Perhaps one or both of my postings weren't clear. 

The devices rest on a basement slab in a conditioned space. They are enclosed by a thick Styrofoam box which significantly reduces air currents. These two factors reduce the temperature variation of the instruments, although I don't doubt there could be a couple of degrees variation over the course of long periods relative to the period of the damped pendulum/filter combinations. Because that temperature varies over relatively long times, however, I judge that it is trivial after considering that the ac-coupled integrator suppresses signals longer than about 20 seconds. 

Also, the instruments use glass fibers for the torsion fiber and glass jars for the structural supports in part to reduce the effects of thermal contraction and expansion. Glass has a significantly lower thermal expansion coefficient than any of the metal pendulums used for the Lehman instruments of which I am aware, for example. Moreover, the dimension of the expansion is orthogonal to the directional sensitivity of the instrument; I hoped that this would further reduce sensitivity to thermal variation.

I recognize that this instrument is not a lab quality instrument, but for something that can be built so cheaply and easily by anyone without a lot of technical or machining skills it's a very useful device. I've detected at least 120 earthquakes from all over the world with it. 

Why discourage folks out there who don't have the skills or $ to build a Porsche when a Ford will get the job done, while spreading the fun of doing science?

Kevin McKee, burke, VA

---- psn-l-digest-request@.............. wrote: 

=============

..------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
| Message 1                                                           |
'------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
Subject: Torsion Seismometer in a jar
From:    
Date:    Thu, 15 Oct 2009 16:27:19 -0400

Hi -
I just joined this e-mail list, and noticed some of the discussion about the seismometer project I posted at http://tinkeringcaveman.googlepages.com/home

There seemed to be some concern that the system would perform poorly or be very noisy. There was some discussion over the LED being a cause for noise. I performed a number of tests before I completed the seismometer (several years ago now) to satisfy my own concerns in that regard. I used various incandescent bulbs, LEDs and sensors, and in some cases, locked the pendulum in a stationary position to test whether the sensor was picking up noise from the LED. In fact, I got a flat line even when the amplifier was maxed out--the LED was significantly quieter than other light sources. 

I've actually found the two devices that I've built (facing orthogonally) to be very sensitive.  I've put together a rough distance-magnitude threshold detection plot that I can forward to anyone who is interested. The weakest quake I've detected is a magnitude 2.0, about 0.05 degrees (about 3 miles) away.  The common phases are easily picked out of the signal, depending on the quake origin. 

The larger factor in the devices' sensitivity is that I mounted them on a basement slab under the stairs to my front porch, which means that  they detects the comings and goings of my family too (I sometimes used this fact to check up on my daughters' returns from dates).

I believe that I am detecting microseismic background with the instrument too, rather than air currents. 

Regarding the fact that the sensor is a displacement sensor vice a velocity sensor--I don't see that as a problem. As with a capacitive /radio frequency force-balance sensor, electronic filters and amplifiers can be used to translate the signal into a velocity response, if that is what is desired. In fact using the appropriate filter amplifier combination gets around any concerns like those expressed about the resolution of the optical sensing devices--electronic integration works very well for that.


Kevin McKee, Burke, VA


..------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
| Message 2                                                           |
'------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
Subject: Re: Torsion Seismometer in a jar
From:    ChrisAtUpw@.......
Date:    Thu, 15 Oct 2009 18:13:04 EDT


-------------------------------1255644784
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit

 
In a message dated 15/10/2009 21:27:46 GMT Daylight Time,  
kevin.mckee@....... writes:

There  seemed to be some concern that the system would perform poorly or be 
very  noisy. There was some discussion over the LED being a cause for 
noise. I used  various incandescent bulbs, LEDs and sensors, and in some cases, 
locked the  pendulum in a stationary position to test whether the sensor was 
picking up  noise from the LED. In fact, I got a flat line even when the 
amplifier was  maxed out--the LED was significantly quieter than other light 
sources.  



