PSN-L Email List Message

Subject: Re: Period
From: "Geoffrey" gmvoeth@...........
Date: Tue, 3 Jul 2007 03:44:57 -0700

Hello tchannel1;

This is a response to your post and intended
for everyone at PSN.

A period range of 0.2 Hz to 2Hz is by far the best
range for seeing P and S of most EQ signals
If one lives in a noisy area with lots of vehicular
traffic this is the best range to use. In this range you
should be able to crank up the gain until you see
both low and high frq noise about +/- two samples
maybe four then you should be able to see a 4.0
regional quake at 6 degrees of distance as the
guideline for a MDL (Minimum discernable level).
The smaller the bandwidth the less the noise.
If you are in a quiet area by all means open up the bandwidth
to see the close quakes better.
If you have like a 24 bit A/D converter you can make the gain
adjustable simply be looking at 8 or 12 or 16 bits and scaling
everything else then you can digitally set the gain from
the keyboard.

If you program a modulo recorder of 12 minutes or so
there are almost no earthquakes of any size
that you will miss with a computer oriented recorder
and a seismic alarm that works something like a venus fly trap
trigger hairs.

Do you know if anyone has ever used a laser pointer
in a seismic sensor ? Possibly using surfaced mirrors
to provide mechanical amplification like a lever arm ?

If you do not build your own electronics and program your own
software then most probably forget these finer details.
Like Always, I am no expert but i do have several broken
years of practical experience.

To me an Amateur can be a layman who simply watches the signals
and cares nothing more about the science since an amateur receives
No pay for his hobby. Laymen Professional Expert are different ideas.
I think you can find all three in any amateur group. I am an Amateur layman
with some formal technical education and experience but no degree\certificates
 that means anything.
The range I tell you here is from personal experience over the past 13 years.
I have found the 500LB geophone used all over the world by the USGS gives the
best signal right down to 20 seconds even tho it is a 1sec free period device.
It does not have the severe natural damping of the smaller geophones
but does have a very powerful magnet.

My thought is this;
The sensor should be in a vacuum and have a very powerful magnet
and in a faraday shield otherwise the sensor will typically be marginal at best.
Everything needs to be shielded and bonded between the Computer including the sensor.
AntiAlising filtration by whatever means prior to conversion/recording by the A/D converter.
Do not use the sensor coil for damping because that reduces sensitivity but instead
use a seperate magnet/Copper Plate arrangement. The preamplifier to be a
differential type with both +/- of the opamp equalized for DC.
Put as little load on the sensor as possible.
It used to be easy to get copper in Arizona but not any more
unless you pay lots of cash, copper has become outrageously expensive
along with everything else.
It seems to me they (My Country/USA) do not want people playing
around with Science and Technology unless they are rich or
Formally educated. I think they (My USA) are overly Religious and overly paranoid.
The majority of people here are not well enough educated (including myself).
It is humbling to note that after study of life that intelligence/education are not necessary for survival.

Winquake is the best public seismic program I have seen for the layman.
I will convert my data to text to use in winquake and find the data
easier to analyze. But to be most precise and accurate I will read
using my own program on a sample by sample bases.
Winquake gives a sort of standard for PSN and lets people
share their data. But I have found that if you are converting
your own data into PSN form you must start the first sample
relating to an exact start of minute or the times will not be right.
In Winquake You must test your sample rate and make sure its right.
Start all your data for winquake at HH:MM:00.0 the seconds always
being zero for your very first sample. You need cal markes from
WWV/CHU or something like that to see the top of second zero
in each minute.
Say Minute 59-60 second mark goes high at sample #1025
use sample #1024 as the very first sample and stop the samples
at the N= The next cal mark high - one sample at the end. Winquake
should agree that the times are all right so long as you got the
sample rate right.
Winquake does not give me good times if I do not follow this idea.
I can not just start out in the middle of a minute and end just anywhere.
Without Cal marks of some standard type the times for the layman are uncertain.

Winquake needs a FFT waterfall display to see a movie like of
shifting frequencies over time.
Possibly in 1024 sample chunks advancing one sample for the next 1024.
Nothing fancy, Black and white and shades of gray,  totally suitable.


Public Seismic Network Mailing List (PSN-L)

[ Top ] [ Back ] [ Home Page ]