In a message dated 16/11/2008, gel@................. writes:
 
For the  last few months I have been working on a vertical  seismometer.

Hi Gary,
 
    There is a reasonably good vertical design on  John's website at
    _http://jclahr.com/science/psn/hill/index.html_ 
(http://jclahr.com/science/psn/hill/index.html) 
    But I don't like either the suspension knife blade  or the oil damping. 
It is quite easy to use either a twin wire or foil  suspension in tension or a 
twin blade/foil suspension in compression. A bit of  weak spring on the input 
suspension doesn't matterin a vertical. Magnetic  damping is vastly preferable 
to oil. It is quite cheap, easy to adjust and  clean.
    See _http://jclahr.com/science/psn/chapman/as1%20damping/index.html_ 
(http://jclahr.com/science/psn/chapman/as1%20damping/index.html) 
    The longest period you can expect to get from  a steel spring vertical is 
about 6 seconds. The elastic properties of steel are  strongly temperature 
dependant and attempts at longer period suspensions simply  collapse with small 
changes in temperature. 6 seconds is fine if you want to  study microseisms, 
but it is not good for the surface waves from  earthquakes.
    However, you can make a 1.5 to 2 second vertical  quite easily as shown 
above and then extend the period to 15 to 20 seconds  to cover the common Love 
and Rayleigh surface waves at about 20 seconds, using  an extra gain of x100. 
    If you do this digitally, you will need a 16 bit  ADC to give you enough 
signal resolution, with the normal sesimic  background gain set at over 100 
counts.
    If you use an electronic period compensating  amplifier, this needs to be 
low noise, but you can use a 12 bit ADC with it  (just!).
    Sorry, but it is only fair to warn you  that you WON'T get extended 
periods with the vertical compression spring  that you are trying to use. The 
spring + the geometry are both wrong. See  Bob's spring calculations?
 
    Regards,
 
    Chris Chapman







In a message dated 16/11/2008, gel@................. writes:

<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>For the=20
  last few months I have been working on a vertical=20
seismometer.
Hi Gary,
 
    There is a reasonably good vertical design on=20
John's website at
    http://jclahr.com/sci=
ence/psn/hill/index.html
    But I don't like either the suspension knife bl=
ade=20
or the oil damping. It is quite easy to use either a twin wire or foil=20
suspension in tension or a twin blade/foil suspension in compression. A bit=20=
of=20
weak spring on the input suspension doesn't matterin a vertical. Magnetic=20
damping is vastly preferable to oil. It is quite cheap, easy to adjust and=20
clean.
    See http=
://jclahr.com/science/psn/chapman/as1%20damping/index.html
    The longest period you can expect to get f=
rom=20
a steel spring vertical is about 6 seconds. The elastic properties of steel=20=
are=20
strongly temperature dependant and attempts at longer period suspensions sim=
ply=20
collapse with small changes in temperature. 6 seconds is fine if you want to=
=20
study microseisms, but it is not good for the surface waves from=20
earthquakes.
    However, you can make a 1.5 to 2 second vertica=
l=20
quite easily as shown above and then extend the period to 15 to 20 seco=
nds=20
to cover the common Love and Rayleigh surface waves at about 20 seconds, usi=
ng=20
an extra gain of x100. 
    If you do this digitally, you will need a 16 bi=
t=20
ADC to give you enough signal resolution, with the normal sesimic=20
background gain set at over 100 counts.
    If you use an electronic period compensating=20
amplifier, this needs to be low noise, but you can use a 12 bit ADC with it=20
(just!).
    Sorry, but it is only fair to warn you=20
that you WON'T get extended periods with the vertical compression sprin=
g=20
that you are trying to use. The spring + the geometry are both wrong. S=
ee=20
Bob's spring calculations?
 
    Regards,
 
    Chris Chapman