Hi Kevin,
 
    You seem to be experiencing apparatus problems.  Ordinary LEDs are much 
more noisy than superbright or IR LEDs. Both vary in  light output by about 
x5 between 0 and 100 C. The is a HUGE variation. They are  both more noisy 
than a filament bulb driven from a regulated voltage source.  Your flat line 
response was probably due to low gain. You should be able to pick  up 
microseisms all the time.
    You can build a very good detector using two large  area Si photocells 
BPW34 connected to a differential amplifier. I get a noise  level of about 
14 nano metres. You can't get adequate stability with a single  detector or 
with a LED which is not temperature compensated. Phototransistors  are 
extremely noisy compared to photodiodes. You can get very fine Nichrome wire  
which makes a good suspension. Alternatively, you can use a single glass or  
carbon filament. 
 
    Regards,
 
    Chris Chapman
 
    

-------------------------------1255644784
Content-Type: text/html; charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable







In a message dated 15/10/2009 21:27:46 GMT Daylight Time,=20
kevin.mckee@....... writes:
There=20
  seemed to be some concern that the system would perform poorly or be ver=
y=20
  noisy. There was some discussion over the LED being a cause for noise.=
 I used=20
  various incandescent bulbs, LEDs and sensors, and in some cases, locked=
 the=20
  pendulum in a stationary position to test whether the sensor was picking=
 up=20
  noise from the LED. In fact, I got a flat line even when the amplifier=
 was=20
  maxed out--the LED was significantly quieter than other light sources.=
=20
  


Hi Kevin,
 
    You seem to be experiencing apparatus problem=
s.=20
Ordinary LEDs are much more noisy than superbright or IR LEDs. Both vary=
 in=20
light output by about x5 between 0 and 100 C. The is a HUGE variation. The=
y are=20
both more noisy than a filament bulb driven from a regulated voltage sourc=
e.=20
Your flat line response was probably due to low gain. You should be able=
 to pick=20
up microseisms all the time.
    You can build a very good detector using two=
 large=20
area Si photocells BPW34 connected to a differential amplifier. I get a no=
ise=20
level of about 14 nano metres. You can't get adequate stability with a sin=
gle=20
detector or with a LED which is not temperature compensated. Phototransist=
ors=20
are extremely noisy compared to photodiodes. You can get very fine Nichrom=
e wire=20
which makes a good suspension. Alternatively, you can use a single gl=
ass or=20
carbon filament. 
 
    Regards,
 
    Chris Chapman
 
    

-------------------------------1255644784--


..------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
| Message 3                                                           |
'------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
Subject: Re: Torsion Seismometer in a jar
From:    "tchannel" 
Date:    Thu, 15 Oct 2009 17:18:47 -0600

This is a multi-part message in MIME format.

------=_NextPart_000_000A_01CA4DBB.8E711840
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Hi Chris,  the two large area Si photocells BPW34 connected to a =
differential amplifier, sound like something I would like to try.  Could =
you help me with a simple schematic of this, even a sketch?   Does the =
amp just use easy to find op amps?

Thanks, Ted
  ----- Original Message -----=20
  From: ChrisAtUpw@..........
  To: psn-l@.................
  Sent: Thursday, October 15, 2009 4:13 PM
  Subject: Re: Torsion Seismometer in a jar


  In a message dated 15/10/2009 21:27:46 GMT Daylight Time, =
kevin.mckee@....... writes:
    There seemed to be some concern that the system would perform poorly =
or be very noisy. There was some discussion over the LED being a cause =
for noise. I used various incandescent bulbs, LEDs and sensors, and in =
some cases, locked the pendulum in a stationary position to test whether =
the sensor was picking up noise from the LED. In fact, I got a flat line =
even when the amplifier was maxed out--the LED was significantly quieter =
than other light sources.=20

  Hi Kevin,

      You seem to be experiencing apparatus problems. Ordinary LEDs are =
much more noisy than superbright or IR LEDs. Both vary in light output =
by about x5 between 0 and 100 C. The is a HUGE variation. They are both =
more noisy than a filament bulb driven from a regulated voltage source. =
Your flat line response was probably due to low gain. You should be able =
to pick up microseisms all the time.
      You can build a very good detector using two large area Si =
photocells BPW34 connected to a differential amplifier. I get a noise =
level of about 14 nano metres. You can't get adequate stability with a =
single detector or with a LED which is not temperature compensated. =
Phototransistors are extremely noisy compared to photodiodes. You can =
get very fine Nichrome wire which makes a good suspension. =
Alternatively, you can use a single glass or carbon filament.=20

