Chris,
Good point about the lack of damping in one horizontal direction.  Mauro's 
web page shows a trace from a similar arrangement which shows 
this (long tail of low amplitude oscillation after the earthquake signal).  
 
I like the symmetry of your suggested coil setup, but I'm not sure  
changing the coil orientation will fix the problem completely though.  There

would still be undamped motion on one horizontal axis and still would be
some sensitivity of the coil/magnet to horizontal motion, although maybe 
less than before.  
 
If a horizontal damping plate with a vertically aligned magnetic field were 
added it would help.  The Wood-Anderson seismometers had a similar 
problem with the torsion wire "twanging" and solved it with a drop of oil 
around the wire as a damper.
 
Jack
 

-----Original Message-----
From: ChrisAtUpw@....... [mailto:ChrisAtUpw@........
Sent: Thursday, January 02, 2003 7:19 PM
To: psn-l@..............
Subject: Re: Notes to UK--Nick


In a message dated 02/01/2003, ivey@.......... writes:



Mauro,
I'm not sure it will be _very_ sensitive to axial movements.  Horizontal
movements of the magnet will not cause as much change in magnetic flux
linkage through the coil as vertical movements because of the geometry of
the coil and magnet.  For movement in one horizontal direction, the magnet
is moving parallel to the wires at the top of the coil.  For the other
horizontal direction, the magnet is moving parallel to the flux between it's
poles.  In either case any flux change through the coil is a second order
effect due to curvature of the coil and curvature of the magnet path.  For
vertical movement the flux change is a first order effect, proportional to
displacement.  As a guess I would expect the horizontal sensitivity to be 5x
to 10x down from the vertical.  
Jack



Hi Jack,

     I am in agreement with your analysis so far as it goes, but it seems to
be incomplete and the conclusion may not be correct. The magnetic damping is
presumably adjusted to be near critical for vertical motion. This should
also give reasonably good damping for horizontal motion in the plane of the
plate. However, the plate will be only lightly damped for horizontal
oscillations perpendicular to it's plane and it will show a resonant
characteristic with enhanced swing amplitudes. The magnetic field is highly
inhomogeneous and movements of the magnet in this plane can be expected to
give significant induced signals. The sensitivity could be decreased by
mounting the magnet and sensor coil at right angles to the original
orientation shown. 

     An arrangement which is less sensitive to horizontal motion and can
give a larger signal, may be made using a horizontally mounted flat coil
with a vertical cylindrical NdBFe magnet in the centre. A 1/2" dia by 1/2"
long magnet Type #13 from www.wondermagnet.com at $4 would be suitable. The
end of the polepiece should be about the centre plane of the coil. A
suitable coil would be about 3/8" long and 5/8" internal diameter, wound
with 3,000 to 5,000 turns.

     Neodymium magnets type #31 or #39 are capable of giving very high
damping forces, probably at considerably less cost than the 100 lb pull
damping magnets in the original design. I use two opposed #31s with a small
1/8" thick Al plate to critically damp a 1.1 kg pendulum. With very strong
cheap magnets readily available, we might as well make use of them.

     Regards,

     Chris Chapman 



From: Mauro Mariotti [mailto:mariotti@..........
Sent: Monday, December 30, 2002 2:01 PM
Subject: Re: Notes to UK--Nick
Hi all,
I suggest a different approach for homemade vertical sensors. The sensor in
the link you propose is VERY sensitive to axial movements
(horizontal movements) if you give a look to 
http://www.infoeq.it/doc02v_e.htm    
and scroll the page until you reach:
Lacoste suspension sensor 
you can see a good approach to a sensitive long period seismometer.
........ 
Regards Mauro



From: John   & Jan Lahr 
I've scanned Jim's Short Period design plans and put a copy on my web site
here: http://jjlahr.com/science/psn/lehman/
 
This will save on postage and make the plans available to more people.  They
are a good example of how to document a seismic sensor design.










Chris,
Good 
point about the lack of damping in one horizontal direction.  Mauro's 

web 
page shows a trace from a similar arrangement 
which shows 
this 
(long tail of low amplitude oscillation after the 
earthquake signal).  
 
