Actually "I" don't need the help and instead was trying to help you out!
Anyway, like I stated earlier, I experimented with a few vibration sensors including piezo speakers that detected vibration (it's resistance varied),
mechanical units that have weighted sensors that would 'bounce' on a pointed contact closure, and 'circular' sensors that seemed to use a ball in the center that would 'sway' toward the case edge to produce a contact closure.
All of these different sensors seemed to behave differently to a variety of testing 'stimuli' i.e. sharp taps, metal on metal (vs say a finger tap). I also noticed that surfaces and bonding of the sensors played a big part on sensitivity. The mechanical ones with the weights seemed to also behave differently depending on the orientation they were mounted in as well.
All except that last one I mentioned have some sort of sensitivity adjustment, but I believe what you seek is more of a 'signature' analysis of the actual vibration waveform to help detect false alarms.
I could not find anything that offered such an analysis. Depending on the accuracy of the actual vibration detection scenario you seek you would most likely in my opinion have better success with the piezo speaker method with excellent bonding to the surface being monitored. The only problem is having to analyze the 'signature' of the produced waveform real time within a reasonable cost..
There is a lot of information on various circuitry using the piezo speaker in this application that I found on various websites. The one I had success with is shown in
THIS gallery. This project was created to detect the vibration of a
DEXCOM diabetes remote monitor. Basically this unit constantly monitors the blood sugar level real time from a sensor a diabetic would wear on their arm. The unit will vibrate on a first preliminary warning, then beep and vibrate again on a 'level 2' warning.
I find it amazing that this is the only type of warning the remote will make, and a diabetic can easily sleep through them. This unit will detect this vibration and produce a contact closure (which was then rigged up to the Elk of the client so lights can turn on and whole house announcements accompanied with the warning).
The device can be set to such a low sensitivity that I can keep my finger pressed on the center of the mount and it will keep going off as it's detecting the blood pulse!
Unfortunately no signature detection is used as it's just a voltage comparator that is triggered when the voltage goes above a set threshold (i.e. the variable resistor sensitivity setting).
As you can see in those other pics I tried a mechanical vibration detection as well. The only problem was I could not get the sensitivity low enough to reliably detect the Dexcom unit's vibrations (as they only last for a second).
I don't have any pics of the circular detector I experimented with, but
HERE is a link to it.
Anyway it was an interesting project and I learned a lot about vibration detection.
I did find some sensors that would be used to detect vibration changes from say, machinery, but those were
VERY expensive! The reason these types may interest you is because they are used to detect vibration pattern changes in the machinery monitored (first clue something is going wrong with the machine).
Good luck with your vibration detection project! If you get any cool results, feel free to post them here as well! B)
