PSN-L Message - Subject: Re: $49 USB seismometer for BOINC based quake-catcher network?

In a message dated 05/09/2009, dsaum@............ writes: Has anyone participated in the quake-catcher network? I have ordered one of the USB sensors to try the system out. Hi Dave, Should you do some basic seismic sums and checks, first? Most of the Macs and Ipods seem to have +/-2g sensors with 8 bit ADCs which can give a maximum sensitivity of about 15.6 milli g per count. To get an actual seismic trace, you need say 5 counts ~ 0.08 g. The JoyWarrior JW24F8 triaxial accelerometer board with a +/- 2g range initially looks more attractive, since it has a built in ADC and an interface chip. However, the ADC still only has 10 bits, giving a sensitivity of about 3.9 milli g per count - about -48 dB. To get a recognisable seismic trace, you might need ~5 counts, say 20 milli g. This roughly represents an M4 quake at 10 km, or an M6 Quake at 80 km. The practical detection range seems to be QUITE LIMITED for most locations outside California and you may not even get a trace from a local M3 quake. See the graphs at _http://earthquake.usgs.gov/eqcenter/waveforms/netq/FAQ.php#howbig_ (http://earthquake.usgs.gov/eqcenter/waveforms/netq/FAQ.php#howbig) Observers start to 'feel' quakes over about 15 milli g in quiet locations. These detection sensitivities are so low that unless you live quite close to an active fault line, or over a seismic 'hot spot', your chances of recording a quake during a whole year seem to be minimal. You might first check if there have been ANY quakes at distances of less that ~50 km from your home site in the last few years and their magnitudes, before spending any money? See _http://earthquake.usgs.gov/_ (http://earthquake.usgs.gov/) _www.iris.edu/seismon_ (http://www.iris.edu/seismon) and _http://jclahr.com/science/psn/magnitude.html_ (http://jclahr.com/science/psn/magnitude.html) Commercial MEMs seismic accelerometers seem to have a noise level below -93 dB, but the older mechanical accelerometers can go down to between -140 and -160 dB. These sensors need to be fixed to the ground to resist accelerations of up to several g. The detection range could be greatly improved by using an existing MEMs sensor with a noise level of about 50 micro g per root Hz and a 16 bit Sigma Delta ADC. Up until now, the QC project seems to have concentrated on just increasing the number of types of low sensitivity devices that can be used. You might ask Dr. Lawrence or Dr Cochran at Quake-Catcher _http://qcn.stanford.edu/_ (http://qcn.stanford.edu/) to state at what ranges earthquake traces can be recorded with their various systems? Check whether any claims relate to just a single count, or to a recognisable seismic trace? Maybe also ask when QC are going to provide a sensor with 16 bits resolution and a noise level roughly comparable to a MEMs seismic accelerometer? The ICs are available. How about building your own accelerometer with a Piezo Sounder disk, a mass and a FET opamp? Digikey 102-1170-ND Regards, Chris Chapman

In a message dated 05/09/2009, dsaum@............ writes:

Has anyone participated in the quake-catcher network?

I have ordered one of the USB sensors to try the system out.

Hi Dave,

Should you do some basic seismic sums and che= cks,=20 first?

Most of the Macs and Ipods seem to have +/-2g= =20 sensors with 8 bit ADCs which can give a maximum sensitivity of about 15.6= milli=20 g per count. To get an actual seismic trace, you need say 5 counts ~ 0.08= =20 g.

The JoyWarrior JW24F8 triaxial accelerometer board= with=20 a +/- 2g range initially looks more attractive, since it has a built= in ADC=20 and an interface chip. However, the ADC still only has 10=20 bits, giving a sensitivity of about 3.9 milli g per count - about -48= =20 dB.

To get a recognisable seismic trace, you migh= t=20 need ~5 counts, say 20 milli g. This roughly represents an M4 qu= ake at=20 10 km, or an M6 Quake at 80 km. The practical detection range= seems=20 to be QUITE LIMITED for most locations outside California and you may not= even=20 get a trace from a local M3 quake.

See the graphs at http://earthquake.usgs.gov/eqcenter/waveforms/netq/FAQ.php#howbig

Observers start to 'feel' quakes over ab= out 15=20 milli g in quiet locations.

These detection sensitivities are so low that= =20 unless you live quite close to an active fault line, or over a seismic 'ho= t=20 spot', your chances of recording a quake during a whole year see= m to=20 be minimal. You might first check if there have been ANY qu= akes=20 at distances of less that ~50 km from your home site in the last few= years=20 and their magnitudes, before spending any money? See&n= bsp;http://earthquake.usgs.gov/ = www.iris.edu/seismon and http://jclahr.com/sc= ience/psn/magnitude.html

Commercial MEMs seismic accelerometers seem= to have=20 a noise level below -93 dB, but the older mechanical accelerometers= can go=20 down to between -140 and -160 dB. These sensors need to be fixed= to=20 the ground to resist accelerations of up to several g.

The detection range could be greatly imp= roved=20 by using an existing MEMs sensor with a noise level of about 50 micro g pe= r root=20 Hz and a 16 bit Sigma Delta ADC. Up until now, the QC project seems to hav= e=20 concentrated on just increasing the number of types of low sensitivit= y=20 devices that can be used.

You might ask Dr. Lawrence or Dr Cochran at= =20 Quake-Catcher http://qcn.stanford.edu/ to=20 state at what ranges earthquake traces can be recorded with thei= r=20 various systems? Check whether any claims relate to just a single cou= nt, or=20 to a recognisable seismic trace?

Maybe also ask when QC are going to provide= a=20 sensor with 16 bits resolution and a noise level roughly comparable to a= MEMs=20 seismic accelerometer? The ICs are available.

How about building your own accelerometer wit= h a=20 Piezo Sounder disk, a mass and a FET opamp? Digikey 102-1170-ND

Regards,

Chris=20 Chapman

PSN-L Email List Message