Brett,
      Everything depends ultimately on the factor that I mentioned as regul=
ating the sensitivity of a capacitive detector; i.e., the size of the elect=
ric field between the plates.  For a given "bias" voltage (necessarily a.c.=
), the 'driver' then is the spacing between the plates.  For sufficiently l=
arge plate areas (relative to the gap spacing), the field is well approxima=
ted as being uniform and given simply by the ratio of voltage to spacing.  =
In the VolksMeter and also the Cavendish balance, the nominal spacing emplo=
yed (better part of 1 mm) is very large compared to what is routinely used =
in commercial force balance instruments (measured in microns).  In other wo=
rds, for reason of user-friendliness, we have opted for a great sacrifice i=
n ultimate-possible sensitivity.  The great performance of the STS instrume=
nts derived from the 'watchmaker-like' artisanship of the master craftsman,=
 Gunar Streckeisen.  On the other hand, I'm like one with 'two thumbs' on e=
ach hand when it comes to trying to copy the man.  That I could build anyth=
ing even close to performance approaching his instruments only in the lowes=
t frequency regimes--is a testament of the 'advantage' of my sensor at low =
frequencies, operating in an unconventional manner.
      One configuration of my sensor allows operation on the basis of gap s=
pacing change, so that it can be used in a force balance arrangement, with =
potentially greater sensitivity than conventional commercial instruments-no=
ne of which, to my knowledge-are fully differential.  Allan Coleman did bui=
ld his instrument (MKXXI) that is described on my webpage to operate in thi=
s manner.  The electronics is shown in his Fig. 8, and he estimated his sen=
sitivity to be 1415 V/m/s.  I don't know how this figure compares with othe=
r instruments in the frequency range important to earthquake waves, since I=
 am accustomed to acceleration specifications rather than velocity.  Of cou=
rse one can transform back and forth between the two, which I haven't done =
for this case.
          I have chosen to avoid force balance for one reason that is not w=
ell appreciated.  Real springs are influenced by non-static metastabilities=
 that influence the shape of the potential energy well; in other words, the=
 non-Hookean spring is not consistent with a parabola.  In terms that some =
engineers appreciate, there is an advantage to dithering because of these m=
etastabilities.  In the more modern physics terminology, we say that there =
is an improvement in SNR that is possible (in the presence of nonlinear (co=
mplex) defect structures involving dislocations) by taking advantage of sto=
chastic resonance.   Thus my usual SDC sensing arrangement is one that oper=
ates on the basis of area variation rather than gap-spacing variation.  By =
means of electrode arrays, it allows a very large mechanical dynamic range =
while retaining a decent, reasonably constant sensitivity over the whole ra=
nge.  By contrast, when using gap-variation for the sensing means, force ba=
lance is required-since otherwise the sensor becomes highly nonlinear.  For=
ce balance tends (at least in the 'low and slow' limit) toward 'latching' o=
f the inertial mass in the small, localized trapping sites.  On the other h=
and, the mode I'm using allows 'skating over the washboard'.  Sometimes fol=
ks in the stochastic resonance world talk about a similar thing by means of=
 a ball rolling on a track similar in shape to an egg carton.
     Randall
Brett,
      Everything depends=
 ultimately on the factor that I mentioned as regulating the sensitivity of=
 a capacitive detector; i.e., the size of the electric field between the pl=
ates.  For a given “bias” voltage (necessarily a.c.), the =
‘driver’ then is the spacing between the plates.  For suff=
iciently large plate areas (relative to the gap spacing), the field is well=
 approximated as being uniform and given simply by the ratio of voltage to =
spacing.  In the VolksMeter and also the Cavendish balance, the nomina=
l spacing employed (better part of 1 mm) is very large compared to what is =
routinely used in commercial force balance instruments (measured in microns=
).  In other words, for reason of user-friendliness, we have opted for=
 a great sacrifice in ultimate-possible sensitivity.  The great perfor=
mance of the STS instruments derived from the ‘watchmaker-like’=
 artisanship of the master craftsman, Gunar Streckeisen.  On the other=
 hand, I’m like one with ‘two thumbs’ on each hand when i=
t comes to trying to copy the man.  That I could build anything even c=
lose to performance approaching his instruments only in the lowest frequenc=
y regimes--is a testament of the ‘advantage’ of my sensor at lo=
w frequencies, operating in an unconventional manner.
      One configuration of my sensor=
 allows operation on the basis of gap spacing change, so that it can be use=
d in a force balance arrangement, with potentially greater sensitivity than=
 conventional commercial instruments—none of which, to my knowledge&#=
8212;are fully differential.  Allan Coleman did build his instrument (=
MKXXI) that is described on my webpage to operate in this manner.  The=
 electronics is shown in his Fig. 8, and he estimated his sensitivity to be=
 1415 V/m/s.  I don’t know how this figure compares with other i=
nstruments in the frequency range important to earthquake waves, since I am=
 accustomed to acceleration specifications rather than velocity.  Of c=
ourse one can transform back and forth between the two, which I haven’=
;t done for this case.  
 &nbs=
p;        I have chosen to avoid fo=
rce balance for one reason that is not well appreciated.  Real springs=
 are influenced by non-static metastabilities that influence the shape of t=
he potential energy well; in other words, the non-Hookean spring is not con=
sistent with a parabola.  In terms that some engineers appreciate, the=
re is an advantage to dithering because of these metastabilities.  In =
the more modern physics terminology, we say that there is an improvement in=
 SNR that is possible (in the presence of nonlinear (complex) defect struct=
ures involving dislocations) by taking advantage of stochastic resonance. &=
nbsp; Thus my usual SDC sensing arrangement is one that operates on th=
e basis of area variation rather than gap-spacing variation.  By means=
 of electrode arrays, it allows a very large mechanical dynamic range while=
 retaining a decent, reasonably constant sensitivity over the whole range.&=
nbsp; By contrast, when using gap-variation for the sensing means, force ba=
lance is required—since otherwise the sensor becomes highly nonlinear=
..  Force balance tends (at least in the ‘low and slow’ lim=
it) toward ‘latching’ of the inertial mass in the small, locali=
zed trapping sites.  On the other hand, the mode I’m using allow=
s ‘skating over the washboard’.  Sometimes folks in the st=
ochastic resonance world talk about a similar thing by means of a ball roll=
ing on a track similar in shape to an egg carton.  
     Randall
=
=