PSN-L Email List Message

Subject: VBB response, contd.
From: S-T Morrissey sean@...........
Date: Wed, 26 Jan 2000 14:50:06 -0600 (CST)

Allan and co:

Some questions about the VBB response:

If you are looking at the response by using MATHCAD to calculate
the transfer function and plot the response, you should get exactly
what I get. This assumes that you are using exactly the same formulations,
constants, etc, as the worksheets I have shown. Have you noted that
the frequency (or period) variable is a complex number, ie s(i) = j*w(i),
where w is omega? The result you get should not be "approximate",
but exactly the same as I get with the same inputs.

The other question that makes me suspect that you are not getting
the proper response is that you ask about the "high end" response:
the VBB response IS the total response, as is shown by the MATHCAD
plot. The VBB response is flat to velocity over a wide range of periods,
not just the long periods. If you change the integrator time constant
and/or the feedback capacitor the long period corner can be changed.
The short period corner frequency is largely determined by "r", the 
output of the displacement detector. Larger is better until the 
instabilities show up. I have one VBB that audibly humms until I
strap in a RC filter into the displacement detector output. (that most
likely produces a proper phase shift so the oscillation doesn't start).

I don't know of anyone else who has had problems with the transfer
function and MATHCAD worksheet expressions. They do work.
The only way I can address the question about the resulting response is 
to point to the calibration data on my web site, where the independent 
calibration data overlay the response predicted by the transfer function.
What you see is what you get.

In fact, the broadband response at the short period/high frequency
end can cause problems with local site noise. My site here in the
basement is about 15 meters from a busy street, so I have considerable
low-pass filtering to reduce the cultural noise, as well as that of
the dogs romping through the house; however, I still pick up a railroad
about 3 km away. And I can record a Mb 3.+ at New Madrid 300 km away,
as well as the long period data that I am more interested in. Since
I am using a vertical sensor, the long period noise is factors of 100
less than the tilt noise a horizontal LP would be subject to. Which 
is why I wish that more PSN people would build vertical long period 



Public Seismic Network Mailing List (PSN-L)

[ Top ] [ Back ] [ Home Page ]

Larry Cochrane <cochrane@..............>