## PSN-L Email List Message

Subject: Re: New Lehman on line (almost)
From: "Randall Pratt" randallpratts@..........
Date: Tue, 1 Oct 2002 22:14:14 -0500

```Allen,

Have you used the method of calibration you referenced?  It is very easy =
to set up but my system does not behave quite as advertised.  With the =
boom blocked I don't get a step function as in fig 4.5.1a but rather an =
exponential decay.  I find that a bit confusing since I put a steady =
battery voltage across the coil but I have attempted to determine the =
curve and adjust subsequent readings by the correct factor over time.  =
I'm also not clear about para 9 where a0 is computed.  What does that =
formula really mean?  How would it be adjusted for swings later in the =
wave train and what is the ' on the end?  Why would later pairs of =
values work when there is a log decay in the swings?

Randy=20
----- Original Message -----=20
From: ACole65464@..........
To: psn-l@.................
Sent: Monday, September 30, 2002 9:04 PM
Subject: Re: New Lehman on line (almost)

In a message dated 10/01/2002 12:14:06 AM !!!First Boot!!!, =
ChrisAtUpw@....... writes:

In a message dated 30/09/02, shammon1@............. writes:=20

The standard rule is to pull the boom back a few inches and let it =
go. The boom=20
should loose 30% of its motion on each swing past center and come =
to rest=20
in 3 1/2 swings.

Hi Steve,=20

I am puzzled as to where this *standard rule* is supposed to =
come from? But using it will give you a quite seriously underdamped =
system! A critically damped system experiences no oscillation at all. =
This is inherent in the maths.=20
This is important if you apply post processing to the recorded =
signal with the assumption that it was critically damped to start with. =
It will also give problems with the amplitudes and frequencies =
calculated in FFT displays and may 'smear' P and S wave recordings.=20
A procedure to get critical damping could involve deflecting =
the beam a very small amount (microns) and recording the amplifier =
output. You progressively increase the damping until the arm just =
returns to the balance position without having crossed the zero line. If =
you increase the damping further, the arm will simply take longer to get =
back to zero. If you use huge deflections like a few inches, you are =
likely to encounter non linear effects which do not apply to the tiny =
(hopefully!) signals that we normally record.  =20
It is helpful if the recording displays just what the earth is =
doing. It is really not helpful if the system adds an oscillating tail =
to every transient.=20

Regards,=20

Chris Chapman=20

Steve,=20

In support of what Chris has stated, please go to:  =
http://www.seismo.com/msop/msop79/inst/inst4.html#aa250  Go to section =
4.5 for a text description, and then click on figure 4.5.1a to see how =
pendulums are supposed to be damped. About Critical is the response you =
should obtain. I hope this helps a little, the diagrams may not make =
much sense at first but it shows how professional instruments =
(electromagnetic, aka Lehman designs) are adjusted.

Regards,

Allen,

Have you used the method of calibration =
you=20
referenced?  It is very easy to set up but my system does not =
behave quite=20
as advertised.  With the boom blocked I don't get a step function =
as in fig=20
4.5.1a but rather an exponential decay.  I find that a bit =
confusing since=20
I put a steady battery voltage across the coil but I have attempted to =
determine=20
the curve and adjust subsequent readings by the correct factor over =
time. =20
I'm also not clear about para 9 where a0 is computed.  What does =
that=20
formula really mean?  How would it be adjusted for swings =
later in the=20
wave train and what is the ' on the end?  Why would later pairs of =
values=20
work when there is a log decay in the swings?

Randy

----- Original Message -----
From:=20
ACole65464@.......
To: psn-l@..............
Sent: Monday, September 30, =
2002 9:04=20
PM
Subject: Re: New Lehman on line =

(almost)
In a =
message dated=20
10/01/2002 12:14:06 AM !!!First Boot!!!, ChrisAtUpw@....... =
writes:
In a message dated 30/09/02, shammon1@............. =
writes:=20

The standard rule is to pull the boom back a few =
inches and=20
let it go. The boom should loose 30% of its motion on each =
swing past=20
center and come to rest in 3 1/2 swings.Hi Steve, =
I am=20
puzzled as to where this *standard rule* is supposed to come from? =
But using=20
it will give you a quite seriously underdamped system! A=20
critically damped system experiences no oscillation at =
all.=20
This is inherent in the maths.       =
This is=20
important if you apply post processing to the recorded signal with =
the=20
assumption that it was critically damped to start with. It will also =
give=20
problems with the amplitudes and frequencies calculated in FFT =
displays and=20
may 'smear' P and S wave recordings. =
A=20
procedure to get critical damping could involve deflecting the beam =
a=20
very small amount (microns) and recording the amplifier=20
output. You progressively increase the damping until the arm =
just=20
returns to the balance position without having crossed the zero =
line. If you=20
increase the damping further, the arm will simply take longer to get =
back to=20
zero. If you use huge deflections like a few inches, you are likely =
to=20
encounter non linear effects which do not apply to the tiny =
(hopefully!)=20
signals that we normally record.  =20
It is helpful if the recording =
displays=20
just what the earth is doing. It is really not helpful if the system =
oscillating tail to every transient. =
=20
Regards,       Chris =
Chapman Steve, In support of what Chris has =
stated,=20
http://www.seismo.com/msop/msop79/inst/inst4.html#aa250  Go to =
section=20
4.5 for a text description, and then click on figure 4.5.1a to see how =

pendulums are supposed to be damped. About Critical is the response =
you should=20
obtain. I hope this helps a little, the diagrams may not make much =
sense at=20
first but it shows how professional instruments (electromagnetic, aka =
Lehman=20