PSN-L Email List Message
Subject: Re: Instrumentation Question
From: chrisatupw@.......
Date: Thu, 27 Dec 2012 09:57:18 -0500 (EST)
From: Dave Nelson
To: psnlist
Sent: Mon, 24 Dec 2012 18:23
Subject: Re: Instrumentation Question
DC drift is very common --almost universal -- in seismometers. Even the=20
best broadband instruments have a drift proportional to the rate of=20
temperature change. The frequency content of the drift are usually well=20
below the seismic information so is of no consequence.
A 24 hour periodic drift is of absolutely no consequence for a geophone. Fo=
r=20
a 1 second instrument, even with period extension, the longest practical=20
seismic signal which can be observed is probably around 20 seconds. The=20
noise level at 20 seconds, after period extension, will be so high that onl=
y=20
very large events will be observable above the instrument/system noise. A=
=20
high pass filter in Winsdr at .05 Hz 1 pole or AC coupling with a time=20
constant of about 3 seconds (20 seconds / 2pi) would be about right.
Cheers,
Dave Nelson
Rolling Hills Estates,=20
Hi Dave,
****The noise that you see from a period extension circuit depends on the o=
pamps used and on the=20
circuit chosen.
Period Extensions of up to x10 can be achieved using the Roberts' Circuit. =
The internal gain is x100.=20
For Period Extensions of up to x50, you need the Lippmann circuit and a VER=
Y low noise amplifier.=20
The output short circuit current is very low. =20
As far as I remember, Geoff is trying to use a Lippmann type circuit, so he=
should get reasonably low=20
noise signals at 20 seconds.=20
****Brett Nordgren wrote ;
Even professional instruments show substantial DC drift. Broadband=20
verticals have an output proportional to the rate of temperature=20
change and over a 24-hour period can easily vary by a volt or more=20
depending on their thermal insulation. We routinely put in a=20
one-pole digital high-pass filter at the lowest frequency available=20
(0.002 Hz in WinSDR) which makes all that go away.
****An offset of over a volt could exceed the input range of some ADCs !!=
=20
Such poor quality electronics would be totally unacceptable in my opinion.=
=20
>> Subject: Instrumentation Question
>>
>> Yet I can not rid a small DC drift which is either related to
>> ambient AC noise level or DC drift of resistance or
>> thermocouple voltage related to the soldering of junctions.
>> This DC drift is related to the 24 hour cycle.
>> The dc drift is on the order of micro volts
>> which seem to be originating on the input.
>> The overall DC gain is 80Dbv or X10000.
>> I should be able to achieve a DC free drift at this gain.
>> I am using a op177G op amp ??
>> The sensor is HS10-1 Geophone.
>> The 40 foot of cable is designed for burial
>> it has silicon grease (I think) impregnated and
>> has a heavy copper jacket.
>> It is soldered at the geophone and also at the sensor
>> has soldered connections.
>
> If you use a CAZ type opamp like the LTC1150 you will get zero=20
> temperature drift.
>
> The problem does not seem to be the amplifier itself.
> But rather, the variables dealing with the input.
> Cable/Geophone/Common mode stuff/
> Yet I have found a wide variation within the
> op amp offset voltages.
> I am currently using a Chopper amp (or so I think)
> In the front end. LTC1050 ???
> Yes, I have troubles finding decent opamps
> at civilian prices, the best are all seem to be MilSpec
> rip offs. Outrageous prices. The milspec parts are by far
> the best and all others should simply be scrapped.
>
> Regards,
> geoff
> LTC1050 - Precision Zero-Drift Operational Amplifier with Internal=20
> Capacitors
***So is the LTC1150 ! Check it out ? !=20
Regards,=20
Chris Chapman
=20
=20
=20
=20
From: Dave Nelson <davef=
nelson@.......>
To: psnlist <psnlist@..............>
Sent: Mon, 24 Dec 2012 18:23
Subject: Re: Instrumentation Question
DC drift is very common -=
-almost universal -- in seismometers. Even the=20
best broadband instruments have a drift proportional to the rate of=20
temperature change. The frequency content of the drift are=
usually well=20
below the seismic information so is of no consequence.
A 24 hour periodic drift is of absolutely no consequence for a geophone. Fo=
r=20
a 1 second instrument, even with period extension, the longest practical=20
seismic signal which can be observed is probably around 20 seconds. The=20
noise level at 20 seconds, after period extension, will be so high that onl=
y=20
very large events will =
be observable above the instrument/system noise. A=20
high pass filter in Winsdr at .05 Hz 1 pole or AC coupling with a time=20
constant of about 3 seconds (20 seconds / 2pi) would be about right.
Cheers,
Dave Nelson
Rolling Hills Estates,=20
Hi Dave,
****The noise that you see=
from a period extension circuit depends on the opamps use=
d and on the
circuit chosen.
Period Extensions of up =
to x10 can be achieved using the Roberts' Circuit. The internal gain is x10=
0.
For Period Extensions of up to x50, you need the Lippmann =
circuit and a VERY low noise amplifier.
The output short circuit current is very low. <=
br>
As far as I remember, Geoff is =
trying to use a Lippmann type circuit, so he should get reasonably low
noise signals at 20 seconds.
****Brett Nordgren wrote ;
Even professional instruments show substantial DC drift. Broad=
band=20
verticals have an output proportional to the rate of temperature=20
change and over a 24-hour period can easily vary by a volt or more=20
depending on their thermal insulation. We routinely put in a=20
one-pole digital high-pass filter at the lowest frequency available=20
(0.002 Hz in WinSDR) which makes all that go away.
****An offset of over=
a volt could exceed the input range of some ADCs !!
Such poor quality=
electronics would be totally unacceptable in my opinion.
>> Subject: Instrumentation Question
>>
>> Yet I can not rid a small DC drift which is either related to
>> ambient AC noise level or DC drift of resistance or
>> thermocouple voltage related to the soldering of junctions.
>> This DC drift is related to the 24 hour cycle.
>> The dc drift is on the order of micro volts
>> which seem to be originating on the input.
>> The overall DC gain is 80Dbv or X10000.
>> I should be able to achieve a DC free drift at this gain.
>> I am using a op177G op amp ??
>> The sensor is HS10-1 Geophone.
>> The 40 foot of cable is designed for burial
>> it has silicon grease (I think) impregnated and
>> has a heavy copper jacket.
>> It is soldered at the geophone and also at the sensor
>> has soldered connections.
>
> If you use a CAZ type opamp like the LTC1150 you will get zero=20
> temperature drift.
>
> The problem does not seem to be the amplifier itself.
> But rather, the variables dealing with the input.
> Cable/Geophone/Common mode stuff/
> Yet I have found a wide variation within the
> op amp offset voltages.
> I am currently using a Chopper amp (or so I think)
> In the front end. LTC1050 ???
> Yes, I have troubles finding decent opamps
> at civilian prices, the best are all seem to be MilSpec
> rip offs. Outrageous prices. The milspec parts are by far
> the best and all others should simply be scrapped.
>
> Regards,
> geoff
> LTC1050 - Precision Zero-Drift Operational Amplifier with Internal=20
> Capacitors
***So is the LTC1150 ! Check it out ?=
!
Regards,
Chris Chapman
[ Top ]
[ Back ]
[ Home Page ]