In a message dated 19/09/02, rlthompson@................. writes:

>  My current project is based on the Shackleford-Gundersen design, and I have
> a few questions about the design.
> 
>      I'd like some help with the following questions:
> - what is a practical pendulum length

       About 12" will give a 1 sec pendulum. Anything bigger tends to get 
unwieldy,  difficult to mount and use. A 2 sec pendulum is about 39" long. K&
S Metals sell various tube in 1' lengths at model making shops.

> - what way would the mass of the pendulum affect performance.
> 
       1 oz is the ~real minimum from noise considerations. ~3oz total is 
more usual. I would suggest making the magnet part of the bob mass, rather 
then putting it part way up the pendulum. This discourages flexing 
oscillations of the pendulum and marginally increases the period. 

> 'm assuming that there is some relationship between the pendulum mass and 
> the amount of force feedback, is there a rule of thumb?  (I'm not found of
> doing the math).  Hopefully, the feedback loop gain adjustment is simply a 
> matter of setting the gain just to the point that oscillation is induced, 
> and then backing it off, will that work?

 Hi Ron,

       There is a large download about the SG from Larry on the PSN Website 
which may be helpful. Also look at http://www.iinet.com.au/~ajbv/index.html 
and http://pages.prodigy.net/fxc/

       To set the damping, you reduce the damping control, displace the 
pendulum and then release it. You will probably see several oscillations. If 
it oscillates continually, reverse the coil connections. Increase the damping 
and repeat the performance until the trace just drops quickly to near zero, 
but without oscillating or actually crossing the zero line. It is then 
critically damped.  

       Regards,

       Chris Chapman
In a message dated 19/09/02, rlthompson@................. writes:



 My current project is based on the Shackleford-Gundersen design, and I have

a few questions about the design.



     I'd like some help with the following questions:

- what is a practical pendulum length




       About 12" will give a 1 sec pendulum. Anything bigger tends to get unwieldy,  difficult to mount and use. A 2 sec pendulum is about 39" long. K&S Metals sell various tube in 1' lengths at model making shops.



- what way would the mass of the pendulum affect performance.



       1 oz is the ~real minimum from noise considerations. ~3oz total is more usual. I would suggest making the magnet part of the bob mass, rather then putting it part way up the pendulum. This discourages flexing oscillations of the pendulum and marginally increases the period. 



I
'm assuming that there is some relationship between the pendulum mass and the amount of force feedback, is there a rule of thumb?  (I'm not found of

doing the math).  Hopefully, the feedback loop gain adjustment is simply a matter of setting the gain just to the point that oscillation is induced, and then backing it off, will that work?




 Hi Ron,



       There is a large download about the SG from Larry on the PSN Website which may be helpful. Also look at http://www.iinet.com.au/~ajbv/index.html and http://pages.prodigy.net/fxc/



       To set the damping, you reduce the damping control, displace the pendulum and then release it. You will probably see several oscillations. If it oscillates continually, reverse the coil connections. Increase the damping and repeat the performance until the trace just drops quickly to near zero, but without oscillating or actually crossing the zero line. It is then critically damped.  



       Regards,



       Chris Chapman