In a message dated 2007/09/15, ptrowe@......... writes:

> I left out one important piece of the circuit. Chris' reply to you jogged 
> my memory. Since you drive the bridge with AC, you have to have a synchronous 
> rectifier after the gain stages to turn the signal back to DC. This is done 
> with an analog switch and an op amp. Easy.
> As Chris mentioned, the homemade liquid filled tube may be your best bet for 
> building the sensor yourself. I saw one of these at Cal Tech many years ago. 
> It was
> about 2 meters long and had a horizontal tube filled with Mercury. They 
> sensed the level of the Mercury in a cup at each end capacitively. With today's 
> capacitive sensor ICs, this should be easy to do.

Hi Chuck,

       There is a distinct advantage in using a sine wave drive if you want 
really low noise. This can be generated with a crystal oscillator driving a 
binary counter and some load resistors into a summing amplifier and a low pass 
filter. It is described on Sean Morrissey's website.

       You might be interested to read Gile, WW, Geophys. J. Roy. Astro. Soc. 
vol 36 1974 pp 153-165 "A Mercury Pendulum Seismometer" Your local library 
should be able to get you a photocopy. He describes a seismometer built by 
Benioff, but he does not go into the technique of using large end cups to extend 
the natural period,  or strong overdamping to flatten the natural pendulum dog 
leg response below the resonant period.

       Mercury is highly toxic, particularly the vapour, even at room 
temperature, so your equipment needs to be 100% sealed. Always handle it over a deep 
plastic tray to catch any spillage. It forms amalgams with many common metals 
and it will 'rot' components made of brass, copper, aluminum, zinc, solder.... 
SS is OK. It is also very expensive! 

       Regards,

       Chris Chapman   
In a me=
ssage dated 2007/09/15, ptrowe@......... writes:



I left out one important piece=20=
of the circuit. Chris' reply to you jogged my memory. Since you drive the br=
idge with AC, you have to have a synchronous rectifier after the gain stages=
 to turn the signal back to DC. This is done with an analog switch and an op=
 amp. Easy.

As Chris mentioned, the homemade liquid filled tube may be your best bet for=
 building the sensor yourself. I saw one of these at Cal Tech many years ago=
.. It was

about 2 meters long and had a horizontal tube filled with Mercury. They sens=
ed the level of the Mercury in a cup at each end capacitively. With today's=20=
capacitive sensor ICs, this should be easy to do.




Hi Chuck,



       There is a distinct advantage in using=20=
a sine wave drive if you want really low noise. This can be generated with a=
 crystal oscillator driving a binary counter and some load resistors into a=20=
summing amplifier and a low pass filter. It is described on Sean Morrissey's=
 website.



       You might be interested to read Gile, W=
W, Geophys. J. Roy. Astro. Soc. vol 36 1974 pp 153-165 "A Mercury Pendulu=
m Seismometer" Your local library should be able to get you a photocopy.=
 He describes a seismometer built by Benioff, but he does not go into the te=
chnique of using large end cups to extend the natural period,  or stron=
g overdamping to flatten the natural pendulum dog leg response below the res=
onant period.



       Mercury is highly toxic, particularly t=
he vapour, even at room temperature, so your equipment needs to be 100% seal=
ed. Always handle it over a deep plastic tray to catch any spillage. It form=
s amalgams with many common metals and it will 'rot' components made of bras=
s, copper, aluminum, zinc, solder.... SS is OK. It is also very expensive! <=
BR>


       Regards,



       Chris Chapman