PSN-L Email List Message

Subject: Re: High noise >10s
From: "Tiago A." tiagoagre@.........
Date: Thu, 9 Aug 2012 02:05:28 +0000


Forgot the picture to compare noise before/now:

http://dl.dropbox.com/u/15802004/com.png

2012/8/9 Tiago A. 

> Thanks for the answer Geoff.
>
> I=B4m not using Larrys Boards (yet), im using a ADC ADS1211P mounted in a
> PCB, image:
> http://imageshack.us/photo/my-images/31/adca.jpg/
>
> The board worked ok for almost a year.
>
> Hadnt much time to test more the noise, but here is a picture which shows
> very well the issues:
> First image - 3 months back. 0.25Hz to 0.75Hz is ocean noise.
> Second image - Taken few minutes ago.
>
>
> >10s noise is probably 10 times higher than before, also i have some high
> count at a specific frequencies, starts at +-1Hz, next day 2Hz, next day
> 3Hz). Restart the board/Amaseis, the high count returns to +-1Hz.
>
> I will follow your advices and i will search some noise sources.
>
> Best Regards,
> Tiago.
>
>
> 2012/8/8 GMVoeth GM 
>
>> Sorry Tiago,
>>
>> I can not now handle SAC files.
>>
>> The only kind of files I handle are PSN.
>>
>> I recommend you change sample rates, Assuming you are using Larrys Type
>> III A/D board.
>>
>> Then you create a test file, convert to WAVE both before and after sampl=
e
>> rate change files.
>>
>> Then look at the results in Audacity FFT Waterfall program.
>>
>> This is the only way i know how to evaluate such a problem.
>>
>> You talk about things I know nothing about like (sample rate to 125sps
>> (-3dB at 36Hz)).
>>
>> (-3dB at 36Hz) means to me a Low Pass Filter At 36Hz.
>>
>> If you filter at 36 hZ your sample rate should be much higher than 100
>> more like 500 sps.
>>
>> You need to filter out even frequencies above the rolloff.
>>
>> When you mix frequencies you get the following,
>>
>> F1 & F2 (Fundamentals),
>>  F1 + F2 (SUM), and,
>>  F1 - F2 (Difference).
>>
>> You will notice small variations in the fundamentals
>> can result in large changes at higher harmonics.
>>
>>
>>
>> I am assuming that when you get alizing you are getting these mixing
>> products.
>>
>> Frequency mixing is typical for superhetrodyne receivers.
>>
>> If your problem is not mixing you need to
>> look at everything else like staic electricity,
>> Commom mode rejection, Thermal variations.
>>
>> Like take a soda straw and carefully blow your warm breath on
>> various components to see what causes changes.
>> Do this while recording.
>>
>> It took me a long time to get rid of most artifacts.
>>
>> The best thing I did was buy a manufactured (used) geophone
>> then lower the overall gain of my system.
>>
>> If you need more than x10000, little problems can easily be big ones.
>>
>> I was unable to stop unwanted artifacts if gain was over X100,000.
>>
>> Good Luck Comrade,
>> Regards,
>> geoff
>>
>>
>>
>>
>> On 8/2/2012 4:19 PM, Tiago A. wrote:
>>
>>> Hi Geoff.
>>>
>>> Increasing the sample rate to 125sps (-3dB at 36Hz), didint solve the
>>> issue and increased the >10s noise..i will try to increase more the
>>> sample rate. But increasing the sample rate reveals a very high count a=
t
>>> 50Hz, i check the earth connections it looks all ok..if i remove the
>>> geophone from the ADC, i have +-2 counts of noise at 16bits and the 50H=
z
>>> spike is gone. Simply touching a wire from earth in any part of the
>>> geophone, will make the geophone stiff, without responding.
>>>
>>> SAC file with 15minutes at 125SPS
>>> http://dl.dropbox.com/u/**15802004/aa.sac
>>>
>>> Netherless the 50Hz issue, getting only 3800ohm from a 5500ohm geophone
>>> is a good indicator that something is wrong with the geophone itself. I=
t
>>> still responds ok to >0.1Hz, but <0.1Hz is just noise.
>>>
>>>
>>>
>>> 2012/8/2 Geoff >
>>>
>>>
>>>     if the geophone is OK then i understand
>>>     the rdamp should be like 4x Rgeo.
>>>     together the final R value is (Rd||Rg)
>>>     the final Ge is like GRg/(Rd + Rg).
>>>     If your geophone has a resistance
>>>     more than like 1% off at room temperature
>>>     you must consider it damaged somehow.
>>>     Regards,
>>>     geoff
>>>
>>>
>>>     On 2012-07-30 14:34, Tiago A. wrote:
>>>
>>>         Thanks for the answer Chris.
>>>
>>>              I am not sure that I understand what you have written - or
>>> what
