Methods to determine the minimum amount of acceleration a MEMS accelerometer can measure

[supersnare]

Hello all,

I'm currently doing research on MEMS accelerometers, frequency range dc to 1000 hz and accelerations less than 3ms^2 (0.306g), in comparisson to piezo electric ones. Since the noise floor of a MEMS sensor is inferior to that of a piezo my goal is to determine the smallest amount of acceleration that a MEMS can measure before the accelerations drown into the noisefloor of the sensor.
I'm using a PAK DAQ that handles the data comming from the piezo and MEMS under test, a signal generator that can generate various signals (sine, sweep, noise, etc) and a shaker. All sensors are attached to a large metal block on rubber feet.
My first guess is to put with noise with a low pass filter set to 1000hz and lower the amplitude untill the noise floor is reached.
Or maybe use a single sine at various frequencies and lower their amplitudes untill the noise floor get's reached.
Does anyone have experience with these kind of measurements and would like to share them with me?
Thanks for any help.

 

[GregLocock]

I'd use a sine wave at various frequencies .Measuring noise floors with white noise is a bit tricky. If you really want to use white noise, measure the coherence between the drive signal and the accelerometer output.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[IRstuff]

I'm confused, why not simply determine the minimum acceptable acceleration and find an accelerometer that does that?

For example

http://www.analog.com/media/en/technical-documentation/data-sheets/ADIS16201.pdf

would seem to meet your stated requirements of 0.3 g at 1000 Hz with about 30:1 SNR at 2+ kHz

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm