Hi Geoff,

  Heavy damping is not the solution. You should adjust damping for a Q
of around 0.71, which gives the flattest response over the sensor
passband. From there, you have several options. You can use an analog
low pass filter as you suggest, but a true Butterworth design cannot
be realized without an inductive element in the circuit. A 10-second
cutoff second order low-pass filter applied to the HS-10 will yield a
flat response in the 1 to 10 second period range, but will result in
loss of signals higher than 1 Hz and a severe loss of overall
amplitude.

  The preferred analog solution is the Roberts circuit used by Chris
Chapman. It compensates for the roll-off in response beyond the
natural frequency and preserve high frequency response. You would be
better off abandoning your balanced input scheme. Use single-ended
signals. It will simplify implementation of the Roberts circuit.

  I have a digital solution, an inverse filter implemented in my
WQFilter utility program. Its gain curve compensates for the loss of
response for periods beyond the natural frequency, and can compensate
for over or under damping. It does nothing to improve whatever
signal-to-noise ratio you record at long periods, however. WQFilter
operates on WinQuake PSN Type 4 files. I know that you work with PSN
text files, but WinQuake will convert these to PSN Type 4. You can
download WQFilter from

  http://bobmcclure90.googlepages.com/winsac

  WQFilter has benefits for everyone with amateur sensors. I,
personally, would be lost without it. I recommend that all amateurs
look up my web site, read all about WQFilter, and give it a try.

  I also urge everyone who uses WinQuake to download the latest version from

  http://www.seismicnet.com/software.html#WinQuake

  Those who submit multiple event files should learn how to use
WinQuake to make volume files, as well.

---Bob

On 2 Jun 2009, Geoff wrote:
I am currently using heavy damping to
help flatten my response because
my compensation circuit filters
like a LPF with a curve at like
10 seconds.

It seems if I amplify 1Hz to 80Dbv (x10000) I can
do almost anything I want and still get a
signal. Flatness of response is what I'm
looking for.

I am using a 470 ohm damping resistor
Split 50/50 between ground and +/- differential inputs.
for a single 450 ohm geophone.
This produces a current loop between the
geophone and amp that should override
the weaker noise.

I run the geophone coil straight into the diff amp
with no resistors in series.

I expect this to reduce overall gain by 1/2
plus geophone damping effect of raising
the point of roll off to over 1 Hz.

Is it possible to tell what the new
geophone curve is with such heavy damping ?
====
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)