In a message dated 2007/12/16, tchannel1@............ writes:

> The movement of the earth resulting from an earthquake, will move the 
> entire apparatuses, and the room it is in.  However the mass at the ends of the 
> inverted T will move less, relative to the sensor and the room it is in.

Hi Ted,

       A T bar torsion system is fine for clock timing, provided you have a 
strip / wire of Elinvar to suspend it, otherwise the torsional temperature 
coefficient is too great. The low energy requirement for clocks is a consequence 
of the long period and simple gearing.

       However, I don't see how you would use it as a seismic sensor? There 
are two modes of oscillation, the torsional one and the whole body as a simple 
pendulum. Only the rotational one has a long period and this is not strongly 
excited by moving the suspension, unless the quake is local, when there may be 
rotational components.

       If you look back into the history of seismometers, one type which does 
not seem to have be much used, except for miniature quartz devices, is the 
torsional type . It has a horizontal light beam with a weight on the end and two 
suspension wires spaced a short distance apart, the center one going upwards 
and the lower one going downwards. They can have quite long periods and were 
mentioned in one of your pendulum references. You might need Invar wire.

                        XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX   Frame          
              
                          \
                            \   WIRE                                          
  
                              \                                               
_____
                        
--------\---------------------------------------------I        I MASS
                        
I------------------------------------------------------I____I
                         \                   RIGID ARM
                           \
                             \   WIRE
                               \
                         XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX Frame

       Regards,

       Chris Chapman   
In a me=
ssage dated 2007/12/16, tchannel1@............ writes:



The movement of the earth resul=
ting from an earthquake, will move the entire apparatuses, and the room it i=
s in.  However the mass at the ends of the inverted T will move less, r=
elative to the sensor and the room it is in.




Hi Ted,



       A T bar torsion system is fine for cloc=
k timing, provided you have a strip / wire of Elinvar to suspend it, otherwi=
se the torsional temperature coefficient is too great. The low energy requir=
ement for clocks is a consequence of the long period and simple gearing.



       However, I don't see how you would use=20=
it as a seismic sensor? There are two modes of oscillation, the torsional on=
e and the whole body as a simple pendulum. Only the rotational one has a lon=
g period and this is not strongly excited by moving the suspension, unless t=
he quake is local, when there may be rotational components.



       If you look back into the history of se=
ismometers, one type which does not seem to have be much used, except for mi=
niature quartz devices, is the torsional type . It has a horizontal light be=
am with a weight on the end and two suspension wires spaced a short distance=
 apart, the center one going upwards and the lower one going downwards. They=
 can have quite long periods and were mentioned in one of your pendulum refe=
rences. You might need Invar wire.



            &nbs=
p;           XXXXXXXXXXXXX=
XXXXXXXXXXXXXXXXXXXXXXX   Frame      =
;            &nb=
sp;     

            &nbs=
p;             \=


            &nbs=
p;            &n=
bsp;  \   WIRE        =
;            &nb=
sp;            &=
nbsp;          

            &nbs=
p;            &n=
bsp;    \        &nbs=
p;            &n=
bsp;            =
             ___=
__

            &nbs=
p;           --------\----=
-----------------------------------------I     &nbs=
p;  I MASS

            &nbs=
p;           I------------=
------------------------------------------I____I

            &nbs=
p;            \ =
            &nbs=
p;     RIGID ARM

            &nbs=
p;            &n=
bsp; \

            &nbs=
p;            &n=
bsp;   \   WIRE

            &nbs=
p;            &n=
bsp;     \

            &nbs=
p;            XXXXXXX=
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX Frame



       Regards,



       Chris Chapman