Ardunio interface to measure Gravity Changes using a Gravimeter

Arduino Interface to measure % of Gravity change. The Gravimeter produces a DC voltage indicating the amount of Earth’s Gravitational change which is read via the Arudunio analog inputs and passed through to Labview for plotting. The Gravimeter kit is available via Nuts&Volts.


Is this a scale, with a known weight?

Does the sensor generate much noise, or is it filtered by the PIC, and only produce a stable measurement?

The sensor is actually 2 copper plates with conductive foam between them acting as a variable resistor with a weight sitting on top. The plate resistance changes as the gravity increases. I don't think the weight size matters too much since the sampling circuit just calculates the time in counts it takes the plate resistor to charge a capacitor. Once the base count is established, the difference in counts between the base count and the new count divided by the base count yields the percent change in gravity. The unit outputs a 12-bit DAC voltage that is centered around 2.5V for zero position, the +/- gravity percent is then added or subtracted to the centered voltage to show the change in gravity. C1 on the board smooths out any ripples, so the signal is pretty clean. I do notice spikes in the signal which I haven't quite figured out why they occur yet, but it has something to do with hot days I've noticed. I need to probably change Labview a bit to smooth out the measurement a little better, but for the most part this kit works pretty well. Having the Arduino board hooked up and talking with Labview is really cool.


Ah nice. Thanks for the explanation.

How sensitive is it? If you pick it up, does the value change? The pressure sensor at work is cool that way - picking it up off the desk changes the value. Great for quick confirmation that the serial link is working!

I read the original article in Nuts & Volts and the author cautioned that the homemade foam sensor was sensitive to temperature and barometric pressure changes. It would seem to me that these changes need to be negated to have a decent chance of detecting gravity changes which were stated as 2% of total.... +/- 1%?

Assembling two as identical as possible sensors and arranging the second at 90 degrees to the earth's centerline and preloading the reference sensor with spring tension would perhaps provide a negating signal to assist to null out the undesired effects. As the spring tension would change with temperature, a few experiments with changing ambient temps should give the data points necessary to compensate for this effect in software.