"zero power" one shot 3 axis gShock tipping sensor?

[max_torque]

Anyone got any clever ideas how to make a solid state "zero power" 3 axis shock sensor that can sit in a product, and should that product experience a certain level of acceleration (different settings required in eaxh axis) irreversably trip some sort of tell tail, and that later when the unit is powered up, can be read my a microcontroller and set a "ive been dropped you fool" flag?

My current best guess is to use a piezo sensor of some sort to trip the gate of a fet, and that fet turns on, shorting a small coil cell across some sort of fuseable element that then blows, and the resistance of that element can be read with a simple digital pin on the micro at any point to see if it is still intact? (could also be a thyristor so it stays on)


Ok, it's not actually "zero power" but with a lithium primary cell it's very close and should have a long shelf life, because it only uses the cell power to trip once it's exceeded a threshold (set with a passivle votlage divider on the fet gate)

Many points awared for a purely PCB based solution btw?  I've thought about gluing some metal blocks to the piezo's and to the pcb, so a accel on the pcb pushes the blocks inertia against the piezo to provide the power to open the gate?


Any other solutions?   

[bill_c]

McMaster has just about anything, except low prices.
https://www.mcmaster.com/shock-indicating-labels/

[Terry Bites]

Make/ buy energy harvester. eg LTC3588-1 Shake before use, give it some light from a solar cell or use a Peltier module. McMaster are one shot aren't they?
Now for a warning:

[Gyro]

Any other solutions?   

Your OP description sounds as if you're trying to use a hammer to crack a nut. 

What is the problem that you're actually trying to solve? How many products is this for? Why is the digital signal important? Are you trying to live track impact as it happens (presumably not if you want 'zero power')? Are you wanting whatever it is to 'phone home' when it is powered up again?

If you want a simple visual indication (not as fancy as a report on an LCD - assuming that it isn't already providing its own severe shock indication) then a Shockwatch type product has to be the most reliable and economical solution...

https://www.shockwatch.co.uk/shockwatch-labels-clips-tubes/

They have been around for ever - I remember them being on PDP-11 era 5 / 10MB removable 14" disk platters and they still look identical!

[max_torque]

I have a product that is somewhat sensitive in nature that is pretty heavy and difficult to handle, but that is externally very robust, ie it probably won't show signs of being dropped.

I want the device to be able to know if it has experienced an impact and be able to refuse to work in that case.

Because the product has a long shelf life i don't really want to use a continuously powered sensor solution, especially as that could potentially be asleep and miss the impact.  A passive sensor does exist, they use a steel ball on a magnet, and when a certain g is exceeded, the ball jumps off the magnet and knocks some over centre contacts apart, and those contacts don't re-join when the ball jumps back onto its magnetic base after the impact has finished.  That sensor could easily be read by a simple micro controller input.

However, those sensors are

1) Large and cannot easily be PCB mounted
2) expensive for what they are
3) can potentially be reset by being disassembled and physically reset
4) do not have different tripping G in each axis, and generally have one axis where they have effectively no sensitivity


I'd love a totally pcb mounted, small, and configerable solution  :-)

[Someone]

Like a thermal fuse, but for mechanic shock.

Its going to need some clever mechanical design, you've already identified the magnetic type, there are analogous methods using spring force.

Your problems with them:

1) Large and cannot easily be PCB mounted
2) expensive for what they are
3) can potentially be reset by being disassembled and physically reset
4) do not have different tripping G in each axis, and generally have one axis where they have effectively no sensitivity


I'd love a totally pcb mounted, small, and configerable solution  :-)

Cannot be solved unless you are planning to run production into the millions. Its not a commodity part, there is no source for such devices that are simultaneously small + cheap + wide range + secure. I'm sure that it could be done on silicon with reasonable accuracy/repeatability as mems parts already have the required characteristics. But if they are easily available on the open market (to make them cheap) then you come back around and fail the "cant be reset" requirement

[max_torque]

TE sell these Piezo film sensors:

https://www.te.com/usa-en/products/sensors/piezo-film-sensors.htm



which look to be small enough to mount onto a pcb, perhaps under a soldered on tin cover for anti-tamper protection.

They look to have a big enough output to trip the gate of a fet or SCR.  I could use a 3.3v lithium battery to keep a cap charged and that could supply a slug of current when the fet turns on to blow a 50mA smd fuse ?

[Gyro]

1) Large and cannot easily be PCB mounted

Just a quick thought. Anything PCB mounted is likely to be less sensitive in the axis vertical to that PCB as the peak shock will be attenuated by PCB flex. This may or may not be a issue, but is one reason that these things are chassis mounted.

[nali]

You don't really describe your equipment, but FWIW I looked at using piezo elements for vandalism detectors in external displays. It worked but the high frequency shockwave from a metallic tap (like a ladder put against the pole or whatever) propagated just as effectively as the LF from a hefty thump. So be wary of false indicators especially if you're rigidly mounting a low mass sensor like that.

[Benta]

I've seen mechanical sensors for this.
Basically it's a spring-loaded mass (normally a steel ball) held in place by a two-piece magnet. On impact, the mass is catapulted away from the magnet and will not return due to the spring. This opens an electrical connection between the two magnet parts.
Original application was automotive fuel cut off in case of a crash, but any application is possible.