      Regards,

      Chris Chapman

      
------=_NextPart_000_000A_01CA4DBB.8E711840
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable








Hi Chris,  the two large area Si photocells BPW34 connected to =
a=20
differential amplifier, sound like something I would like to try.  =
Could=20
you help me with a simple schematic of this, even a sketch?   =
Does the=20
amp just use easy to find op amps?
 
Thanks, Ted

  
----- Original Message ----- 
  From:=20
  ChrisAtUpw@....... 
  
To: psn-l@.............. 
  
Sent: Thursday, October 15, =
2009 4:13=20
  PM
  
Subject: Re: Torsion =
Seismometer in a=20
  jar
  


  

  
In a message dated 15/10/2009 21:27:46 GMT Daylight Time, kevin.mckee@....... =
writes:
  There=20
    seemed to be some concern that the system would perform poorly or be =
very=20
    noisy. There was some discussion over the LED being a cause for =
noise. I=20
    used various incandescent bulbs, LEDs and sensors, and in some =
cases, locked=20
    the pendulum in a stationary position to test whether the sensor was =
picking=20
    up noise from the LED. In fact, I got a flat line even when the =
amplifier=20
    was maxed out--the LED was significantly quieter than other light =
sources.=20
    

  

  
Hi Kevin,
  
 
  
    You seem to be experiencing apparatus =
problems.=20
  Ordinary LEDs are much more noisy than superbright or IR LEDs. Both =
vary in=20
  light output by about x5 between 0 and 100 C. The is a HUGE variation. =
They=20
  are both more noisy than a filament bulb driven from a regulated =
voltage=20
  source. Your flat line response was probably due to low gain. You =
should be=20
  able to pick up microseisms all the time.
  
    You can build a very good detector using =
two=20
  large area Si photocells BPW34 connected to a differential amplifier. =
I get a=20
  noise level of about 14 nano metres. You can't get adequate stability =
with a=20
  single detector or with a LED which is not temperature compensated.=20
  Phototransistors are extremely noisy compared to photodiodes. You can =
get very=20
  fine Nichrome wire which makes a good suspension. Alternatively, =
you can=20
  use a single glass or carbon filament. 
  
 
  
    Regards,
  
 
  
    Chris Chapman
  
 
  
    

------=_NextPart_000_000A_01CA4DBB.8E711840--



..------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
| Message 4                                                           |
'------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
Subject: Re: Torsion Seismometer in a jar
From:    ChrisAtUpw@.......
Date:    Thu, 15 Oct 2009 21:49:17 EDT


-------------------------------1255657757
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit

 
In a message dated 16/10/2009, tchannel@............ writes:

Hi Chris,  the two large area Si photocells BPW34 connected to a  
differential amplifier, sound like something I would like to try.  Could  you help me 
with a simple schematic of this, even a sketch?   Does  the amp just use 
easy to find op amps?
 
Thanks, Ted



Hi Ted,
 
    Check out the notes and the circuit at  
_http://jclahr.com/science/psn/chapman/photo_detect/index.html_ 
(http://jclahr.com/science/psn/chapman/photo_detect/index.html) 
    This circuit works fine. You need the INA118 or  similar to get the 
very high CMRR common mode rejection ratio. You may need to  alter filter 
values to suit your application. The photocells drive current into  the opamps, 
so they run at 0V and there is no leakage current. I usually use  a miniature 
filament bulb driven from a stabilised voltage and a lens to  give a 
~parallel light beam. You can use an IR LED if you provide an auxiliary  Si 
photodiode and a feedback amplifier to stabilise the photo output. Otherwise  it 
drifts all over the place. I use two small razor blades for the sides of the  
shutter, painted matt black, but with clean edges. They move just clear of 
the  surface of the photocells.  I measured the noise level at about 14 nano 
 metres. There is inevitably some heat associated with the bulb. I suggest  
placing the bulb in a separate housing on top of the sensor. You under run 
the  bulb voltage supply to give ~infinite life. 
 
