Ted,

I think you are finding that simply having long period does not necessarily 
make a good sensor.  What seems to be happening is that you are largely 
lengthening the period by reducing the restoring force on a distributed 
pendulum.  In the design you describe, that is also having the effect of 
reducing its horizontal motion sensitivity, pretty much in proportion.

Another design which would act in the same way would be a heavy disk with 
an axle drilled just slightly off-center.  It would rotate heavy side down 
and would oscillate quite slowly, but its sensitivity to horizontal motion 
would be very small.  It might still be fairly sensitive to tilt 
oscillation, though.

Brett

At 12:01 PM 2/24/2008 -0700, you wrote:
>Hi John and All,   This is clear to me, because its right in front of me, 
>so I hope you can picture the mock up from my description.
>My latest attempt go like this;
>
>A 48" dowel with a pivot, at 24".  Equal mass (two machine nuts) one fixed 
>on the bottom of the propeller, and the top nut adjustable up or down.
>The goal being to adjust the top nut in the up direction until it near 
>reached the top of the dowel, which would equal the mass and position of 
>the fixed bottom nut.   Stopping just short of the end of the dowel should 
>result in a longer period, and still always come to rest at vertical.
>
>Earlier all this did happen, but the propeller tended to come to rest at 
>the 2:oo o'clock position, and as the nut moved up the resting position 
>wanted to become horizontal.
>
>I though, my pivot hole was off center, just a bit, and that could cause 
>the propeller to become other than vertical.
>So I added this:    I drilled and installed a piece of wire about 12" long 
>into the bottom of the propeller.  Now it looks like an inverted "T" and 
>by moving the wire to the left or right, I can bring the propeller back to 
>vertical.   The wire shift the weight of the prop. until it hang vertical.
>
>After this correction, I continued to move the top nut up to increase the 
>period;   I was able to get 45 seconds maximum, but it was unstable.   I 
>lowered the top nut a bit, and got 20 seconds, still a bit unstable.   I 
>lowered more and got 13 seconds stable and hanging vertical.
>
>I have no idea what to do with it now, except to confirm it works, but it 
>is touch-y.   A 24" pendulum was extended from 1.5 seconds to 45 seconds max.
>Thanks, Ted


__________________________________________________________

Public Seismic Network Mailing List (PSN-L)