In a message dated 2006/10/17, tchannel@.............. writes:

> I have finished my sensor, and have had it running for three days.

Hi Ted,

       Sorry, but you haven't finished your sensor until you have evaluated 
and proved it!!

> 1.  During the night hours, 11pm to 6am, the tract goes from normal to 
> noisy, getting worse each hour. The sensor sits in the garage on a concrete slab. 
> No draft and no heat.  

       This is the evening cooling problem due to air convection. You need to 
house the seismometer in a thermally insulated case and place a heater giving 
about 10 Watts inside the top. I use two Al cased power resistors bolted to a 
sheet of Al. This maintains a positive vertical internal temperature gradient 
at all times and suppresses convection. Make the enclosure out of 2" Celotex. 
This has fibre glass skins and Al foil top coats. Tape the joins with 3" 
gaffer tape and glue the joints with foam grouting. You can make double windows 
with either photocopy film or cellulose sheet. See 
http://pages.prodigy.net/fxc/seiscoverslg.jpg More work, but 100% essential!

> 2.   I think I am getting a good recording, I have see the last two quakes 
> and to me they look normal.
> I am in Boise Idaho and have tried to compare my .PSN with any others close 
> by.  However our local one in Boise is down and the next, nearest one is in 
> Washington State.  Now here is a hard question to ask: I created my expanded 
> trace, using the same time plotted on the .PSN in Washington (
> 061017.013725.ebgz.psn M6.5 10356km from Ellensburg, Wa.,) This was around one hour of trace. 
>  I did this so I could compare the two side by side.  I was a little 
> surprised to see the Wa. psn to show a large spike on the P and a short spike on the 
> S.  My trace looks very similar, but I show a large spike on the S and a 
> smaller spike on the P. From the description of the Ellensburg Wa file it has a 
> Z in it, is that a Vertical sensor? and since my sensor is a Horz pointing 
> N/S, would that explain the different. 

       Z signifies a vertical sensor. The two other designations are E east 
and N north, both true, not magnetic. The P and Rayleigh waves are 
predominantly vertical, so a Lehman, which is a horizontal sensor, will have a low 
sensitivity to them. The S and Love waves are horizontal and you should see 
relatively large signals. The amplitude of the signals also depends both on the type 
and on the angular range.
   
>  I would appreciate someone looking at them and voicing an opinion as to my 
> level of gain, not sure if it is correct, and some advise as to how to best 
> filter this event to enhance the P and S. 

       I suggest that you set the lowpass filter at 2 Hz and the highpass 
filter at 0.4 Hz, both at least 4 pole.

> 3.  Presently I am building a Plexiglas box. 

       See my previous EMail and the advice above. You need a thermally 
insulated enclosure with sealed joints and a top heater, probably not a perspex 
box.

> 4. I know weather can alter the way the trace looks, I thought my situation 
> was weather, but it was clear today.
> What is there in a front moving thru that creates the noise?  Lets say a 
> front with little wind and no lightning.

       Cold fronts tend to give more seismic noise that warm fronts. All 
fronts have wind and turbulence / pressure variations associated with them, at 
altitude, if not so much at ground level. Broad band wind noise is a common 
seismic problem. 

       Regards,

       Chris Chapman
In a message dated 2006/10/17, tcha=
nnel@.............. writes:



I have finished my sensor, and=20=
have had it running for three days.




Hi Ted,



       Sorry, but you haven't finished your se=
nsor until you have evaluated and proved it!!



1.  During the night hour=
s, 11pm to 6am, the tract goes from normal to noisy, getting worse each hour=
.. The sensor sits in the garage on a concrete slab. No draft and no heat.&nb=
sp; 




       This is the evening cooling problem du=
e to air convection. You need to house the seismometer in a thermally insula=
ted case and place a heater giving about 10 Watts inside the top. I use two=20=
Al cased power resistors bolted to a sheet of Al. This maintains a positive=20=
vertical internal temperature gradient at all times and suppresses convectio=
n. Make the enclosure out of 2" Celotex. This has fibre glass skins and Al f=
oil top coats. Tape the joins with 3" gaffer tape and glue the joints with f=
oam grouting. You can make double windows with either photocopy film or cell=
ulose sheet. See http://pages.prodigy.net/fxc/seiscoverslg.jpg More work, bu=
t 100% essential!



2.   I think I am get=
ting a good recording, I have see the last two quakes and to me they look no=
rmal.

I am in Boise Idaho and have tried to compare my .PSN with any others close=
 by.  However our local one in Boise is down and the next, nearest one=20=
is in Washington State.  Now here is a hard question to ask: I created=20=
my expanded trace, using the same time plotted on the .PSN in Washington (061017.013725.ebgz.psn M6.5 10356km from Ellensburg, Wa.,) This was=20=
around one hour of trace.  I did this so I could compare the two side b=
y side.  I was a little surprised to see the Wa. psn to show a large sp=
ike on the P and a short spike on the S.  My trace looks very similar,=20=
but I show a large spike on the S and a smaller spike on the P. From the des=
cription of the Ellensburg Wa file it has a Z in it, is that a Vertical sens=
or? and since my sensor is a Horz pointing N/S, would that explain the diffe=
rent. 




       Z signifies a vertical sensor. The two=
 other designations are E east and N north, both true, not magnetic. The P a=
nd Rayleigh waves are predominantly vertical, so a Lehman, which is a horizo=
ntal sensor, will have a low sensitivity to them. The S and Love waves are h=
orizontal and you should see relatively large signals. The amplitude of the=20=
signals also depends both on the type and on the angular range.

   

 I would appreciate some=
one looking at them and voicing an opinion as to my level of gain, not sure=20=
if it is correct, and some advise as to how to best filter this event to enh=
ance the P and S. 




       I suggest that you set the lowpass fil=
ter at 2 Hz and the highpass filter at 0.4 Hz, both at least 4 pole.<=
FONT  COLOR=3D"#000000" BACK=3D"#ffffff" style=3D"BACKGROUND-COLOR: #ffffff"=
 SIZE=3D3 PTSIZE=3D12 FAMILY=3D"SERIF" FACE=3D"Times New Roman" LANG=3D"0"><=
BR>


3.  Presently I am buildi=
ng a Plexiglas box. 




       See my previous EMail and the advice a=
bove. You need a thermally insulated enclosure with sealed joints and a top=20=
heater, probably not a perspex box.



4. I know weather can alter the=
 way the trace looks, I thought my situation was weather, but it was clear t=
oday.

What is there in a front moving thru that creates the noise?  Lets say=
 a front with little wind and no lightning.




       Cold fronts tend to give more seismic n=
oise that warm fronts. All fronts have wind and turbulence / pressure variat=
ions associated with them, at altitude, if not so much at ground level. Broa=
d band wind noise is a common seismic problem. 



       Regards,



       Chris Chapman