Accelerometer with 0.2mG resolution - seismography??

I've just been gifted some old equipment. I grabbed it mostly for the very nice weather resistant metal case it was in.

On taking it apart, I find that it has a sensor, an ADXL105, and whilst it's almost 15 years old, the technology seems similar to what is used in these things today.

I wonder if the 0.2mG resolution makes it at all useful for a sensor in a seismograph?

It has an analogue output and the (analog) sensor output is 250mV per G.

I guess the output is noise limited and they quote a noise figure of 225uV/sqrt(Hz) but considering that you're going to be looking for seismic signals at the very low frequency end of the spectrum, 1/f noise is going to be far more significant isn't it?

That noise figure suggests for a DC to 10Hz bandwidth, the noise figure (0.7mV) will be at about 2.8mG. That corresponds to being tilted by 0.2 of a degree (which is significant because the device is (was) also touted as an inclinometer).

How do I calculate what the 1/f noise will be?

Some further research suggests that an earthquake is noticeable to humans at about 10 to 20 mG, so (at best) 2 orders of magnitude better than this is probably not going to be really useful.

 

Not so much better. My calcs for noise on the new one were 2.8mg, this one says 2.6mg -- the difference will be in the rounding.