I recently acquired and began to explore the sensor capabilities of the Dro=
id-x, manufactured by Motorola.  This smartphone has some impressive "apps"=
 (software applications).  To date, my most useful app is the one called 'S=
ensor Dump'.  It records output from any one of the several remarkable MEMS=
 sensors onboard the device, in the form of a comma separated variable (csv=
) file.  The csv file thus generated can be directly imported into Excel (u=
sing a USB cable between phone and computer), where any number of analysis =
routines can then be used.  Ones that I routinely employ include analyses i=
n both the time domain and frequency domain.  Some examples have been descr=
ibed in a pair of arxiv papers that I wrote, titled "Physics education usin=
g a smartphone accelerometer" (online at http://arxiv.org/abs/1012.3416) an=
d "Smartphone sensor of pendulum motion", online at http://arxiv.org/ftp/ar=
xiv/papers/1012/1012.1800.pdf.  Though the phone could be used as the bob o=
f a pendulum to form a crude seismometer, its sensitivity is limited to loc=
al earthquake detection, largely because the ADC of the MEMS accelerometer =
is only 12 bits, and the full-scale deflection is 2 g's.  Thus like Larry's=
 comments concerning Analog Devices' air bag sensor in years past, you're n=
ot going to pick up teleseismic events with it.  On the other hand, one can=
 make a (highly nonlinear) motion detector with higher sensitivity by the f=
ollowing means.  With a wooden (or other) dowel, fix (such as by tape) a ba=
ll point pen on one end.  If the barrel-end is ferrous, this pen will be at=
tracted to a rare earth magnet, serving as a high-Q pendulum suspension.  A=
nother magnet on the other end of the dowel can be used (as the pendulum sw=
ings) to measure output from the magnetic field sensor that is three-axis, =
using Hall effect sensors.  The sensitivity of each sensor is about one mic=
ro-Tesla; and so the system is quite sensitive to small displacements of th=
e pendulum when the phone is placed (by trial and error) close to an optima=
l position for observing disturbances. Somebody could no doubt write a Java=
 app to sound an alarm when the motion exceeds a defined threshold.  Anyone=
 in California or other earthquake active regions interested in doing so?  =
Or maybe use the same approach to build an intrusion alarm?
 
I recently acquired and began to explore the sens=
or capabilities of the Droid-x, manufactured by Motorola.  This smartp=
hone has some impressive "apps" (software applications).  To=
 date, my most useful app is the one called 'Sensor Dump'.  It records=
 output from any one of the several remarkable MEMS sensors onboard the dev=
ice, in the form of a comma separated variable (csv) file.  The csv fi=
le thus generated can be directly imported into Excel (using a USB cable be=
tween phone and computer), where any number of analysis routines can then b=
e used.  Ones that I routinely employ include analyses in both the tim=
e domain and frequency domain.  Some examples have been described in a=
 pair of arxiv papers that I wrote, titled "Physics education using a =
smartphone accelerometer" (online at http://arxiv.org/abs/1012.3416) a=
nd "Smartphone sensor of pendulum motion", online at http://arxiv=
..org/ftp/arxiv/papers/1012/1012.1800.pdf.  Though the phone could be u=
sed as the bob of a pendulum to form a crude seismometer, its sensitivity i=
s limited to local earthquake detection, largely because the ADC of the MEM=
S accelerometer is only 12 bits, and the full-scale deflection is 2 g's.&nb=
sp; Thus like Larry's comments concerning Analog Devices' air bag sensor in=
 years past, you're not going to pick up teleseismic events with it.  =
On the other hand, one can make a (highly nonlinear) motion detector with h=
igher sensitivity by the following means.  With a wooden (or other) do=
wel, fix (such as by tape) a ball point pen on one end.  If the barrel=
-end is ferrous, this pen will be attracted to a rare earth magnet, serving=
 as a high-Q pendulum suspension.  Another magnet on the other end of =
the dowel can be used (as the pendulum swings) to measure output from the m=
agnetic field sensor that is three-axis, using Hall effect sensors.  T=
he sensitivity of each sensor is about one micro-Tesla; and so the system i=
s quite sensitive to small displacements of the pendulum when the phone is =
placed (by trial and error) close to an optimal position for observing dist=
urbances. Somebody could no doubt write a Java app to sound an alarm when t=
he motion exceeds a defined threshold.  Anyone in California or other =
earthquake active regions interested in doing so?  Or maybe use the sa=
me approach to build an intrusion alarm?
=