I like 
the symmetry of your suggested coil setup, but I'm not sure  

changing the coil orientation will fix the 
problem completely though.  There 
would 
still be undamped motion on one horizontal axis and still would 
be
some 
sensitivity of the coil/magnet to horizontal motion, although maybe 

less 
than before.  
 
If a horizontal damping plate with a vertically 
aligned magnetic field were 
added 
it would help.  The Wood-Anderson seismometers had 
a similar 
problem with the torsion wire "twanging" 
and solved it with a drop of oil 

around 
the wire as a damper.
 
Jack
 

  
-----Original Message-----
From: ChrisAtUpw@....... 
  [mailto:ChrisAtUpw@........
Sent: Thursday, January 02, 2003 7:19 
  PM
To: psn-l@..............
Subject: Re: Notes to 
  UK--Nick

In a message dated 02/01/2003, 
  ivey@.......... writes:


  
Mauro,
I'm not sure it will be _very_ sensitive to axial 
    movements.  Horizontal movements of the magnet will not cause as much 
    change in magnetic flux linkage through the coil as vertical movements 
    because of the geometry of the coil and magnet.  For movement in one 
    horizontal direction, the magnet is moving parallel to the wires at the top 
    of the coil.  For the other horizontal direction, the magnet is moving 
    parallel to the flux between it's poles.  In either case any flux 
    change through the coil is a second order effect due to curvature of the 
    coil and curvature of the magnet path.  For vertical movement the flux 
    change is a first order effect, proportional to displacement.  As a 
    guess I would expect the horizontal sensitivity to be 5x to 10x down from 
    the vertical.  
Jack


Hi 
  Jack,

     I am in agreement with your analysis so 
  far as it goes, but it seems to be incomplete and the conclusion may not be 
  correct. The magnetic damping is presumably adjusted to be near critical for 
  vertical motion. This should also give reasonably good damping for horizontal 
  motion in the plane of the plate. However, the plate will be only lightly 
  damped for horizontal oscillations perpendicular to it's plane and it will 
  show a resonant characteristic with enhanced swing amplitudes. The magnetic 
  field is highly inhomogeneous and movements of the magnet in this plane can be 
  expected to give significant induced signals. The sensitivity could be 
  decreased by mounting the magnet and sensor coil at right angles to the 
  original orientation shown. 

     An arrangement 
  which is less sensitive to horizontal motion and can give a larger signal, may 
  be made using a horizontally mounted flat coil with a vertical cylindrical 
  NdBFe magnet in the centre. A 1/2" dia by 1/2" long magnet Type #13 from 
  www.wondermagnet.com at $4 would be suitable. The end of the polepiece should 
  be about the centre plane of the coil. A suitable coil would be about 3/8" 
  long and 5/8" internal diameter, wound with 3,000 to 5,000 
  turns.

     Neodymium magnets type #31 or #39 are 
  capable of giving very high damping forces, probably at considerably less cost 
  than the 100 lb pull damping magnets in the original design. I use two opposed 
  #31s with a small 1/8" thick Al plate to critically damp a 1.1 kg pendulum. 
  With very strong cheap magnets readily available, we might as well make use of 
  them.

     Regards,

     
  Chris Chapman 


  

    
From: Mauro Mariotti 
      [mailto:mariotti@..........
Sent: Monday, December 30, 2002 2:01 
      PM
Subject: Re: Notes to UK--Nick
Hi all,
I suggest a different approach for homemade vertical 
      sensors. The sensor in the link you propose is VERY sensitive to axial 
      movements
(horizontal movements) if you give a look to 
      
http://www.infoeq.it/doc02v_e.htm  
and scroll the page until you reach:
Lacoste 
      suspension sensor 
you can see a good approach to a sensitive long 
      period seismometer.
....... 
Regards Mauro


      
From: John & 
        Jan Lahr 
I've scanned Jim's Short Period design plans and put a 
        copy on my web site here: http://jjlahr.com/science/psn/lehman/
This 
        will save on postage and make the plans available to more people.  
        They
are a good example of how to document a seismic sensor 
        design.