if i want to detect earthquake based on acceleration, what kind of accelerometer is better?

Piezo or MEMS

  • \$\begingroup\$ If you could provide datasheets of some devices you are considering in each category, that might evoke answers... else this is rather open-ended a question. Thanks. \$\endgroup\$ Oct 21, 2012 at 11:20
  • \$\begingroup\$ Also: What do you consider to be an earthquake? \$\endgroup\$
    – ARF
    Oct 21, 2012 at 12:01
  • \$\begingroup\$ i used to think of adxl203 , but i dont know from where i became aware of piexo existance, in the mysterious world of sensors i am some kind of LOST .. :( \$\endgroup\$
    – shampoo
    Oct 22, 2012 at 13:24
  • \$\begingroup\$ i want to make an automatic signal trying to close the gas pipes valves at a destructive earthquake ... and i should be able to predict the earthquake as fast and accurate as possible.. \$\endgroup\$
    – shampoo
    Oct 22, 2012 at 13:29
  • \$\begingroup\$ Hydraulic fracturing is common Gas production method of forcing deep pockets of gas towards wellbores. It creates micro-seismic activity and hopefully below the risk of damage to boreholes and pipelines. More instrumentation is needed to ensure safety. Is your task for benign areas away from exploration for normal distribution or for areas near gas boreholes? \$\endgroup\$ Oct 22, 2012 at 15:05

2 Answers 2


You are asking the wrong question.

What you should do is first define what you need a accellerometer to do to measure earthquakes and any other criteria that are important to you, like possibly size, power consuption and price. This is what we call a specification. Once you have the spec, compare it to accellerometers of any technology and see what fits best. If it meets the spec, then you shouldn't care if it is piezo, MEMS, or anything else.

I'm not a geologist, but most likely the important criterion will be to properly react to relatively low frequency vibrations. I'm guessing that for earthquakes it needs to work well down to maybe 100 mHz or so. High frequency operation above maybe 10 Hz (I'm guessing again) is probably irrelevant, so you can use that to advantage to increase signal to noise ratio.

  • 1
    \$\begingroup\$ After reading this document I decided definitely not to answer this question :-). \$\endgroup\$
    – stevenvh
    Oct 21, 2012 at 16:27
  • \$\begingroup\$ @steven: Yeah, looks like my guess of 100 mHz to 10 Hz needs to be extended at least by a order of magnitude on each side. This is going to be beyond any simple accellerometer, and then you still have to integrate to get actual displacement. Oh well, I did say I was guessing. \$\endgroup\$ Oct 22, 2012 at 13:56

Traditional "Geophone" design is based on moving coil in a permanent magnet assembly low Q dampening factor near 1, which is a very low for a spring mass sensor but then that has the advantage of voltage proportional to acceleration at low frequencies. The breakpoint was around 1Hz and could be equalized down to 0.1 Hz but most important ground waves were below 20Hz. "As I recall 40 yrs ago" important "Rayleigh surface waves" were higher frequency but higher density core waves travelled faster in velocity (or visa versa). We used both 10" reel multi-track mag tape Ampex recorders and 24 channel strip chart galvanic recorders on light sensitive paper. (development was like doing chinese laundry in mass production)

You need an array of sensors to triangulate the wave directions. So I suspect there must be an underground movement for cheap sensors today.... (pun intended)

You also need to oversample around 200Hz to render better anti-alias filters

I have used large 3 axis piezo electric accelerometers which are no-where near as sensitive or quiet as geophones as they require very sensitive charge amplifiers (CCA) and any slight coax motion is detected a modulation of dielectric charge or acceleration noise. But rigid coax would be pretty standard now.

Think of 1g as your calibration signal when you drop it into foam and go for a sensitivity as low on the Richter scale as you need.

My guess is they use teflon coated aluminum piston with copper wire coil inside rare earth magnets to get the dampening qualities, or a thin viscous dampener, but resonance must be avoided to get a dampening factor of 1 or have electronic EQ compensation for each sensor.

