SparkFun Forums

[jenes]

Hello everyone.

I am new to accelerometers and just bought an ADXL330 board with buffers (http://www.sgbotic.com/index.php?target ... uct_id=109) - (US$41.66) for signal conditioning so that it would be directly connected to a PIC u-controller. Supply voltage is 3.3V and 0g bias is around 1.6V.
But the problem is that its sensitivity is around 0.3V/g. And even when I used the +vref and -vref (+vref=2V, -vref=1V) of the microcontroller to somehow improve its resolution, it would still be not enough to measure or sense those very small (minute) movement.

I also found about this accelerometer, the ADXL203 with a sensitivity of 1V/g (http://www.sparkfun.com/commerce/produc ... cts_id=844) - (US$39.95). What do you think of this? Could it be directly connected to a PIC's ADC without buffers?

Does anyone happen to know of a much more sensitive accelerometer, a higher V/g sensitivity? With a price range close to those I've shown. Our project really needs a higher sensitivity accelerometer to sense those minute earth movements (earthquakes of small intensities). If you might have any suggestions, please post it here.
Thanks a lot in advance.

[stevech]

sensing distant earthquakes - lots of hobbyists have done this.
I think it's normally a mass suspended on a spring (Y axis) or a fulcrum (X axis, times 2). The inertia of the mass is such that the suspension moves faster than the mass. The difference in speed is what you measure.

I've seen homemade jigs: A weight hanging on a spring, with a large cylindrical magnet beneath. The magnet is within a coil of thousands of turns of small wire. The flux lines create a current flow that you measure and calibrate (somehow) to velocity.

For horizontal movement, perhaps the same idea, but using a fulcrum or balance.

Scientific American has over the years published lots of these.

I have, and you can find on the surplus market, a geophone. This is a beer-can sized thing as above, for the Y axis. Sensitive enough to catch cars going by tens of ft. away if the geophone is securely affixed to a slab. Probably not good enough for distant earthquakes.

Because of the lack of mass, the MEMS devices probably cannot get down to, what, 10 to the minus 4 G or some such.

So start with Scientific American, I'd say, to get the first principles.

[silic0re]

The instrument you're looking for is called a 'geophone'. There used to be online electronic surplus places that would sell various types of geophones -- here's a cheap one I just googled from BG Micro:

http://www.bgmicro.com/index.asp?PageAc ... odID=10783

hope that helps

[Philba]

A seismometer was one of my favorite projects when I was growing up. The Amateur Scientist column in the back of SciAm had several good designs over the years. I used to spend hours pouring over that column. They are a lot easier to build now because you have so many cheap options for the readout and amplifier electronics are quite inexpensive to build. Google turns up lots of hits.

The problem with geophones is that you need to filter out the local and higher frequency noise. Accelerometers could work but I think they aren't sensitive enough. You are probably looking for frequencies in sub .1 hz range. The magnetic pickup coil sensor is probably the easiest for an amateur to build but it needs a lot of amplification. There may be better sensors out there.

[jenes]

Thank you so much guys for your time and effort to reply. I also searched the net for possible applications of an accelerometer for seismic sensing and I have found only one site that uses an accelerometer for seismic sensing. The Applied MEMs Inc. model SF1500A tri-axial accelerometers, http://www.inl.gov/geosciences/seismicnetwork.shtml, you could also google "SF1500A," and you could see their paper "Earthquake Data Recorded by the MEMS Accelerometer" by Austin Holland. But I really did not find any SF1500A on most common electronic stores. As of now I did not find any data sheet to view its sensitivity.

Our application is really on accelerometers. If there would be an IC like size of geophone then why not. But we've committed on using an accelerometer. We just have to bear with the fact that it cannot measure lower intensity earthquakes.

We just really have to find a very sensitive MEMS accelerometer, those with higher V/g, and could measure lesser accelerations, like lesser than 1mg. If you might happen to know of this kind of MEMS accelerometer, please post it here.

[edmoore]

I think you will be disappointed with what you will find - MEMS technology is really just too noisy and insensitive for this kind of application. There may be some (very) expensive exceptions to that rule, but I don't know where you would find them, if indeed they exist.

You may have to un-commit yourself, as it's really the wrong technology for what you want to do.

[shake]

Bump. Wondering if there's any new concensus on this topic since it's been a few years? I'd like to build a seismometer for measuring earth movements using a MEMS accelerometer. Totally new to this hobby, so would appreciate some insight. Thanks!

[jremington]

Why bother with a noisy MEMS accelerometer, when you can buy a geophone cheaply? https://www.sparkfun.com/products/11744