>>>         system you are using. What ADC / number of bits, are you using =
?
>>>
>>>         The system (ADC) is a ADS1211P, the geophone is connected
>>>         directly to
>>>         the ADC inputs with a 20cm wires. geIm using the A/D set to
>>>         22bits with
>>>
>>>         the PGA set to 2x. With a Decimate Factor of 5600 with Turbo
>>> Mode 8,
>>>         gives a sample rate of 27.86, which sets -3dB at 8Hz (3rd order
>>>         digital
>>>         filter) .
>>>
>>>         You MUST damp a 1 Hz geophone correctly to get a flat output
>>>         over 1 Hz.
>>>         What is the value of the damping resistor should i use? Its a
>>>         5500ohm
>>>         geophone. But the strangest thing is i can only measure 3800ohm
>>>         (tested
>>>         today), i think when i got the geophone it had a higher
>>>         resistance. This
>>>         geophone has be altered, the previous owner changed the origina=
l
>>>         Bendix
>>>         6/8 pin connector to a 4 pin Bendix connector.
>>>
>>>         How are you getting your period expansion below 1 Hz down to 0.=
1
>>>         Hz ?
>>>         Only using AmaSeis. One of the few times i could get down to
>>>         0.03Hz was
>>>         in the Mexico M 7.4 earthquake (~7500km). It was offline in the
>>>         Sumatra
>>>         earthquakes. I live in a small island in the middle of the
>>>         ocean, so i
>>>         have always a lot of noise in the 2~5s period.
>>>
>>>         What is the noise on your ADC - when you connect it to a
>>>         resistor with
>>>         the same value as the geophone ?
>>>         I can check that in the next days.
>>>
>>>
>>>         Thanks.
>>>
>>>         Regards,
>>>         Tiago.
>>>
>>>
>>>
>>>         2012/7/30 
>>>         >>
>>>
>>>              From: Tiago A. tiagoagre@.........
>>>          >>
>>>         >
>>>
>>>              Sent: Mon, 30 Jul 2012 3:08
>>>              Subject: High noise >10s
>>>              Hi.
>>>
>>>              Im using a L4 1Hz geophone and for almost a month im
>>>         struggling with
>>>              high noise from 10s up to ..., but from <10s it looks good=
.
>>> The
>>>              noise comes and goes, this image is self explaining:
>>>
>>>         http://dl.dropbox.com/u/__**15802004/noise.png
>>>
>>>         
>>> >
>>>
>>>              Changed cables, re centered the geophone, but didint
>>> help..the
>>>              geophone is connected directly to a ADC without damping,
>>>         external
>>>              amplification or filtering, i only use digital filtering
>>>         from the
>>>              ADC. Before this issue start, i had a pretty clean spectru=
m
>>>          >10s for
>>>              almost 6 months, which allowed me to received LQ and LR
>>>         from some
>>>              strong events.
>>>
>>>              This sac contains 15minutes with high noise >10s, notice
>>>         that at
>>>              3:37 it recorded a Ml2.1 at 70km.
>>>         http://dl.dropbox.com/u/__**15802004/noise.sac
>>>
>>>         
>>> >
>>>
>>>              Hi Tiago,
>>>                   I am not sure that I understand what you have written
>>>         - or what
>>>              system you are using.
>>>              You MUST damp a 1 Hz geophone correctly to get a flat
>>>         output over 1 Hz.
>>>              How are you getting your period expansion below 1 Hz down
>>>         to 0.1 Hz ?
>>>              Any period expansion DEPENDS on the geophone being
>>>         correctly damped.
>>>              What is the noise on your ADC - when you connect it to a
>>>         resistor
>>>              with the same value as the geophone ?
>>>              You do need a low noise metal film resistor.
>>>              What ADC / number of bits, are you using ?
>>>              Regards,
>>>              Chris Chapman
>>>
>>>
>>>
>>>
>>>     ______________________________**______________________________
>>>
>>>
>>>     Public Seismic Network Mailing List (PSNLIST)
>>>
>>>     To leave this list email PSNLIST-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 (PSNLIST)
>>
>> To leave this list email PSNLIST-REQUEST@.............. with the body of
>> the message (first line only): unsubscribe
>> See http://www.seismicnet.com/**maillist.htmlfor more information.
>>
>
>
Forgot the picture to compare noise before/now:

