Geoff,

That is exactly what you need to make a calibration file for your
geophone.  I few years ago I migrated to a suite of sismic analysis
programs called Seisan, which is free and freely available.  It includes
a program called resp.exe.  find it at ftp://190.14.211.35/ with
username "unam" and password "unam".

When you run it you will be prompted for all the items you listed and a
few you didn't list.  Running it is easy what you will have at the end
of the process is a calibration file in one of several formats.  Go for
the GSE2 PAZ (poles and zeros)

The begged question is what do you do with a GSE2 PAZ response file once
you have it.  Let's not go there yet.

What follows is cut and pasted from the Seisan manual.

Saludos,

Angel


This roughly how it works:

Type RESP to start the program. You will then get a series of questions
as indicated below in upper case letters. All input is format free. A
sample run is shown below. 

CHOSE OUTPUT FORMAT: 0: NO OUTPUT FILE 
1: SEISAN FAP 
2: SEISAN PAZ 
3: GSE2 FAP 
4: GSE2 PAZ 

Answer with 0-4, options 1-4 will create respective response files in
selected format, option 0 will only calculate and show the response on
the screen. SEISAN PAZ can only be used if number of poles and number of
zeros are less than 38. If more are input, a table will be generated
automatically in FAP format. 

TYPE OF SENSOR: 
1: NONE 
2: SEISMOMETER 
3: ACCELEROMETER 

Answer with 1, 2 or 3. Number 1 is used when only calculation of filters
or amplifiers are desired, 2 is a standard velocity transducer and 3 a
standard accelerometer. If a seismic sensor is used, you will get
additional questions on the constants of the sensor. If a seismometer is
chosen, the following questions must be answered: 

SEISMOMETER NATURAL PERIOD ? 

This is measured in seconds. For most short period systems the value
would be 1.0 second. 

SEISMOMETER DAMPING RATIO ? 

The damping ratio should ideally be 0.7. This depends on the damping
resistance. 

For both the seismometer and accelerometer, the following question is
given: 

SENSOR LOADED GENERATOR CONSTANT (V/M/S OR V/G) ? 

This is the generator constant of the sensor in terms of volt per unit
of of ground motion (meter/second or g). It is important to note that
this is the loaded constant, which means the effective output of the
sensor taking into account amplifier input and damping resistances.

Now comes questions about amplifier, filter and recording unit. 

RECORDING MEDIA GAIN (COUNT/V OR M/V) ? 

If you have a recording media, the gain can be given here, otherwise
just enter 1.0 

If the output format is GSE, the response is always calculated in
displacement units, while for SEISAN output and seismometer or
accelerometer, the following options appear: 

TYPE OF RESPONSE: 
1: DISPLACEMENT 
2: VELOCITY 
3: ACCELERATION 

Normally for a seismometer, one wants to calculate the displacement
response and for an accelerometer, the acceleration response. However it
might sometimes be interesting to look at e.g. the velocity response for
a seismometer (after all, the seismometer is normally a velocity
transducer !!). Enter the appropriate number. 

AMPLIFIER GAIN (DB) ? 

This is the amplifier gain in dB. Since this question is only asked
once, this gain must include gain of all units except the recorder
(asked below). This could e.g. include gain of the VCO system. 

NUMBER OF FILTERS (0-10), RETURN FOR NONE ? 

Up to 10 filters can be specified. If you answer 0, no filters are used
and no more questions on filters will appear. Otherwise one line of
input must be given for each filter as follows: 

FREQUENCY AND NUMBER OF POLES FOR EACH FILTER, 
POLES NEGATIVE FOR HIGH PASS 

Each line requires two numbers, the corner frequency of the filter and
the number of poles. A high pass filter is given by letting number of
poles be negative. It is not always easy to know whether a filter is
e.g. one 2 pole or two 1 pole filters, the user needs to experiment with
this. 

FILE NAME FOR FILE WITH POLES AND ZEROS, RETURN FOR NO FILE 

Here a file with poles and zeros can be entered. If seismometer
constants have been chosen above, the values calculated with poles and
zeros are multiplied with the values previously calculated. The free
format file contains: 

1. line: NP: Number of poles, NZ: Number of zeros, Norm: Normalization
constant 
Following NP lines contain one pair each of real and imaginary poles 
Following NZ lines contain one pair each of real and imaginary zeros 

NOTE: The unit of frequency is radian/s so if in Hz, multiply with $ 2
\pi$ and normalization constant in $ radian= (normalization constant in
Hz) 2\pi^{(number of poles-number of zeroes)}$. 

The next 2 options are only shown if the output file is selected to be
FAP: 

FILE NAME FOR TABULATED VALUES, RETURN FOR NO FILE 

Here a file with tabulated values are entered. If seismometer constants
or poles and zeros have been chosen above, the tabulated values will be
interpolated and multiplied with the values previously calculated for
from above. The free format file contains: 

1. line: N: Number of tabulated values, Norm: Normalization constant 
Following N lines contain one each frequancy, amplitude and phase(deg) 

GIVE FILE NAME FOR MEASURED VALUES, RETURN FOR NONE 

Give file name for measured values. In most cases you have none so just
make a return. The format of the input file is as follows: 

frequency, amplitude, phase 
frequency, amplitude, phase 
etc. 

e.g. 

0.2,0.7,200 
0.7,0.8,100 
10.0,0.1,33 

The file has no blank lines and can contain up to 60 data sets. It is
important to note that the amplitude values should be NORMALIZED at 1.0
Hz. 

