setting up a shock test.

[Simon]

I need to carry out a test that involves something being dropped and an accelerometer measuring the impact and recording it.

The test specification seems to have been created by an industry body rather than a specification body so feels a bit like they are not sure how to explain it.

The system requires 1kHz of bandwidth, 10kS/s and the results needs to be low pass filtered with a -3dB filter @30Hz. There are some filter ripple requirements, at this point given that a -3dB filter is a simple RC filter I assume the ripple stuff is for digital filtering of the data.

So I need an accelerometer with at least 1kHz of bandwidth and a low pass filter of 3dB. Given the brevity of the test I was considering using an oscilloscope to capture the waveform. I have a picoscope with 250MS/s so that is 10kS/s satisfied but I would put the signal through a physical RC filter so that sampling rate comes after the filter. But as the result is to be capped at 30Hz and the scope has 250MS/s it's the same thing?

If my sensor has a higher than 1kHz bandwidth should I filter it to 1kHz ? or maybe 10kHz, the bit that I don't understand is that if this is all being filtered to 30Hz then why all the other numbers? Nothing faster than a 30Hz sine wave will pass that filter and unless a square wave were less than 3Hz it would not be accurately shown.

[jonpaul]

Dear Simon: Great question....shock impulses need wider BW than 1 kHz.

We used to evaluate instrumant packing vs UPS/USPS handling and roughness...
 

Box up the device, kick down a flight of concrete stairs 3 times.

Open up packaging to see damages.....

Jon

[Simon]

Well this is basically to evaluate the effects on a person. I am trying to understand if what is being asked for is doable or if there are just too many numbers in there.

If the question is poorly framed then I need to come up with a way of answering it so that either they think it is what they want, or so that I can explain how it's the same thing.

[Simon]

If I put a 33.333ms pulse through a 30Hz filter (33.333ms = 1/30Hz) then and only then does the output get to 100% of the input. If I put a 1kHz sine wave through a 30Hz filter it comes out attenuated to 2.6% of the input.

So if at worse case scenario of 1kHz @ 100% I only get 2.6% of the signal out of my filter the 10kHz sampling is for what?

[Overspeed]

Hello

Accelerometer shall be first selected by technology as piezo + charge amplifier or piezo ICP or strain gage

A very common accelerometer use in shock test is the PCB piezotronic Model 350B21 able to reach 30 kHz
a shock is not really a simple pulse

mounting affect also a lot accelerometer measurement

Regards
OS