Understanding Shock Sensor G-Force Ranges & Outputs

[samodelov.1]

I have no experience with shock sensors and am looking for some insight. The ones I have found say a range in g's (example - range: 10,000g) and then give a maximum value. I was wondering if the range would be similar to a minimum that the sensor is able to read or how close to the given range I would have to be in order for it to give me an output. Supposedly, it is not going to give any output if there is just some weight on the sensor (if I understand correctly), but can it be damaged if a weight (well below the maximum) is placed on it for a sustained amount of time? Thanks!

 

[MATLABdude]

I'd doubt that you can damage it by putting on a small (but reasonable) load on top of the IC, because, well, you're probably not actually causing the sensing element to actually do anything (more than likely, you're just applying this load to the device package). I'm not a devices expert, however, the designs that I've seen have micro paddles inside of a cavity that deflect whenever the device is accelerated. However, something like a strain gauge or force meter (where a load is applied and then measured) is a different beast altogether.

So, if you use this chip to prop up, say, a 5-ton hydraulic press, you'll probably smash the sensor. If you're attempting to load the sensor using, say, a sledgehammer, you'll also probably smash it, and do so before you can get anything out of the sensor. If you're just squeezing the device into your enclosure, you're probably okay.

 

[timmay]

Chances are if it's a shock sensor it's probably a piezo crystal sensor, such as this. Piezoelectric sensors generally exhibit significant signal decay at frequencies comparable to the reciprocal of their time constant, meaning that a (quasi)statically applied loading may show a brief load but will tail off to zero. As a result, any static loads applied to the sensor shouldn't give a reading, but may affect the sensor's frequency response.

If you multiply the sensitivity (e.g. in my example above, 0.5 mV/g) by the acceptable range of accelerations (e.g. [itex]\pm[/itex] 10 000 g), you'll see the range of voltages the sensor can produce, in this case [itex]\pm[/itex]5 V. Accelerations above this will cause the sensor to read full output (e.g. 20 000 g will actually be output as 10 000 g or +5 V due to saturation).

The maximum tolerable acceleration or shock (here 50 000 g) is the maximum overload acceleration the manufacture has confidence in that the sensor can handle for a very short duration before damage is sustained.

So if you have a test setup where shocks of up to 7 000 g will be measured frequently but occasionally due to operator error can produce a single shock of up 20 000 g, you should be fine - a shock outside the maximum measureable range will be clipped.

Have a read through the information on the PCB website for more information regarding sensors in general, especially piezo ones.

 

[samodelov.1]

Thanks for all the information! It has been really useful, especially the PCB website. I would be using the sensor more as a switch to trigger the next action and just need the step response, rather than an exact and repeatable voltage.

I want to make sure that I get a sensor with the correct range to give me a response but last as long as possible. I assume that there would be a decay over time of the response. Do you know how I can calculate the life of a piezoelectric sensor? If anyone knows on average how many cycles a piezoelectric sensor can still give a decent step response, that would also be helpful. It would be best though if I could calculate how low I would need to put the threshold voltage for the next action, figuring in decay over time. I'd like to use it for a couple years, if possible, with only a couple cycles a day.

Thanks!