Now there is no more input to the response parameters, and the output
is: 

GAIN FACTOR AT 1.0 HZ: 12345.6 

This is the gain of the system at 1.0 Hz and is also the value for
normalizing the response curve, that is, all calculated values are
divided by this number. There is no unit for gain of an amplifier and
for displacement response using a seismometer and drum recording. If the
recording is digital, the unit would be counts/meter and for a velocity
response counts/meter/second etc. If a file with poles and zeros is used
without any other information, the normalization constant must have the
unit of count/m, similar for the tabulated input. 

Further output is given in a file called resp.out, see Table 1 for an
example. 

The response curves (amplitude and phase) are now printed/plotted on the
screen. First comes the amplitude response (amplitude in db versus log
frequency). By pushing return, the phase response is shown (phase shift
(deg) versus log frequency). 






On Fri, 2010-04-09 at 22:09 -0700, Geoffrey wrote:
> Thank you much sismos/Angel,
> 
> More questions to ask.
> 
> The specs for my geophone is like an ideal 2.99V/IN/SEC
> Is that like ( 117.7V/M/sec) or (1.177V/cm/SEC).
> [which seems more like 1.5V/IN/SEC when I look at it myself]
> through the filtered electronics. Then I amplify like X1025 (60Dbv)
> Then into a 12bit A/D running at +/- 2.045 Volts.
> This means about 1023 count == output of 1.023Volts
> Which is (0.001V/count)/1025 or 0.975 microvolts per count
> on the geophone output which is possibly 114.7micrometers/sec of motion ?
> Is this the kind of information they want ?
> 
> The order goes something like this:
> 1. Geophone input:   1uM/sec
> 2. Geophone output: 117.7 uV
> 3. Amplifier Input:      117.7 uV
> 4. Amplifier Output:  120.7 mV
> 5. Counts out of A/D: +120
> Does any of this sound right ?
> Does this come anywhere close to calibration ideas ?
> 
> The only way I know how to calibrate or evaluate the
> sensitivity is to go to the USGS website for
> Theoretical values relating to a specific EQ then
> translating the numbers backward through the system
> to match the theoretical intensity of the EQ
> as it passed through my system at my location.
> Might this be a proper way to look at my
> own system ??
> 
> Right now for regional quakes Winquake gives me a very good idea
> of the actual Ml (Richter magnitude). Might I
> use the correctional numbers for Winquake to
> enhance the accuracy and precision of my sensor.
> What I really mean is this, if I can achieve zero correction
> for correct ML readings, then my calibration numbers must be right ?
> Then I may send this number to IRIS with my data ??
> 
> Do I need to bandpass filter the read data at 1Hz before applying the numbers
> to the data since the USGS states these numbers to be at 1Hz?
> 
> Best regards,
> geoff
> 
> ----- Original Message ----- 
> From: "sismos" 
> To: 
> Sent: Friday, April 09, 2010 7:46 PM
> Subject: Re: Question to Techno-Philes
> 
> 
> > Hi Geoff,
> > 
> > I have to admit you're a tough one.
> > 
> > 
> > About metrology;
> > 
> > Calibrating a thermometer to an accuracy of +/- 1 degree so you can
> > measure the temp of baby formula can be done in most any kitchen and no
> > one can argue that it is not calibrated. 
> > 
> > Calibrating seismographs is much the same, it just does not have to be
> > that accurate. A few parts per thousand is good enough and that is
> > within the reach most anyone that can build a home seismometer.  I was
> > calibrating a sensor during a class and sweating the decimals and the
> > instructor said to calm down, +/- 5% was fine.
> > 
> > He asked me to image setting off a firecracker at one end of a pile of
> > construction ruble and sensing it at the other end with a very accurate
> > sensor and one that was not so accurate and trying to say much about the
> > pile of ruble or nature of the firecracker with either sensor.
> > 
> > About amateurs;
> > 
> > Being an amateur has very little if anything to do with the quality of
> > what is being done. In it's simplest context it just means you don't get
> > paid for doing what you do. Many amateur do great work and in some cases
> > better work than paid professional.  Paid professional and scientist
> > have to be concerned about the bottom line and us amateurs can ignore
> > that reality.  I think you would be surprised that many amateurs are
> > building seismometers that are just as good as what the professionals
> > build, and better than some, and I will add if you can calibrate and
> > keep the temperature constant then the data from any seismometer is good
> > data. 
> > 
> > About IRIS
> > 
> > Yes, you can send raw data to IRIS along with the calibration file and
> > they will make a power density function display for you, but you will
> > need the calibration file.  IRIS knows nothing about your seismometer
> > and the calibration file allows them not to care.  That's what the
> > calibration file is about, not caring about the sensor and it's
> > associated electronics.
> > 
> > Geoff, I have no doubts that calibrating a seismograph is well within
> > your abilities.  I will add that if all you care about is picking phases
> > (P and S and their variants) then calibration is not an issue.  And for
> > small local earthquakes you can use the duration or coda to determine
> > magnitude (Mc or Md) so you have no need for calibration, it does not
> > matter.  And if you had a bunch of sensors like that and some decent
> > timestamping of the data then you could locate epicenters.  
> > 
> > Regards,
> > 
> > Angel
> > 
> > 
> > 
> > 
> > __________________________________________________________
> > 
> > Public Seismic Network Mailing List (PSN-L)
> > 
> > To leave this list email PSN-L-REQUEST@.............. with 
> > the body of the message (first line only): unsubscribe
> > See http://www.seismicnet.com/maillist.html for more information.
> >
> __________________________________________________________
> 
> Public Seismic Network Mailing List (PSN-L)
> 
> To leave this list email PSN-L-REQUEST@.............. with 
> the body of the message (first line only): unsubscribe
> See http://www.seismicnet.com/maillist.html for more information.
> 


__________________________________________________________

Public Seismic Network Mailing List (PSN-L)