Chris,

I'm sorry that I never posted the AGU abstract. I still think that
I can get much more done than reality allows.  I used to be able
to run a dozen projects at the ends of the earth, but my disabilities 
are now an unfortunate limitation. And then someone shared the 
two-week flu that is going around. 

My major effort in early October was to complete the preliminary
technical report on the beam-balance tiltmeter: a very difficult
effort to define the project from all the theoretical, mechanical,
and electronic aspects, as well as to present some very initial data. 
The report runs 60 pages of text and 40 of figures, and none of it
is in a condition to mark up to HTML. I have sent it to those who I
thought could provide a really critical review, but so far I have only
got back "nicely written ... as far as I got ... about page 5". So I guess
that if it was complicated for me to write, maybe it IS unreadable. I had 
hoped for (and they were anxious to provide) good criticism from ASL
(the national Albuquerque Seismological Lab.) but they are in the 
process of relocating because their lease on the reservation went up
by over 10X. They moved their servers, etc, on Oct. 18. I don't know
the fate of the ANMO test site (a very quiet deep tunnel) where we were
planning on testing the new tiltmeter. (So I am setting up a side-by-side
test with an STS-2 with an 8-channel, 24-bit digitizer in my remote vault.)

Another problem came up last week when the leaf-spring vertical seis
in the basement here bonked; details next.

Here is the AGU abstract. I have submitted it with Dr. Mitchell as a
second author and presenter, since I don't think I will be able to attend;
but I am looking to attend the SSA meeting in SFO in April with new data.
 
Regards,
Sean-Thomas

________________________________________________
An abstract for the fall 2000 AGU meeting.

A Beam-Balance Broadband Tiltmeter 
That is Insensitive to Horizontal Acceleration.

 Sean-Thomas Morrissey (sean@............
 Brian J. Mitchell (mitchell@............
 (Dept. of Earth and Atmospheric 
Sciences, St.Louis University, 3507 Laclede; St. Louis 
MO, 63103; 314 977 3129; 

A new beam-balance tiltmeter has been developed that does not respond 
to horizontal ground movement. Since all horizontal seismometers are 
also tiltmeters they are sensitive to tilt, especially at longer periods,
and seismic data can be compromised in the period range of 20 to 3000 
seconds.  The most important noise source in the horizontal data is 
tilting of the pier, mostly due to barometric loading.  But all tiltmeters,
up to now, are also seismometers, (ie. they are sensitive to horizontal 
translation), so can not be used to separate tilt noise from seismic signals. 

Our new beam balance tiltmeter does not respond to horizontal 
acceleration because the masses at each end of the horizontal beam are 
suspended through the exact center of mass. This system is inherently 
unstable, so broadband feedback is used to control it.  With appropriate
feedback, the beam remains relatively horizontal when the base is tilted,
and the output is the relative motion at the displacement detector. 
There is no rotation or output when the base is translated horizontally 
along the axis of the beam.

The beam-balance tiltmeter is designed with separated lead masses 
mounted in an aluminum bar that is suspended exactly at the center 
of mass of the horizontal beam with a new low torque hinge flexure.
The center of mass is trimmed by a unique vertical mass centering 
adjustment above the flexures.  Displacement transducers and  compact 
force feedback coils with rare-earth magnets are placed at both ends 
of the beam.  

Three sensors have been assembled as "proof of concept" prototypes.
One instrument is operating at station CCMO, near Saint Louis University,
and data from it are being digitized.  Another is on a table and can be 
used to demonstrate that tilt can be separated from horizontal acceleration 
by simply sliding it horizontally.  The static or DC tilt sensitivity of 
the prototype is about 120 millivolts per microradian, and the resolution 
is better than 0.1 nanoradian. Initial comparisons using data generated 
by large quakes on the horizontal components of nearby broadband seismic 
stations (18 km distant) show that the response to horizontal acceleration
is reduced by a factor greater than 1000 while maintaining the equivalent 
tilt sensitivity of the seismometer. 

Ideally, the noise recorded in the tiltmeter output will exactly emulate 
the tilt noise from the seismometer, at least in the flat portions of 
their broadband velocity response.  The success of this new instrument
has a tremendous implications for broadband stations in all regions of 
the world  where tilting from barometric, thermal, hydrologic, etc., 
effects may limit the usefulness of the horizontal data.

____________________

__________________________________________________________

Public Seismic Network Mailing List (PSN-L)