All,

Several days ago I received a strong motion sensor prototype from one of
our members Roger Baker in Texas, USA. I placed some pictures of it here
http://www.seismicnet.com/strongmotion/. The first picture show an overall
view of the device and the others show a closer look at the end of the
boom. The boom is made from a hack saw blade. The device is forced-balanced
with an optical displacement sensor using a red LED and photo transistor
with an interrupter mounted on the end of the pendulum.  The device uses a
disk drive head actuator with the coil (~133 ohms)  mounted on the
pendulum. Mounted on the frame, made of wood, are two flat magnets.

The device worked right out of the box. It has two outputs, one is a low
gain output and another higher gain. The low gain output gives a nice DC
change of voltage as one rotates the device 90 degrees. It looks like the
device has a 2.45 v / g output. The higher gain output needs to have an
offset pot adjusted so that the output is in a linear range. I think this
can be fixed by AC coupling this output. Connecting a scope to this output
I can see me jumping around on the ground  similar to what a geophone would
pick up. 

What Roger has given me (BTW, thanks a lot Roger!) is a mechanical
structure and electronics that I can improve on. As I told Roger, I do ok
with electronic and software design, but mechanical design and construction
are hard for me to do.

I have a lot of questions about what can be done to improve and make this
type of sensor more reproducible etc. For now I have two questions....

The first one is about the natural period of the pendulum when trying to
construct an accelerometer. I understand that to create an accelerometer
one can use a pendulum and a displacement sensor, and for frequencies below
the period of the pendulum you will be recording acceleration. The device
Roger gave me has a period of around 8-10 hz. A little low..I will be
shortening the boom it to increase the period, but I was wondering if the
feedback or output can be modified so that it has a flat response to
acceleration even if the pendulum period is lower then what you want to
record? If I understand how Sean-Thomas's VBB sensor works, the feedback is
controlled so that the sensor produces a flat response to velocity, and
that the actual period of the pendulum is not that important. Can the same
thing be done with acceleration, or does the period of the pendulum need to
be above the max frequency? I ask this more for curiosity. I think it will
be easy to simply shorten the pendulum to get a 20 - 30 hz natural period
and use the more simple force-balance feedback setup. 

My other question, for now, is about using light, and photo transistor and
LEDs in particular, for the displacement sensor. The parts are easy to get
but I wonder about linearity, temperature and long term stability. Any
thoughts???? This weekend I will be checking the linearity and temperature
stability. Long term stability is harder to check for....

Regards,
Larry Cochrane
Redwood City, PSN





  

_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)