I just came across this thread and for once I can contribute a little.
Back in the eighties I used to build gravimeters. They are indeed
exquisite instruments. If there was someplace to post pictures to I would
post one of a sensor. Most of the speculations are on the right track.
They all have a proof mass and a means of support, usually a spring, and
a position detection system which can be optical or electronic. Our
instruments operated at a reduced pressure to reduce convection currents
but still high enough to provide some damping. The gas does not have to
be air. A two stage oven will stabilize the sensor to within about one
millidegree and then we would compenstate in software below that
resolution. The proof mass weighed approx. one gram, and was suspended
from a spring. Gravity variations were determined by measuring the
voltage required to electrostatically return the mass to a null position.
This was a 'relative' gravimeter. Absolute gravimeters are very different
things, usually employing a corner cube reflector dropped or tossed in a
vacuum. FYI the gravity unit is a Gal (for Galileo) and since it is so
large a milliGal is the common unit. Field strength is approx 980000
milliGal and our portable gravimeters could measure to 1 microGal. A
worldwide instrument requires a range of 7000 milliGal. Not all older
instruments could operate at all locations on Earth without resetting.
The principals that these instruments operate on are not 'rocket
science', I will agree, but the skills required to build them are rare.

(Dan Caster) wrote in news:3183eab.0312262345.71cd47e9
@posting.google.com:

The temperature is only one of the things that needs to stabilize.
Another problem is air currents inside the box and since the device
must have very low hysteris, it takes a long time to dampen out the
oscilations. I was told that some of the gravity meters had a partial
vacuum in the box to reduce the problem with air currents. A full
vacuum did not provide the damping.
But the guy probably just went to lunch.

I also understand that the usual procedure for gravity surveys used
two gravity meters. One stayed at the starting location and the
gravity was recorded there every so often. The other meter was moved
around the area being surveyed. The reason is that the gravity
measured varies with time. Well with the position of the sun and moon
mostly.

Dan

(brownnsharp) wrote in message

We had the gravity constant checked at our laboratory in Houston. We
were using precision dead-weight testers to calibrate pressure gauges
(How precision? 0.1 psi resolution at 20,000 psi static). A shaky old
man in a baseball cap came in with a box about a foot cubed with a
battery pack attached. He set the box on our bench, stared through an
eyepiece, and said it had not stabilized yet. He said he would be
back later, and disappeared for an hour. Then he came in and read the
gravity constant at our location. I think the box had nothing more
precise than a weight with a spring on it, and a scale read with a
microscope. The precision, and stability, came from maintaining a
VERY precise temperature in the box. I.e., the wait for the box to
stabilize better was a wait for the temperature to stabilize.

In summary, "gravity meters" have been around a long time, and don't
require rocket-science to build.