http://dl.dropbox.com/u/15802004/com.png

2012/8/9 Tiago A. &l= t;tiagoagre@....... com>
Thanks for the answer Geoff.

I=B4m no= t using Larrys Boards (yet), im using a ADC ADS1211P mounted in a PCB, imag= e:
http://imageshack.us/photo/my-images/31/adca.jpg/

The board worked ok for almost a year.

Hadnt much time to test m= ore the noise, but here is a picture which shows very well the issues:
F= irst image - 3 months back. 0.25Hz to 0.75Hz is ocean noise.
Second imag= e - Taken few minutes ago.


>10s noise is probably 10 times higher than before, also i have = some high count at a specific frequencies, starts at +-1Hz, next day 2Hz, n= ext day 3Hz). Restart the board/Amaseis, the high count returns to +-1Hz. <= br>
I will follow your advices and i will search some noise sources.
Best Regards,
Tiago.


2012/8/8 GMVoeth GM <gmvoeth@.........>
Sorry Tiago,

I can not now handle SAC files.

The only kind of files I handle are PSN.

I recommend you change sample rates, Assuming you are using Larrys Type III= A/D board.

Then you create a test file, convert to WAVE both before and after sample r= ate change files.

Then look at the results in Audacity FFT Waterfall program.

This is the only way i know how to evaluate such a problem.

You talk about things I know nothing about like (sample rate to 125sps (-3d= B at 36Hz)).

(-3dB at 36Hz) means to me a Low Pass Filter At 36Hz.

If you filter at 36 hZ your sample rate should be much higher than 100
more like 500 sps.

You need to filter out even frequencies above the rolloff.

When you mix frequencies you get the following,

F1 & F2 (Fundamentals),
=A0F1 + F2 (SUM), and,
=A0F1 - F2 (Difference).

You will notice small variations in the fundamentals
can result in large changes at higher harmonics.



I am assuming that when you get alizing you are getting these mixing produc= ts.

Frequency mixing is typical for superhetrodyne receivers.

If your problem is not mixing you need to
look at everything else like staic electricity,
Commom mode rejection, Thermal variations.

Like take a soda straw and carefully blow your warm breath on
various components to see what causes changes.
Do this while recording.

It took me a long time to get rid of most artifacts.

The best thing I did was buy a manufactured (used) geophone
then lower the overall gain of my system.

If you need more than x10000, little problems can easily be big ones.

I was unable to stop unwanted artifacts if gain was over X100,000.

Good Luck Comrade,
Regards,
geoff




On 8/2/2012 4:19 PM, Tiago A. wrote:
Hi Geoff.

Increasing the sample rate to 125sps (-3dB at 36Hz), didint solve the
issue and increased the >10s noise..i will try to increase more the
sample rate. But increasing the sample rate reveals a very high count at 50Hz, i check the earth connections it looks all ok..if i remove the
geophone from the ADC, i have +-2 counts of noise at 16bits and the 50Hz spike is gone. Simply touching a wire from earth in any part of the
geophone, will make the geophone stiff, without responding.

SAC file with 15minutes at 125SPS
http:= //dl.dropbox.com/u/15802004/aa.sac

Netherless the 50Hz issue, getting only 3800ohm from a 5500ohm geophone
is a good indicator that something is wrong with the geophone itself. It still responds ok to >0.1Hz, but <0.1Hz is just noise.