With a tiny piezo sensor, you need enough mass and sit it on bedrock or solid quiet terra-firma. If all want is level 5 and up maybe an iPhone App will do it with your phone buried underground. (chortle, chortle)

side notes

love this kind of questions as it brings back memories of two summers of seismic research I did during my undergrad days for the Dept of Earth Sciences. We installed a dozen or more in a 1/4 mile string of wire with breakout points every hundred yards with the wire coiled on a large spool on our backpacks. All traffic had to be stopped in the country from 10 miles away not because of the ton of C4 explosives used to create the impulse for seismic research but because these sensors could pick up a deer walking 100yds away or the random noise from a nearby creek. When I built and tested the first CMOS portable recorders designed by local phD Physics student, I could do a couple knee-bends standing in the Earth sciences lab in the basement on the concrete floor with the "geophone" as it was called on the other side and with full gain, I could "peg" the meter side to side (AC coupled < 0.1Hz) without much effort.

... I remember our earliest tests in early 70's on an abandoned island near an old mine shaft filled up with water with the Forestry Dept's gas water pumps and their permission and a few hundred pounds of geo-gel submerged by our licensed retired miner and when we went for a cover a hundred feet away and waited for countdown with recording going on only 30 miles away, an old 3ft girth tree flew out of the mineshaft over our heads like a piece of cork...hmm.. potent stuff

... the next site was recording near the arctic circle on a road in the middle of "nowhere". Recording crew far away never feel or hear anything and we alternated from manning truck load of electronics or stopping traffic miles away. There was never any traffic except the day I was on road duty with "white hard hat" and a bus full of miners coming off shift work. The bus driver had every intention of driving over me in the middle of the road with my big hand held red stop sign... in the middle of no where. We played chicken to the end as he lurched to a stop in front of me and he stuck his head out and swore at me with a gusto. I replied quietly.. I hope you don't mind sir ... we doing some blasting .. can you wait a few minutes? (phew)he said .. sure no problem take your time.. I didn't have the heart to tell him the blasting crew was 500 miles south.

Months later after forensic Fourier filtering & sensor time phase synchronization of the digitized geo-data on an IBM360, the Geoscience post-grad students mapped all the way down to 30 miles of the earth's core in a broad stretch of land where they were looking for minerals not oil and discovered Uranium in unlikely places as we discovered the fascinating world of Seismic sensing.... I'll leave out the Black Bear and galvo calibration story....

  • 1
    \$\begingroup\$ This is in some places interesting ramblings, but if there is a answer to the question in here I missed it. \$\endgroup\$ Oct 22, 2012 at 13:54
  • \$\begingroup\$ The 1st sentence.. "Traditional "Geophone" design is based on moving coil in a permanent magnet assembly low Q dampening factor near 1 Hz ( response). .. Sensitivity depends on requirements which affects location for sensing... quiet bed-rock location. The OPs question limits the choices to sensors that have poorer S/N ratio and noise immunity. research required to determine accuracy needed... \$\endgroup\$ Oct 22, 2012 at 14:17
  • \$\begingroup\$ Although technology has improved I see fundamentals have not changed. ["Most modern seismic sensors are electro-magnetic"] (earthquakescanada.nrcan.gc.ca/info-gen/smeters-smetres/…) \$\endgroup\$ Oct 22, 2012 at 15:34
  • \$\begingroup\$ It may also be possible that Geophones use step up transformer method of moving coil to achieve the high dampening factor like a woofer being driven by a low impedance except the moving coil around a permanent magnet has fine stator windings and thus step up the voltage and ability to load dampen the low impedance source coil suspended by springs. ( just my guess on how they design it. \$\endgroup\$ Oct 22, 2012 at 15:46

Not the answer you're looking for? Browse other questions tagged or ask your own question.