Hi John,

       I suggest that you have someone at the Colorado School of Mines=20
Project download all the PSN letters and put them into annual files. They ca=
n=20
then be searched with 'Find' from the WP programme using keywords.

       Following Larry's letter, I checked on a base for seismometers. I=20
found that I could buy an 18" hexagonal concrete slab, some 1/2" square stee=
l=20
bar and three socket cap screws for less than $10. Other suggestions include=
d=20
stone / granite slabs which had defects preventing them from being saleable,=
=20
armoured glass from when banks updated their security systems, metal disks e=
x=20
industrial processes, offcuts from metal suppliers, various offerings from=20
scrap metal merchants, polyester fibreglass + sand mouldings, melamine=20
covered high density chipboard offcuts from kitchen worktops and laminated,=20
varnished hardwood. My own favourite is a 'puddle it yourself' cement and=20
sand mix, so that you can mould in all the necessary holes for the support=20
and mounting bolts.=20

       I also looked at circuit board options. Pin board and punched=20
stripboard are not expensive and neither are the components. If you are daft=
=20
enough to specify DIP board for general purpose printed circuits, you do get=
=20
into the idiot price bracket.=20


> Sensitivity
> For the system to be useful in any location on the Earth, it must be able=20
> to record large earthquakes at any distance -- (teleseisms).  This require=
s=20
> a good signal to noise ratio in the period range from 1 to 20 seconds.
> The following table gives a rough idea of the amplitude and velocity of=20
> ground motion that would be expected from a magnitude 7 earthquake at 90=20
> and 180 degrees distance.
>=20
> Magnitude 7 at 90 degrees distance:           amplitude        velocity
> Expected 20s period surface wave amplitude [  5.61E+01 =B5m]  [  1.76E+01
> =B5m/s]
> Expected 1s period body wave amplitude     [  7.93E-01 =B5m]  [  4.98E+00
> =B5m/s]
>=20
> Magnitude 7 at 180 degrees distance:
> Expected 20s period surface wave amplitude [  1.81E+01 =B5m]  [  5.68E+00
> =B5m/s]
>=20

       This should not be difficult.

> Sensor
>=20
> The sensor could be based on the voltage generated by relative motion
>=20

       OK.

>=20

       OK.

on the amount of=20
>=20

       OK.

>=20

       OK for down to 0.1 Hz, but the longer times may be a problem.=20

> machined accelerometer similar to those used to=20
>=20

       These are very noisy.

       Two more systems both based on the NE5521N chip, which can obtain nm=20
resolution.

       LVDT sensors. These can be made for ranges up to +/- 10 mm
       LCDT sensors. Better for ranges <+/-1 mm and FFB seismometers

> Given that new, high-density, low-cost disk drives are continually coming
> to market, there may well be some sophisticated, mass-produced, and yet=20
> inexpensive components that could also be used for this application.

       I have not seen the 100 MB zip drives mentioned?

       Regards,

       Chris Chapman=20

Hi John,



       I suggest that you have someone at=20=
the Colorado School of Mines Project download all the PSN letters and put th=
em into annual files. They can then be searched with 'Find' from the WP prog=
ramme using keywords.



       Following Larry's letter, I checked=
 on a base for seismometers. I found that I could buy an 18" hexagonal concr=
ete slab, some 1/2" square steel bar and three socket cap screws for less th=
an $10. Other suggestions included stone / granite slabs which had defects p=
reventing them from being saleable, armoured glass from when banks updated t=
heir security systems, metal disks ex industrial processes, offcuts from met=
al suppliers, various offerings from scrap metal merchants, polyester fibreg=
lass + sand mouldings, melamine covered high density chipboard offcuts from=20=
kitchen worktops and laminated, varnished hardwood. My own favourite is a 'p=
uddle it yourself' cement and sand mix, so that you can mould in all the nec=
essary holes for the support and mounting bolts.=20



       I also looked at circuit board opti=
ons. Pin board and punched stripboard are not expensive and neither are the=20=
components. If you are daft enough to specify DIP board for general purpose=20=
printed circuits, you do get into the idiot price bracket.=20





Sensitivity

For the system to be useful in any location on the Earth, it must be abl=
e to record large earthquakes at any distance -- (teleseisms).  This re=
quires a good signal to noise ratio in the period range from 1 to 20 seconds=
.

The following table gives a rough idea of the amplitude and velocity of=20=
ground motion that would be expected from a magnitude 7 earthquake at 90 and=
 180 degrees distance.



Magnitude 7 at 90 degrees distance:       =
    amplitude        =
velocity

Expected 20s period surface wave amplitude [  5.61E+01 =B5m]  =
[  1.76E+01

=B5m/s]

Expected 1s period body wave amplitude     [  7=
..93E-01 =B5m]  [  4.98E+00

=B5m/s]



Magnitude 7 at 180 degrees distance:

Expected 20s period surface wave amplitude [  1.81E+01 =B5m]  =
[  5.68E+00

=B5m/s]

(No direct P-phase at this distance.)



       This should not be difficult.



Sensor



The sensor could be based on the voltage generated by relative motion

between a coil and magnet,



       OK.



on the proximity of a magne=
t to a Hall-effect sensor, 



       OK.



on the amount of=20

light reaching a photo diod=
e that is partially blocked by a flag, 



       OK.



on the strain of a piezoelectric 
crystal, 



       OK for down to 0.1 Hz, but the long=
er times may be a problem.=20



on the output of a micro 
ma=
chined accelerometer similar to those used to trigger the release of an auto=
mobile's air bag.



       These are very noisy.



       Two more systems both based on the=20=
NE5521N chip, which can obtain nm resolution.



       LVDT sensors. These can be made for=
 ranges up to +/- 10 mm

       LCDT sensors. Better for ranges <=
;+/-1 mm and FFB seismometers



Given that new, high-densit=
y, low-cost disk drives are continually coming

to market, there may well be some sophisticated, mass-produced, and yet=20=
inexpensive components that could also be used for this application.<=
FONT  COLOR=3D"#000000" SIZE=3D3 FAMILY=3D"SANSSERIF" FACE=3D"Arial" LANG=
=3D"0">



       I have not seen the 100 MB zip driv=
es mentioned?



       Regards,



       Chris Chapman=20