2012/8/2 Geoff <g= mvoeth@......... <mailto:gmvoeth@.........>>


=A0 =A0 if the geophone is OK then i understand
=A0 =A0 the rdamp should be like 4x Rgeo.
=A0 =A0 together the final R value is (Rd||Rg)
=A0 =A0 the final Ge is like GRg/(Rd + Rg).
=A0 =A0 If your geophone has a resistance
=A0 =A0 more than like 1% off at room temperature
=A0 =A0 you must consider it damaged somehow.
=A0 =A0 Regards,
=A0 =A0 geoff


=A0 =A0 On 2012-07-30 14:34, Tiago A. wrote:

=A0 =A0 =A0 =A0 Thanks for the answer Chris.

=A0 =A0 =A0 =A0 =A0 =A0 =A0I am not sure that I understand what you have wr= itten - or what
=A0 =A0 =A0 =A0 system you are using. What ADC / number of bits, are you us= ing ?

=A0 =A0 =A0 =A0 The system (ADC) is a ADS1211P, the geophone is connected =A0 =A0 =A0 =A0 directly to
=A0 =A0 =A0 =A0 the ADC inputs with a 20cm wires. geIm using the A/D set to=
=A0 =A0 =A0 =A0 22bits with

=A0 =A0 =A0 =A0 the PGA set to 2x. With a Decimate Factor of 5600 with Turb= o Mode 8,
=A0 =A0 =A0 =A0 gives a sample rate of 27.86, which sets -3dB at 8Hz (3rd o= rder
=A0 =A0 =A0 =A0 digital
=A0 =A0 =A0 =A0 filter) .

=A0 =A0 =A0 =A0 You MUST damp a 1 Hz geophone correctly to get a flat outpu= t
=A0 =A0 =A0 =A0 over 1 Hz.
=A0 =A0 =A0 =A0 What is the value of the damping resistor should i use? Its= a
=A0 =A0 =A0 =A0 5500ohm
=A0 =A0 =A0 =A0 geophone. But the strangest thing is i can only measure 380= 0ohm
=A0 =A0 =A0 =A0 (tested
=A0 =A0 =A0 =A0 today), i think when i got the geophone it had a higher
=A0 =A0 =A0 =A0 resistance. This
=A0 =A0 =A0 =A0 geophone has be altered, the previous owner changed the ori= ginal
=A0 =A0 =A0 =A0 Bendix
=A0 =A0 =A0 =A0 6/8 pin connector to a 4 pin Bendix connector.

=A0 =A0 =A0 =A0 How are you getting your period expansion below 1 Hz down t= o 0.1
=A0 =A0 =A0 =A0 Hz ?
=A0 =A0 =A0 =A0 Only using AmaSeis. One of the few times i could get down t= o
=A0 =A0 =A0 =A0 0.03Hz was
=A0 =A0 =A0 =A0 in the Mexico M 7.4 earthquake (~7500km). It was offline in= the
=A0 =A0 =A0 =A0 Sumatra
=A0 =A0 =A0 =A0 earthquakes. I live in a small island in the middle of the<= br> =A0 =A0 =A0 =A0 ocean, so i
=A0 =A0 =A0 =A0 have always a lot of noise in the 2~5s period.

=A0 =A0 =A0 =A0 What is the noise on your ADC - when you connect it to a =A0 =A0 =A0 =A0 resistor with
=A0 =A0 =A0 =A0 the same value as the geophone ?
=A0 =A0 =A0 =A0 I can check that in the next days.


=A0 =A0 =A0 =A0 Thanks.

=A0 =A0 =A0 =A0 Regards,
=A0 =A0 =A0 =A0 Tiago.



=A0 =A0 =A0 =A0 2012/7/30 <chrisatupw@....... <mailto:chrisatupw@.......>
=A0 =A0 =A0 =A0 <mailto:chrisatupw@....... <mailto:chrisatupw@.......>>>

=A0 =A0 =A0 =A0 =A0 =A0 =A0From: Tiago A. tiagoagre@.........
=A0 =A0 =A0 =A0 <mailto:tiagoagre@.........> <mailto:tiagoagre@.........

=A0 =A0 =A0 =A0 <mailto:tiagoagre@.........>>

=A0 =A0 =A0 =A0 =A0 =A0 =A0Sent: Mon, 30 Jul 2012 3:08
=A0 =A0 =A0 =A0 =A0 =A0 =A0Subject: High noise >10s
=A0 =A0 =A0 =A0 =A0 =A0 =A0Hi.

