not sensitive enough to provide a strong enough signal of the mechanical impulse of the escapement

The problem you are having is not noise. The physics of semiconductor electronics creates white noise at every junction. With no signal applied to the input a high gain instrumentation amplifier, the noise is amplified. Fortunately the spectral density of the noise is random and useful information can be extracted from a signal + noise using hardware and/or software techniques: thresholding, filtering, FFT spectral analysis, etc.. The higher the signal to noise ratio, a more successful and stable discrimination can be achieved. So what is needed is a stronger input signal from the sensor, especially for wristwatches."

I have stated that I want to keep the product as it is good enough for my needs. If we cannot determine from my end whether it's working as designed, I'll exchange it. He has already offered to exchange or refund it.

I also mentioned that, if he knew of a watch vise sensor that doesn't cost me as much as a timing machine, I might consider purchasing it. But, I'm only regulating a couple of watches at this time. This is an intermittent hobby. I cannot spend a whole lot more than I have.

For what it's worth, I did use my new tools to open my slow watch that was losing about 6 min / week. I managed to move the rate adjustment lever, maybe too much, and put the cover back on without destroying the watch. Wild Spectra and Clock Tuner now say it's gaining about 35 sec / day. I'll see what's really happening as I wear it this week. I discovered that there is about a 1/3 sec - 1/2 sec delay between me moving a tool under the microscope and it showing up on the internal or the HDMI screen. But, if I move the tool slow, it's not a problem. It's also hard to determine depth on the microscope screen, as in how far down my tool is close to the watch. I just kept moving the tool down very slowly until I felt it touch the adjustment lever. The numbers like 1200X and 1500X magnification on the microscope ad are rubbish. Actual magnification is from about 50X to 100X comparing the sample to the screen.

I talk about microscope features here:

Tips for Purchasing an Inexpensive Digital Microscope

Hi all. I am an intermittent participant to this forum with an interest in time, timekeeping, clocks, watches, and even things like GPS timekeeping. I do a lot of research on topics of interest. When I learn something new, I try to share that. I have the following mechanical clocks: grandmother...

mb.nawcc.org

Here's my tool kit:

BYNIIUR Watch Repair Kit - $ 15
Amazon product: B07HL3VJD3

I originally ordered this microscope:

Leanking 7 inch LCD Digital USB Microscope - $ 75
Amazon product: B08H8BKL85

I couldn't screw the gantry to the base without damaging the threads.

Then I ordered this one:

Hayve 7'' HDMI Digital Microscope,1200X Coin Microscope with IPS Screen
Amazon product: B0BMTHM3LS

That probably would have been fine except it arrived with a cracked screen.

Finally, I ordered this one, got it, and I like it very much.

Hayve 7'' Digital Microscope 1500X, HDMI LCD Microscope
B0B4638J6Z

May your beat rate be stable and your beat error be small. Ron

 

[rfrazier]

PS, I will shortly post pictures of the microscope in the microscope thread linked above.

May your beat rate be stable and your beat error be small. Ron