In a message dated 27/10/2007, tchannel1@............ writes:

Hi Chris,  What you described below, I think, was one of the  illustration,  
Let me describe it this way...........
Lets say a 36" horz rod.............from the balance point 18" connect a  
short 2" x 4"  (L) shape rod, the 2" leg is welded to the balance point  of the 
horz rod.
Now the 4" leg is inserted into some sort of bearing, and the whole thing  
swings back and forth from this 4" axle.   You have two equal  masses, one on 
the left and one on the right of the 36" horz. rod.    NOW CAN  one extend the 
period of the device by relocating one or other  of the masses?



Hi Ted,
 
    You would need to move both masses equally, or they  would not balace. If 
the rod is of low weight and the masses are K from the  centre, the moment of 
inertia is roughly 2MK^2. The restoring force is roughly  2Mg x 4" x dTheta. 
Changes in the length of the 4" rod will make a big  contribution to the 
period.
 
    As John has mentioned, the sensitivity to seismic  motion is low, but if 
you use a quad NdFeB magnet + coil sensor, you can get 10x  the output of a U 
Alnico magnet design. If you also design your amplifier for  low noise, you 
can get maybe 1/10 the noise of available amplifiers. 
    Referring to _http://jclahr.com/science/psn/cochrane/index.html_ 
(http://jclahr.com/science/psn/cochrane/index.html)  if  you reduce R29 from 400 K to 
12 K and R10 from 5.1 K to 150 Ohm you will reduce  the shot noise by ~x5 and 
also reduce the 1/f current noise. Using  discreet low noise input transistors 
can give a further noise reduction of x5,  or more. Using such a system as 
part of a chopper amplifier will dramatically  reduce the 1/f voltage and 
current noise. See 
_http://www.linear.com.cn/pc/downloadDocument.do?navId=H0,C1,C1010,C1203,P1248,D4136_ 
(http://www.linear.com.cn/pc/downloadDocument.do?navId=H0,C1,C1010,C1203,P1248,D4136) 
 
    I like the reference 
_http://www.phy.mtu.edu/~suits/PH3110/pendulums.html_ (http://www.phy.mtu.edu/~suits/PH3110/pendulums.html)  The  wide angle 
crossed wire suspension system is very interesting. But crossed foils  might be 
more robust. You can also extend the period of Lehman and vertical  pendulums up 
to x10 by magnetic repulsion. This modifies the force /  position 
relationship, but you may then be very sensitive to local magnetic  field noise. 
 
    Regards,
 
    Chris Chapman



   






In a message dated 27/10/2007, tchannel1@............ writes:
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
Hi Chris,  What you described below, I think, was one of the=20
  illustration,  Let me describe it this way...........
  
Lets say a 36" horz rod.............from the balance point 18" connec=
t a=20
  short 2" x 4"  (L) shape rod, the 2" leg is welded to the balance poi=
nt=20
  of the horz rod.
  
Now the 4" leg is inserted into some sort of bearing, and the whole t=
hing=20
  swings back and forth from this 4" axle.   You have two equal=20
  masses, one on the left and one on the right of the 36" horz. rod. &n=
bsp;=20
  NOW CAN  one extend the period of the device by relocating one or oth=
er=20
  of the masses?

Hi Ted,
 
    You would need to move both masses equally, or=20=
they=20
would not balace. If the rod is of low weight and the masses are K from the=20
centre, the moment of inertia is roughly 2MK^2. The restoring force is rough=
ly=20
2Mg x 4" x dTheta. Changes in the length of the 4" rod will make a big=20
contribution to the period.
 
    As John has mentioned, the sensitivity to seism=
ic=20
motion is low, but if you use a quad NdFeB magnet + coil sensor, you can get=
 10x=20
the output of a U Alnico magnet design. If you also design your amplifier fo=
r=20
low noise, you can get maybe 1/10 the noise of available amplifiers. 
    Referring to http://jclahr.com=
/science/psn/cochrane/index.html if=20
you reduce R29 from 400 K to 12 K and R10 from 5.1 K to 150 Ohm you will red=
uce=20
the shot noise by ~x5 and also reduce the 1/f current noise. Using=
=20
discreet low noise input transistors can give a further noise reduction of x=
5,=20
or more. Using such a system as part of a chopper amplifier will dramaticall=
y=20
reduce the 1/f voltage and current noise. See http://www.linear.com.cn/pc/downloadDocument.do?navId=3DH=
0,C1,C1010,C1203,P1248,D4136
 
    I like the reference http://www.phy.=
mtu.edu/~suits/PH3110/pendulums.html The=20
wide angle crossed wire suspension system is very interesting. But crossed f=
oils=20
might be more robust. You can also extend the period of Lehman and vertical=20
pendulums up to x10 by magnetic repulsion. This modifies the force /=20
position relationship, but you may then be very sensitive to local magnetic=20
field noise. 
 
    Regards,
 
    Chris Chapman