This is a variation of an inverted pendulum design.
Using two verticals,  it has two stable points which are horizontally
displaced.  The displacement becomes more obvious as the supporting sides
are made more vertical to increase the pendulum period, but it is present no
matter what vertical angles are used in this design.

An observer can see that this is true by observing that for  horizontal
motion to the left (from centered position), the left side of the tee will
raise less than the right hand falls.  As a result, the center (which is the
support for the  mass) will have a small drop in height.  The same thing is
true for horizontal shifts to the right.

The drop in height of the tee center is countered by an increase in height
of the pendulum mass as it begins motion.  Unfortunately, the rise of
the pendulum as it moves from center (lowest point) is slower than the drop
from the change in tee height when moving off center so it can not
completely overcome (prevent or compensate) the drop of the support center.
This is no longer correct if the pendulum length is made so long that the
pendulum
motion arc is below the  lower pivot points of the sides, which offers a way
to make the device have only one stable point.

In other words, try raising the lower side pivot points while leaving the
cross
plate height the same.  Make the sides shorter and more nearly vertical.
I think that will lengthen the period and cure the bistable problem.

Thanks for making this interesting design available for consideration,

Roger

> .------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
> | Message 2                                                           |
> '------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
> Subject: Different pendulum design
> From:    John or Jan Lahr 
> Date:    Mon, 09 Feb 2004 08:51:16 -0700
>
> I've built a prototype horizontal-pendulum seismic sensor using a design
that
> Brady Romberg, a School of Mines student, came up with.  This still needs
> a lot of work, but I thought the mechanical design was quite interesting.
>
> I've posted a seismogram and some photos here:
> http://jclahr.com/science/psn/romberg/
>
> The weakness in this design is the vertical rod and it's attachment to the
> horizontal plate.  I think flexure in the rod and/or attachment is what
limits
> the period to about 2 seconds.  Perhaps the mass needs to hang from a V-
> shaped support rather than a single rod.
>
> Cheers,
> John
>
>
>
> .------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
> | Message 3                                                           |
> '------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
> Subject: RE: Different pendulum design
> From:    "John D Nelson" 
> Date:    Mon, 9 Feb 2004 11:14:58 -0600
>
> It seems to me that there might be considerable friction losses in the
> supporting points because of the weight that they bear-could this be a
> factor?
>
> -----Original Message-----
> From: psn-l-request@.............. [mailto:psn-l-request@...............
> On Behalf Of John or Jan Lahr
> Sent: Monday, February 09, 2004 9:51 AM
> To: psn-l@..............
> Cc: bromberg@.........
> Subject: Different pendulum design
>
> I've built a prototype horizontal-pendulum seismic sensor using a design
> that
> Brady Romberg, a School of Mines student, came up with.  This still
> needs
> a lot of work, but I thought the mechanical design was quite
> interesting.
>
> I've posted a seismogram and some photos here:
> http://jclahr.com/science/psn/romberg/
>
> The weakness in this design is the vertical rod and it's attachment to
> the
> horizontal plate.  I think flexure in the rod and/or attachment is what
> limits
> the period to about 2 seconds.  Perhaps the mass needs to hang from a V-
> shaped support rather than a single rod.
>
> Cheers,
> John
>
> __________________________________________________________
>
> Public Seismic Network Mailing List (PSN-L)
>
> To leave this list email PSN-L-REQUEST@.............. with=20
> the body of the message (first line only): unsubscribe
> See http://www.seismicnet.com/maillist.html for more information.
>
>
> .------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
> | Message 4                                                           |
> '------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
> Subject: RE: Different pendulum design
> From:    John or Jan Lahr 
> Date:    Mon, 09 Feb 2004 12:58:09 -0700
>
> Yes, very possibly.  I haven't come up with a better support system.
>
> John
>
> At 10:14 AM 2/9/2004, John Nelson wrote:
>
> >It seems to me that there might be considerable friction losses in the
> >supporting points because of the weight that they bear-could this be a
> >factor?
>
>
>
> __________________________________________________________
>
> Public Seismic Network Mailing List (PSN-L)
>
> To leave this list email PSN-L-DIGEST-REQUEST@.............. with
> the body of the message (first line only): unsubscribe
> See http://www.seismicnet.com/maillist.html for more information.

__________________________________________________________

Public Seismic Network Mailing List (PSN-L)