PSN-L Email List Message

Subject: Re: Filters
From: ChrisAtUpw@.......
Date: Sun, 7 Jan 2007 22:19:51 EST


In a message dated 07/01/2007, tchannel@.............. writes:

Hi Folks,  One issue where I seem to need  help, is how to filter. I have 
keep notes on the Low and High Pass filter  values used by others.
I also know that the filter values will change  with all the variables.  Let 
us assume, using a vertical similar to the  AS1, what would be good starting 
points for Local, Regional, and Teleseismic  events?  

Hi Ted,
 
    If you are using an AS-1, the sample rate is set to  6 sps, so the max 
frequency that can be detected is 3 Hz. You have no  choice.
 
    However, if you are using a Lehman, you might wish  to set the low pass 
at 20 Hz for Local / Volcanic events, 10 Hz for Regional  events or 5 Hz for 
Teleseismic events. The high pass filter should be set at or  a bit above the 
set period of the sensor - lower in frequency.
    But you are usually presented with an amplifier  with built in frequency 
/ period values. The filters are used to remove man  made / environmental 
noise, which is usually serious at frequencies of 20  Hz and above and VLF 1/f 
noise. 

Also could someone explain what the poles  numbers do?


    The corner frequency is set by the capacitor  values. Increasing the 
number of poles (stages) increases the cut-off slope by 6  dB per octave per pole 
above the corner frequency and this also requires  an increase in the number 
of opamps. It is usual  to assign two poles (RC stages) per opamp. However, 
each stage adds to  the signal delay through the filter. A five or six stage 
filter is usually  adequate. 
    Bessel filters have a constant delay over the  passband and this figure 
may be entered into the SDR program to compensate. The  signal 'shape' is 
fairly well maintained. Butterworth filters have a large  peak delay at the edge of 
the passband and the signal 'shape' may be seriously  distorted if any 
frequency component is near the edge of the passband.
 
    Regards,
 
    Chris Chapman





In a message dated 07/01/2007, tchannel@.............. writes:
<= FONT=20 style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size= =3D2>
Hi Folks,  One issue where I seem to= need=20 help, is how to filter. I have keep notes on the Low and High Pass filter=20 values used by others.
I also know that the filter values will c= hange=20 with all the variables.  Let us assume, using a vertical similar to t= he=20 AS1, what would be good starting points for Local, Regional, and Teleseism= ic=20 events?  
Hi Ted,
 
    If you are using an AS-1, the sample rate is se= t to=20 6 sps, so the max frequency that can be detected is 3 Hz. You have no=20 choice.
 
    However, if you are using a Lehman, you might w= ish=20 to set the low pass at 20 Hz for Local / Volcanic events, 10 Hz for Regional= =20 events or 5 Hz for Teleseismic events. The high pass filter should be set at= or=20 a bit above the set period of the sensor - lower in frequency.
    But you are usually presented with an amplifier= =20 with built in frequency / period values. The filters are used to remove= man=20 made / environmental noise, which is usually serious at frequencies of = 20=20 Hz and above and VLF 1/f noise. 
Also could someone explain what the p= oles=20 numbers do?
    The corner frequency is set by the capacitor=20 values. Increasing the number of poles (stages) increases the cut-off slope=20= by 6=20 dB per octave per pole above the corner frequency and this also requires=20 an increase in the number of opamps. It is usual=20 to assign two poles (RC stages) per opamp. However, each stage add= s to=20 the signal delay through the filter. A five or six stage filter is usually=20 adequate.
    Bessel filters have a constant delay over the=20 passband and this figure may be entered into the SDR program to compensate.=20= The=20 signal 'shape' is fairly well maintained. Butterworth filters have a la= rge=20 peak delay at the edge of the passband and the signal 'shape' may be serious= ly=20 distorted if any frequency component is near the edge of the passband.
 
    Regards,
 
    Chris Chapman

[ Top ] [ Back ] [ Home Page ]