Hi Karl
 I would be interested in any force feedback formulas you might have so I can
try to predict the performance of my sensor. I was going to try STM's formulas
with values for the integration portion  such that it can be ignored and the
diff and proportional set according to the values I am currently using for my
FB horizontal.
Barry

Ps
    would posting something on your website be possible?


Karl Cunningham wrote:

> Larry --
>
> Glad to help.
>
> >Karl,
> >That would be a big help. What information would you need for the number
> >crunching?
>
> Off the top of my head, I think I would need the feedback circuit, the
> characteristics of the feedback transducer (voice coil, or whatever), and
> either the period or the amount of mass and the spring constant.  These may
> not all be easy to come up with, but some relatively simple experiments can
> be done to figure it out.
>
> The calculations are very similar to what Sean-Thomas has posted on his web
> site for the VBB, except for the circuit equations.
>
> >Yes, but does the output at that point have a flat freq. response to
> >acceleration above the period of the pendulum or does it fall of at the
> >12db/octave or 6db/octave rate?
>
> If the loop gain is high enough, the acceleration response should be flat
> with repsect to frequency.  The idea is that if the loop gain is high, the
> feedback will be maintaining the mass in a fixed position with respect to
> the base (and the earth).  If this is the case, then the current in the
> coil has to be proporational to acceleration (barring any friction in the
> mechanical system), independent of frequency.  Of course, the response will
> drop off at the extremes, but this will be predicted by loss of loop gain
> at these frequencies.
>
> Saying it another way, the loop gain also determines the measurement
> accuracy, and for our purposes inaccuracy is approximately the inverse of
> loop gain.  With a loop gain of 10, the inaccuracy will be within about 0.1
> (10%).  I would think that determining acceleration amplitude within 10% is
> probably good enough for us.
>
> Loop gain in the center of the frequency range will probably be much higher
> -- maybe 100 or more.
>
> -- Karl
>


_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)