    Regards,
 
    Chris Chapman

-------------------------------1255657757
Content-Type: text/html; charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable







In a message dated 16/10/2009, tchannel@............ writes:

  
Hi Chris,  the two large area Si photocells BPW34 connected to=
 a=20
  differential amplifier, sound like something I would like to try. =
 Could=20
  you help me with a simple schematic of this, even a sketch?  =
 Does=20
  the amp just use easy to find op amps?
  
 
  
Thanks, Ted

Hi Ted,
 
    Check out the notes and the circuit at  =
htt=
p://jclahr.com/science/psn/chapman/photo_detect/index.html
    This circuit works fine. You need the INA118=
 or=20
similar to get the very high CMRR common mode rejection ratio. You may nee=
d to=20
alter filter values to suit your application. The photocells drive current=
 into=20
the opamps, so they run at 0V and there is no leakage current. I usually=
 use=20
a miniature filament bulb driven from a stabilised voltage and a lens=
 to=20
give a ~parallel light beam. You can use an IR LED if you provide an auxil=
iary=20
Si photodiode and a feedback amplifier to stabilise the photo output. Othe=
rwise=20
it drifts all over the place. I use two small razor blades for the sides=
 of the=20
shutter, painted matt black, but with clean edges. They move just clear of=
 the=20
surface of the photocells.  I measured the noise level at about 14 na=
no=20
metres. There is inevitably some heat associated with the bulb. I suggest=
=20
placing the bulb in a separate housing on top of the sensor. You under run=
 the=20
bulb voltage supply to give ~infinite life. 
 
    Regards,
 
    Chris Chapman

-------------------------------1255657757--


..------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
| Message 5                                                           |
'------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
Subject: Re: Torsion Seismometer in a jar
From:    "Geoffrey" 
Date:    Thu, 15 Oct 2009 19:07:15 -0700

If you used a current source to keep a steady
10ma or 20 ma to the LED would that not be
adequate stability for the light output also ?


----- Original Message ----- 
From: "tchannel" 
To: 
Sent: Thursday, October 15, 2009 4:18 PM
Subject: Re: Torsion Seismometer in a jar


Hi Chris,  the two large area Si photocells BPW34 connected to a differential amplifier, sound like something I would like to try. 
Could you help me with a simple schematic of this, even a sketch?   Does the amp just use easy to find op amps?

Thanks, Ted
  ----- Original Message ----- 
  From: ChrisAtUpw@.......
  To: psn-l@..............
  Sent: Thursday, October 15, 2009 4:13 PM
  Subject: Re: Torsion Seismometer in a jar


  In a message dated 15/10/2009 21:27:46 GMT Daylight Time, kevin.mckee@....... writes:
    There seemed to be some concern that the system would perform poorly or be very noisy. There was some discussion over the LED 
being a cause for noise. I used various incandescent bulbs, LEDs and sensors, and in some cases, locked the pendulum in a stationary 
position to test whether the sensor was picking up noise from the LED. In fact, I got a flat line even when the amplifier was maxed 
out--the LED was significantly quieter than other light sources.

  Hi Kevin,

      You seem to be experiencing apparatus problems. Ordinary LEDs are much more noisy than superbright or IR LEDs. Both vary in 
light output by about x5 between 0 and 100 C. The is a HUGE variation. They are both more noisy than a filament bulb driven from a 
regulated voltage source. Your flat line response was probably due to low gain. You should be able to pick up microseisms all the 
time.
      You can build a very good detector using two large area Si photocells BPW34 connected to a differential amplifier. I get a 
noise level of about 14 nano metres. You can't get adequate stability with a single detector or with a LED which is not temperature 
compensated. Phototransistors are extremely noisy compared to photodiodes. You can get very fine Nichrome wire which makes a good 
suspension. Alternatively, you can use a single glass or carbon filament.

