In a message dated 2007/03/15, mckimzey@........... writes:

> In a previous post I had mused that the Volksmeter electronics might be 
> retrofitted to a Shackleford-Gundersen type assembly.  Well, it would seem 
> that someone in Austria did exactly that.  Here is the link:  
> http://www3.htl-hl.ac.at/homepage/IP_SEISMO/index_en.php?page=abstract
> 
> It looks like a fairly easy setup, but some things I noticed - 1) they have 
> not seemed to insulate against air currents as much as others with a true SG 
> 
> sensor.  2) they are using active damping like an SG using data from the 
> AD7745.

Hi Mike,

       You need good mechanical and thermal construction as well as good 
electronics if you are to get excellent seismic results.
       Although this seismometer is housed in a basement, it's cover is far 
to large and it is not thermally insulated. No internal electric top heater 
seems to have been provided to ensure a stable air coulumn. I would expect it to 
be moderately noisy.
       A much better construction would be to use a thick wall vertical 
Aluminum tube to house the pendulum and to pad out any air spaces with open cell 
plastic foam.
       Putting a unshielded feedback magnet on a seismometer arm is a NO, NO 
- NEVER !! It will happily pick up noise in the Earth's field, noise from the 
power wiring of the building and passing trucks and cars. You need the 
feedback force to act ~through the centre of gravity, NOT half way up the arm! You 
can get excellent results from a horizontal bar magnet mounted on the frame and 
a Maxwell feedback coil mounted on the arm near the mass. Electronic velocity 
force feedback systems are inherently noisy. If you only want to damp the 
pendulum, a simple quad magnet + plate system will give excellent performance, 
without the cost of the electronics.
       There are two suspension types which should be avoided like the 
plague! One is the point in a cup / on a plane and the other is the metallic knife 
blade on a flat. Both stress the metal contact near to or over the deformation 
stress limit and give a variable and unreliable performance. They are highly 
likely to limit the maximum period which can be attained. The 'so called' knife 
edge bearings used in chemical balances are nothing of the sort. They use a 
60 degree triangular section with a lapped tip. This effectively gives a tiny 
cylinder rolling on a plane. There are many perfectly satisfactory alternative 
suspensions - ball bearing on a plane, crossed cylinders, various types of 
foil or wire suspension.....
       I certainly would NOT mount a CRT monitor, which uses powerful 
magnetic defection coils, anywhere near an unshielded magnet on a seismometer! This 
is just asking for trouble! And every time that you switch a colour monitor on, 
a field coil surrounding the CRT gives a huge magnetic pulse to degauss the 
shadow mask plate..... This is the thunk you hear when you switch on a monitor. 


       Thank you for bringing it to our attention!

       Regards,

       Chris Chapman   
In a me=
ssage dated 2007/03/15, mckimzey@........... writes:



In a previous post I had mused=20=
that the Volksmeter electronics might be 

retrofitted to a Shackleford-Gundersen type assembly.  Well, it would s=
eem that someone in Austria did exactly that.  Here is the link: =20=


http://www3.htl-hl.ac.at/homepage/IP_SEISMO/index_en.php?page=3Dabstract



It looks like a fairly easy setup, but some things I noticed - 1) they have=20=


not seemed to insulate against air currents as much as others with a true SG=
 

sensor.  2) they are using active damping like an SG using data from th=
e 

AD7745.




Hi Mike,



       You need good mechanical and thermal co=
nstruction as well as good electronics if you are to get excellent seismic r=
esults.

       Although this seismometer is housed in=20=
a basement, it's cover is far to large and it is not thermally insulated. No=
 internal electric top heater seems to have been provided to ensure a stable=
 air coulumn. I would expect it to be moderately noisy.

       A much better construction would be to=20=
use a thick wall vertical Aluminum tube to house the pendulum and to pad out=
 any air spaces with open cell plastic foam.

       Putting a unshielded feedback magnet on=
 a seismometer arm is a NO, NO - NEVER !! It will happily pick up noise in t=
he Earth's field, noise from the power wiring of the building and passing tr=
ucks and cars. You need the feedback force to act ~through the centre of gra=
vity, NOT half way up the arm! You can get excellent results from a horizont=
al bar magnet mounted on the frame and a Maxwell feedback coil mounted on th=
e arm near the mass. Electronic velocity force feedback systems are inherent=
ly noisy. If you only want to damp the pendulum, a simple quad magnet + plat=
e system will give excellent performance, without the cost of the electronic=
s.

       There are two suspension types which sh=
ould be avoided like the plague! One is the point in a cup / on a plane and=20=
the other is the metallic knife blade on a flat. Both stress the metal conta=
ct near to or over the deformation stress limit and give a variable and unre=
liable performance. They are highly likely to limit the maximum period which=
 can be attained. The 'so called' knife edge bearings used in chemical balan=
ces are nothing of the sort. They use a 60 degree triangular section with a=20=
lapped tip. This effectively gives a tiny cylinder rolling on a plane. There=
 are many perfectly satisfactory alternative suspensions - ball bearing on a=
 plane, crossed cylinders, various types of foil or wire suspension.....

       I certainly would NOT mount a CRT monit=
or, which uses powerful magnetic defection coils, anywhere near an unshielde=
d magnet on a seismometer! This is just asking for trouble! And every time t=
hat you switch a colour monitor on, a field coil surrounding the CRT gives a=
 huge magnetic pulse to degauss the shadow mask plate..... This is the thunk=
 you hear when you switch on a monitor. 



       Thank you for bringing it to our attent=
ion!



       Regards,



       Chris Chapman  =20