Circuit Wanted To Buy...

I agree with Steve... :-D:-D:-D
Although electronics is cheap in bulk, that's purely because of mass production. Nobody can afford to give you this sort of thing for the same price as it would be if you'd found what you want all ready to go in a retail shop. There's too much work in it!
Electronics design is a high-stress activity which uses all a person's brains. Good electronics is designed by clever people who think so hard for so long that they find simplicities where you just wouldn't expect them. Such people have trained for years. It's hard work. What is their time worth? Hard to say, I know my workspace cost my company $300/hour 25 years ago... what does it cost now even to set someone up with everything they need for technical work? I had at least $50,000 worth of gear at my disposal, back then. That's what a bespoke designer would need.
So! Are you a billionaire with a nice attitude? I can do such work for you if we become friends, part-time only (else I won't stay fresh and creative), for a good salary.
You will need real wherewithal to get what you want!

 

Just for fun, just how easy could this be done?

A tiny speaker with a weight attached to the cone could be used to detect vibration, the signal could be fed to a transistor that fills a capacitor that runs an op-amp to switch the main 9V. Without filling the capacitor runs down in time ~RC

Even so, developing such a project would cost big bucks as has been noted here.

 

Not bad, Shiek! Not bad at all!

 

True, my proposal was not well thought out, and far, far from perfect.

 

... like for example the generating coil might be grounded on 1 side and feed the base of a high-gain darlington pair with the other. Half-pulses would charge a 100 microfarad capacitor in the collector with 10 Megohms in series. A separate RC network would smooth Vc with respect to ground, out to the right on the drawing. The averaged Vc would be measured against a 10 meg potentiometer by a low-power comparator, whose output would pull down to switch a P-channel Mosfet in the positive rail.
If the speaker were too low power we could use the spring out of a ball-point pen, with a little rare earth magnet and a good wrapping of litz wire.
... right about now, we're off to see the customer for development costs, with progress payments to be made at defined stages of the project.
How many did you say you wanted?

 

I see it still needs lots of attention.

 

A ceramic record player stereo pick-up would do two axes, it would have a relatively high voltage suitable for matching to a fet. A weight would need to be added instead of the needle.

 

You might consider using a self charging magnetic flashlight modified so that internal capacitor drives a DMOSFET gate on. A 9 volt battery can be installed in series with the FET and the remaining leads would be your output. Maybe a TN0604N3 would work. The circuit draw should be less than a few microamps. Self powered motion detectors are hard to come by but that's what you need to drive the FET.

 

Newbie here, just came across this thread...

I know it's old, and as others have said custom electronic design is not cheap but there are people like me that will work at a reasonable rate to design custom circuits when I have time or when the project interest me...

As far as this circuit goes, it's not complicated at all in fact it's pretty straight forward and simple to build, if the original poster is still around and ready to give more input I might be interested in working out a deal...

One, thing I will comment upon is the fact that the build requirement requires no draw when idle, with todays low power sleep modes on µ-chips this is really impractical for most designs... The low power sleep modes generally draw less or equivalent to the normal idle discharge of a battery just sitting there alone, thus from a cost perspective it's simply easier and more cost effective design wise to put a chip to sleep vs designing and implementing a zero current drain wake up circuit... The current draw when active will be hugely and exponentially more than when just sitting idle and will really be the limiting factor on battery life, in sleep mode with a few AA or AAA batteries you can easily expect 6 months of idle sitting around without killing a battery, theoretically you could get years of standby.

The other big factor in this as has been touched upon is the sensitivity of the design, its dead simple to make a mechanical motion sensor but will that suit this build, or is it going to require some high end solid state device with fine tuning ability? You can build a dead simple motion detector by soldering a BB or small ball bearing onto the end of a small spring or piano wire, and inserting that into a conductive metal tube... The ball bearing on the end of the spring is held centered in the tube by the spring or wire,when jarred it makes contact with the side of the tube completing a circuit, presto motion sensed... The diameter of the tube and/or the stiffness of the spring will determine the sensitivity, by trial and error mostly. Place two or three of these at different axises and you will have a very basic but reliable multi-direction mechanical motion sensor...

You can pick one up at the dollar store and tear it apart, they generally have a small ball bearing in there there rolls inside a tube, upon contact with each end in completes a count circuit, it's only on and drawing power when the ball bearing is moving, and as stated above the sleep drain with modern µ-chips is insignificant overall. Heck the sensor out of one of them could be used in this project in the most basic form...

 

I don't know much about "raw" electronics, but that kind of sounds like it could be done with an Arduino board. Arduino is really easy to use if you're shown how from a good source and should fit in to the size if you were to build your own board. I think the alternative of having a custom built circuit would be more complicated.

Seeing as you can already speak in variables, I think you should give arduino a shot.

 

Overkill, the main concern is the development/choice of the sensor (sensing circuit) itself not the micro handling the logic, a 75¢ micro can handle the logic required for this project, no need for a Ardunio developer board...

The obstacle is designing and developing the actual sensor circuit and that is mostly independent of the micro or logic chips that will take the sensor data and do the appropriate things...

 

Here is a piezo vibration detector
https://www.sparkfun.com/products/9197

 

Will be interested to design and guide till the end. Just thinking about 2.5" x 2.5" x .5" plastic box though. If anybody wants to join me as I design, you are welcome. Let me know if you need videos of the built process in advance. Please don't ask the cost, am considering it my project as well.