Hi Chris,
=A0
Thank you very much for your helpful comments!
=A0
Yes, your explanation have definetly made me see=A0that noise is a fact I h=
ave to deal with using external filtering, tuned based on real response obt=
ained from this environment, that is very noisy. An important part of this =
city is contructed on what was a lake before the spaniard conquest. It is v=
ery common here to feel on building's structure the vibrations produced by =
trucks or buses on the street.=20
=A0
Electronics is not a problem at all,=A0and=A0I advance some experimentation=
 will be needed, particulary to choose the correct useful band of the filte=
rs. The 3 axis accelerometer board block diagram=A0is a good start point re=
ference for me, but I feel (just that, because I insist I have no practical=
 criteria in this matter) the 20Hz cut off frequency is too high.=A0The vib=
ration produced by trucks or busses on=A0buildings looks to me lower in fre=
quency, but the experimentation will say..!
=A0
I will try to get the 3052 sensors and go on with my own design of the filt=
er/amps and microcontroller interface, and=A0once I have some results, will=
 be glad to share them with the forum.
=A0
Many thanks again,
=A0
Alexander

--- El jue, 23/4/09, ChrisAtUpw@.......  escribi=F3:


De: ChrisAtUpw@....... 
Asunto: Re: Seismic detector with AIS326DQ
Para: psn-l@..............
Fecha: jueves, 23 abril, 2009 5:39



In a message dated 22/04/2009, alexmirabal2000@........ writes:

I need to build a device that alarms when a seism takes place, preferably i=
n an early state, let's say, for being used in public buildings or schools.=
 It is not intended to measure and record the magnitude of the earth moveme=
nt at all, just detect once it reaches certain level in any of the 3 axis a=
nd activate a sound/visual indication.
=A0
As a first approximation, I was taking a look to this IC: AIS326DQ http://w=
ww.st.com/stonline/products/literature/ds/14956/ais326dq.pdf, but honestly =
I have no practical criteria to determine in advance=A0if it sensitive enou=
gh for my purpose.=A0 The manufacturer web page mentions it is for automoti=
ve applications, but my feeling is that the accelerations changes in such e=
nvironments are far beyond the ones I need to deal with.
Hi Alexander,
=A0
=A0=A0=A0=A0If you look at the data sheet, the minimum detection level is 1=
 milli g and you only have a 12 bit ADC. These MEMs devices do not have the=
 high sensitivity, low noise and resolution that I would normally consider =
for seismic applications. They use 3.3V supplies and low level digital outp=
uts, not 5V. There is NO low pass filter to remove noise.
=A0
=A0=A0=A0=A0You still have to take measurements in the real world, with all=
=A0the environmental background noise that this entails.=20
=A0
=A0=A0=A0=A0See the comments about noise and drift at http://psn.quake.net/=
strongmotion.html
=A0
=A0=A0=A0=A0Larry does offer a triaxial sensor board http://psn.quake.net/p=
snaccel/

In case you consider it not suitable for this application, could any of you=
, please, recommend some other chip?
=A0=A0=A0=A0Not really. However, you can make a sensitive detector using a =
44 mm OD piezo disk from www.digikey.com such as 102-1170-ND @ $1.62 each w=
ith added load weights, but you do have to produce all the electronics, fil=
ters and an ADC. This is quite a lot of work. How good are you at designing=
 / building electronics?=20
=A0
=A0=A0=A0=A0I use similar discs in flex bending mode, not as a loaded drum.=
 You can also buy high capacity disks already mounted in plastic cases, but=
 their output is nearly square law, not linear like a simple flex disk. The=
 flex disk that I use gives over 5x the voltage output of a 4.5 Hz L15B geo=
phone see http://psn.quake.net/geophone/index.html=A0and the setup can go d=
own to 10 seconds period. Note that you may want to detect P and S waves fr=
om 0.5 Hz to 5 Hz.=20
=A0
=A0=A0=A0=A0Let me know if you want any further help?