=A0 =A0 =A0 =A0 =A0 =A0 =A0Im using a L4 1Hz geophone and for almost a mont= h im
=A0 =A0 =A0 =A0 struggling with
=A0 =A0 =A0 =A0 =A0 =A0 =A0high noise from 10s up to ..., but from <10s = it looks good. The
=A0 =A0 =A0 =A0 =A0 =A0 =A0noise comes and goes, this image is self explain= ing:

=A0 =A0 =A0 =A0 http://dl.dropbox.com/u/__15802004/noise.png
=A0 =A0 =A0 =A0 <http://dl.dropbox.com/u/15802004/noise.png>=

=A0 =A0 =A0 =A0 =A0 =A0 =A0Changed cables, re centered the geophone, but di= dint help..the
=A0 =A0 =A0 =A0 =A0 =A0 =A0geophone is connected directly to a ADC without = damping,
=A0 =A0 =A0 =A0 external
=A0 =A0 =A0 =A0 =A0 =A0 =A0amplification or filtering, i only use digital f= iltering
=A0 =A0 =A0 =A0 from the
=A0 =A0 =A0 =A0 =A0 =A0 =A0ADC. Before this issue start, i had a pretty cle= an spectrum
=A0 =A0 =A0 =A0 =A0>10s for
=A0 =A0 =A0 =A0 =A0 =A0 =A0almost 6 months, which allowed me to received LQ= and LR
=A0 =A0 =A0 =A0 from some
=A0 =A0 =A0 =A0 =A0 =A0 =A0strong events.

=A0 =A0 =A0 =A0 =A0 =A0 =A0This sac contains 15minutes with high noise >= 10s, notice
=A0 =A0 =A0 =A0 that at
=A0 =A0 =A0 =A0 =A0 =A0 =A03:37 it recorded a Ml2.1 at 70km.
=A0 =A0 =A0 =A0 http://dl.dropbox.com/u/__15802004/noise.sac
=A0 =A0 =A0 =A0 <http://dl.dropbox.com/u/15802004/noise.sac>=

=A0 =A0 =A0 =A0 =A0 =A0 =A0Hi Tiago,
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 I am not sure that I understand what yo= u have written
=A0 =A0 =A0 =A0 - or what
=A0 =A0 =A0 =A0 =A0 =A0 =A0system you are using.
=A0 =A0 =A0 =A0 =A0 =A0 =A0You MUST damp a 1 Hz geophone correctly to get a= flat
=A0 =A0 =A0 =A0 output over 1 Hz.
=A0 =A0 =A0 =A0 =A0 =A0 =A0How are you getting your period expansion below = 1 Hz down
=A0 =A0 =A0 =A0 to 0.1 Hz ?
=A0 =A0 =A0 =A0 =A0 =A0 =A0Any period expansion DEPENDS on the geophone bei= ng
=A0 =A0 =A0 =A0 correctly damped.
=A0 =A0 =A0 =A0 =A0 =A0 =A0What is the noise on your ADC - when you connect= it to a
=A0 =A0 =A0 =A0 resistor
=A0 =A0 =A0 =A0 =A0 =A0 =A0with the same value as the geophone ?
=A0 =A0 =A0 =A0 =A0 =A0 =A0You do need a low noise metal film resistor.
=A0 =A0 =A0 =A0 =A0 =A0 =A0What ADC / number of bits, are you using ?
=A0 =A0 =A0 =A0 =A0 =A0 =A0Regards,
=A0 =A0 =A0 =A0 =A0 =A0 =A0Chris Chapman




=A0 =A0 ____________________________________________________________=


=A0 =A0 Public Seismic Network Mailing List (PSNLIST)

=A0 =A0 To leave this list email PSNLIST-REQUEST@..............
=A0 =A0 <mailto:PSNLIST-REQUEST@SEISMICNET.COM> with the body of = the message
=A0 =A0 (first line only): unsubscribe
=A0 =A0 See http://www.seismicnet.com/__maillist.html

=A0 =A0 <http://www.seismicnet.com/maillist.html> for more info= rmation.



__________________________________________________________

Public Seismic Network Mailing List (PSNLIST)

To leave this list email PSNLIST-REQUEST@.............. with the body of the m= essage (first line only): unsubscribe
See h= ttp://www.seismicnet.com/maillist.html for more information.



[ Top ] [ Back ] [ Home Page ]