      Regards,

      Chris Chapman




..------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
| Message 6                                                           |
'------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
Subject: Re: Torsion Seismometer in a jar
From:    ChrisAtUpw@.......
Date:    Thu, 15 Oct 2009 22:12:32 EDT


-------------------------------1255659152
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit

 
In a message dated 16/10/2009, gmvoeth@........... writes:

If you  used a current source to keep a steady
10ma or 20 ma to the LED would that  not be
adequate stability for the light output also  ?



Hi Geoff,
 
    No chance. The photo output at a constant current  depends on the 
junction temperature.
 
    Regards,
 
    Chris Chapman

-------------------------------1255659152
Content-Type: text/html; charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable







In a message dated 16/10/2009, gmvoeth@........... writes:
If you=20
  used a current source to keep a steady
10ma or 20 ma to the LED would=
 that=20
  not be
adequate stability for the light output also=20
?


Hi Geoff,
 
    No chance. The photo output at a constant cur=
rent=20
depends on the junction temperature.
 
    Regards,
 
    Chris Chapman

-------------------------------1255659152--


..------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
| Message 7                                                           |
'------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
Subject: Re: Torsion Seismometer in a jar
From:    =?ISO-8859-1?Q?J=F3n_Fr=EDmann?= 
Date:    Fri, 16 Oct 2009 04:04:54 +0000

Hi,

Is is possible to edit this design so it doesn't use glass jar and other
weak material ?

Regards,
J=F3n Fr=EDmann.

On fim, 2009-10-15 at 16:27 -0400, kevin.mckee@....... wrote:
> Hi -
> I just joined this e-mail list, and noticed some of the discussion about =
the seismometer project I posted at http://tinkeringcaveman.googlepages.com=
/home
>=20
> There seemed to be some concern that the system would perform poorly or b=
e very noisy. There was some discussion over the LED being a cause for nois=
e. I performed a number of tests before I completed the seismometer (severa=
l years ago now) to satisfy my own concerns in that regard. I used various =
incandescent bulbs, LEDs and sensors, and in some cases, locked the pendulu=
m in a stationary position to test whether the sensor was picking up noise =
from the LED. In fact, I got a flat line even when the amplifier was maxed =
out--the LED was significantly quieter than other light sources.=20
>=20
> I've actually found the two devices that I've built (facing orthogonally)=
 to be very sensitive.  I've put together a rough distance-magnitude thresh=
old detection plot that I can forward to anyone who is interested. The weak=
est quake I've detected is a magnitude 2.0, about 0.05 degrees (about 3 mil=
es) away.  The common phases are easily picked out of the signal, depending=
 on the quake origin.=20
>=20
> The larger factor in the devices' sensitivity is that I mounted them on a=
 basement slab under the stairs to my front porch, which means that  they d=
etects the comings and goings of my family too (I sometimes used this fact =
to check up on my daughters' returns from dates).
>=20
> I believe that I am detecting microseismic background with the instrument=
 too, rather than air currents.=20
>=20
> Regarding the fact that the sensor is a displacement sensor vice a veloci=
ty sensor--I don't see that as a problem. As with a capacitive /radio frequ=
ency force-balance sensor, electronic filters and amplifiers can be used to=
 translate the signal into a velocity response, if that is what is desired.=
 In fact using the appropriate filter amplifier combination gets around any=
 concerns like those expressed about the resolution of the optical sensing =
devices--electronic integration works very well for that.
>=20
>=20
> Kevin McKee, Burke, VA
> __________________________________________________________
>=20
> Public Seismic Network Mailing List (PSN-L)
>=20
> To leave this list email PSN-L-REQUEST@.............. with=20
> the body of the message (first line only): unsubscribe
> See http://www.seismicnet.com/maillist.html for more information.



__________________________________________________________

Public Seismic Network Mailing List (PSN-L)

To leave this list email PSN-L-DIGEST-REQUEST@.............. with 
the body of the message (first line only): unsubscribe
See http://www.seismicnet.com/maillist.html for more information.
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)