I am located in Mexico City.

=A0=A0=A0=A0Regards,
=A0
Hi Chris,
 
Thank you very much for your helpful comments!
 
Yes, your explanation have definetly made me see that noise is a =
fact I have to deal with using external filtering, tuned based on real resp=
onse obtained from this environment, that is very noisy. An important part =
of this city is contructed on what was a lake before the spaniard conquest.=
 It is very common here to feel on building's structure the vibrations prod=
uced by trucks or buses on the street. 
 
Electronics is not a problem at all, and I advance some expe=
rimentation will be needed, particulary to choose the correct useful band o=
f the filters. The 3 axis accelerometer board block diagram is a good =
start point reference for me, but I feel (just that, because I insist I hav=
e no practical criteria in this matter) the 20Hz cut off frequency is too h=
igh. The vibration produced by trucks or busses on buildings look=
s to me lower in frequency, but the experimentation will say..!
 
I will try to get the 3052 sensors and go on with my own design of the=
 filter/amps and microcontroller interface, and once I have some resul=
ts, will be glad to share them with the forum.
 
Many thanks again,
 
Alexander

--- El jue, 23/4/09, ChrisAtUpw@....... <Chr=
isAtUpw@.......> escribi=F3:


De: ChrisAtUpw@....... <ChrisAtUpw@.......>=

Asunto: Re: Seismic detector with AIS326DQ
Para: psn-l@..............
m
Fecha: jueves, 23 abril, 2009 5:39



In a message dated 22/04/2009, alexmirabal2000@........ writes:

I need to build a device that alarms when a seism takes place, prefera=
bly in an early state, let's say, for being used in public buildings or sch=
ools. It is not intended to measure and record the magnitude of the earth m=
ovement at all, just detect once it reaches certain level in any of the 3 a=
xis and activate a sound/visual indication.
 
As a first approximation, I was taking a look to this IC: AIS326DQ http://www.st.com/stonline/products/lit=
erature/ds/14956/ais326dq.pdf, but honestly I have no practical criteri=
a to determine in advance if it sensitive enough for my purpose. =
 The manufacturer web page mentions it is for automotive applications, but =
my feeling is that the accelerations changes in such environments are far b=
eyond the ones I need to deal with.
Hi Alexander,
 
    If you look at the data sheet, the minimum det=
ection level is 1 milli g and you only have a 12 bit ADC. These MEMs device=
s do not have the high sensitivity, low noise and resolution that I would n=
ormally consider for seismic applications. They use 3.3V supplies and low l=
evel digital outputs, not 5V. There is NO low pass filter to remove noise.<=
/DIV>
 
    You still have to take measurements in the rea=
l world, with all the environmental background noise that this entails=
.. 
 
    See the comments about noise and drift at http://psn.quake.net/strongmotion.html
 
    Larry does offer a triaxial sensor board http:/=
/psn.quake.net/psnaccel/

In case you consider it not suitable for this application, could any o=
f you, please, recommend some other chip?
    Not really. However, you can make a sensitive =
detector using a 44 mm OD piezo disk from www.digikey.com such as 102-1170-ND @=
 $1.62 each with added load weights, but you do have to produce all the ele=
ctronics, filters and an ADC. This is quite a lot of work. How good are you=
 at designing / building electronics? 
 
    I use similar discs in flex bending mode, not =
as a loaded drum. You can also buy high capacity disks already mounted in p=
lastic cases, but their output is nearly square law, not linear like a simp=
le flex disk. The flex disk that I use gives over 5x the voltage output of =
a 4.5 Hz L15B geophone see http://psn.quake.net/geophone/index.htm=
l and the setup can go down to 10 seconds period. Note that you ma=
y want to detect P and S waves from 0.5 Hz to 5 Hz. 
 
    Let me know if you want any further help?

I am located in Mexico City.

    Regards,
 
    Chris Chapman in UK

[ Top ] 
[ Back ] 
[ Home Page ]