## PSN-L Email List Message

Subject: Re: Surface Waves
From: Bobhelenmcclure@.......
Date: Sun, 3 Jun 2007 14:45:12 EDT

```On Thu, 31 May 2007 11:16:35 EDT, ChrisAtUpw wrote: Where is this capacitor
and what value does it have? The PSN circuit on John Lahr's website has either
a  1 or a 2 uF POLYESTER coupling capacitor into 10 Meg Ohms, giving RC of
either  10 or 20 sec, coupling to 60 or 120 sec periods?<

Hi Randy,

I use an old version of the Cochrane amplifier on my Z channel. It  has a 1
uF polyester coupling capacitor, giving a time constant of 10 seconds.  This
was entirely too short for my needs. I ended up shorting out the capacitor,
which yielded a DC amplifier with a lot of output bias. I was just able to null
out the bias by readjusting the bias pot to its full limit. The resulting
performance is quite satisfactory. There is a little bit of bias response to
ambient temperature rises and falls. I had to add a 1 uf shunt capacitor across
the input terminals to suppress a parasitic oscillation in the amplifier,
invisible at the output, but causing the output bias to be unstable whenever
the  sensor was connected to the amplifier.

If you want to assess the performance of your amplifier, I suggest a  step
function test. Make up a voltage divider consisting of, say, 1 megohm in
series with 1K. Connect the amplifier to the 1K. Apply a 1.5v battery across the
series combination of 1 meg and 1K while recording. You should see a sharp
jump  in ADC output, following by an exponential decay. The time constant of this
decay will give you the RC cutoff period. Removing the battery should yield
another transient of opposite sign. The high frequency gain of the amplifier
can  be calculated from the peak height of the transient, the voltage divider
ratio,  the amplifier input resistance, and the applied voltage.

Bob

************************************** See what's free at http://www.aol.com.

On Thu, 31 May 2007 11:16:35 EDT, ChrisAtUpw wrote: Where is this capac=
itor=20
and what value does it have? The PSN circuit on John Lahr's website has eith=
er a=20
1 or a 2 uF POLYESTER coupling capacitor into 10 Meg Ohms, giving RC of eith=
er=20
10 or 20 sec, coupling to 60 or 120 sec periods?<

Hi Randy,

I use an old version of the Cochrane amplifier on my Z channel.=20=
It=20
has a 1 uF polyester coupling capacitor, giving a time constant of 10 second=
s.=20
This was entirely too short for my needs. I ended up shorting out the capaci=
tor,=20
which yielded a DC amplifier with a lot of output bias. I was just able to n=
ull=20
out the bias by readjusting the bias pot to its full limit. The resulting=20
performance is quite satisfactory. There is a little bit of bias response to=
=20
ambient temperature rises and falls. I had to add a 1 uf shunt capacitor acr=
oss=20
the input terminals to suppress a parasitic oscillation in the amplifier,=20
invisible at the output, but causing the output bias to be unstable whenever=
the=20
sensor was connected to the amplifier.

If you want to assess the performance of your amplifier, I sugge=
st a=20
step function test. Make up a voltage divider consisting of, say, 1 megohm i=
n=20
series with 1K. Connect the amplifier to the 1K. Apply a 1.5v battery across=
the=20
series combination of 1 meg and 1K while recording. You should see a sharp j=
ump=20
in ADC output, following by an exponential decay. The time constant of this=20
decay will give you the RC cutoff period. Removing the battery should yield=20
another transient of opposite sign. The high frequency gain of the amplifier=
can=20
be calculated from the peak height of the transient, the voltage divider rat=
io,=20
the amplifier input resistance, and the applied voltage.

BobSee wh=
at's free at AOL.c=
om.
```