PSN-L Email List Message

Subject: low gain VBB
From: S-T Morrissey sean@...........
Date: Tue, 2 Nov 1999 13:08:35 -0600 (CST)


Barry,

In high-gain/low-gain recording it is desirable that the
signals be identical for purposes of interpretation.

The VBB system has a dynamic range of over 120db, or 10^6:1,
which is why it needs 24-bit digitizing for a single gain level.
For the affordable world of 16-bit (~10,000:1) recording, a
low gain channel of 1/100 the high gain will realize most
of the dynamic range while still recording something once in
a while. A lower gain "low-gain", like 1/1000 will give more
of a total range approaching 24-bits (16x10^6:1), but will 
very rarely record much. If the VBB signal is typically 2000 
volts/meter/second, 1/1000 th of it is 2 v/m/s. So if the 
digitizer is +,- 10 volts, it will clip at a velocity of
5 meters/second. In acceleration this is about 0.8m/sec^2,
or about 0.13 of g, at 1 second (hz ).

Modifying the feedback parameters of the VBB is not a good idea.
THey exactly describe your response at the VBB output.

The way I realize a low gain signal is to attenuate the output
after the final line-driver amplifier. This is a 5x to 20x gain amp
after the 1000-second high pass filter. I use precision resistors
with the total as low as practical without loading the op-amp.
Like a total of 10k ohms made with 9.9k and 100 ohms for a 100:1
ratio. If your high-pass filter follows your line amplifier, you
will need to buffer it with another amplifier. However, I would
place the filter before the line-driver so you can have identical
high and low gain signals. This also avoids the added noise of
stringing together many amplifiers: the fewer the better. (the VBB
only needs three).
With a 5x gain, an output of 10 000 v/m/s will clip (at 10 volts) at
1 millimeter/second. If the amplifier noise floor is 10 microvolts,
the equalivent noise would be 10 nanometers/second.

Somewhere I have data on earthquake magnitude/distance figures
in terms of the velocity output of a seismometer.

Regards,
Sean-THomas

_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)


[ Top ] [ Back ] [ Home Page ]

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