Targets frequency response

[Aziz]

Hello B^C,

studying noise types on the link above, I found that there is more noise to find and detect than signals. It is probably one of the most complex problems in electronics.
But your acquired signals looks really pretty well. Noise is lower than 1 mV.
Under real conditions, we see often more noise.

It is interesting to see, what lock-in amps can do and not. I made another extreme noise test. It looks pretty well. My (surrealistic) goal is to measure nano volts on the coil. I am not sure, if this is ever possible.
Regards,
Aziz

[Aziz]

Gday Aziz,
Thanks for that, very interesting!.

I have been testing reading very small voltages from some coils & getting them to the computer.
I have "nearly" got a clean signal to the computer with 32760 samples but can't get rid of the remaining noise--not much hair left as i have near pulled it all out trying. --see attachment---

Any ideas on how to get rid of small variations, not sure exactly what is causing this??

Hi B^C,

noises can have various sources.
See the following link for noise types and their characteristics:

http://www.educypedia.be/electronics/noise.htm

I don't know where this noise is coming from (too few info).
One noise source can be the dumping resistor (PI) which is converting the magnetic energy into heat (-> Thermal noise, Johnson noise, shot noise).

Another noise sources:
Amplifier stages, filter stages, measuring instrument itself, supply voltage, not capacitively decoupled op-amp supply voltages (when used on high bandwidth), not enough frequency compensated op-amps (on fast op-amps), inductively and/or capacitive coupled noise sources, ground-loops, etc...

You should use just lock-in amp's.

Aziz

PS: A spectrum analyser should locate the noise source probably.

[B^C]

OOP'S forgot attachment!

[B^C]

Gday Aziz,
Thanks for that, very interesting!.

I have been testing reading very small voltages from some coils & getting them to the computer.
I have "nearly" got a clean signal to the computer with 32760 samples but can't get rid of the remaining noise--not much hair left as i have near pulled it all out trying. --see attachment---

Any ideas on how to get rid of small variations, not sure exactly what is causing this??

[Aziz]

Even the signal below can be handled without any problems.

[Aziz]

Gday Aziz,

Interesting Graph, not sure if i have it correct?

The top line--signal + noise-- is this the eddy current decay?

Excuse my ignorance.

Hi B^C,

any form of exponential signal decay ( f(t)=a*exp(-t/TC), TC=L/R, e.g. induced voltage on the RX coil or negative of it). It is just for noise analysis and modulation to see the signal frequency spectrum response. As it is quite a broad band spectrum, a broad band noise is also concerned. So it is a big challenge, to keep the broad band noise small and construct the original signal from the noise.
If the modulation of the signal is done properly, the spread noise spectrum can be kept small.
Following attachement shows the reconstruction of the decay curve using the lock-in amp.

Regards,
Aziz

[B^C]

Gday Mike,

Thanks for that, i may have found more information for you, i will read through it & if it seems useful i will post it for you.

[B^C]

Gday Aziz,

Interesting Graph, not sure if i have it correct?

The top line--signal + noise-- is this the eddy current decay?

Excuse my ignorance.

[Aziz]

We can carefully estimated expect, a much stable and more depth sensitiv new PI MD's in the future!
There is still huge potential, which is not found or used in the PI's yet.
I will be working intensively on this stuff now.

Aziz

[Aziz]

First Test:

The signal decay function has a quite broad band frequency spectrum. As the signal has to be modulated for the lock-in amp, the lock-in reference frequency should be synchronized with the PI timings.

First quick & dirty hacks shows, the windowed integration of the signal can be improved by using the lock-in amp. The noise will be reduced and the SNR will grow.

Now it makes sense, to analyse this deeply and to write the software for this.

It will be not easy to use a real lock-in amp in PI's. Particularly the modulation of the signal must be done properly to avoid additional noises.
Attached: Frequency spectrum of the signal decay function.
Aziz

[B^C]

It is known that iron meteorites have very weak magnetizm – relative permeability is faintly above 1. Unfortunately the dry soil here has near the same permeability. The problem is how to iluminate environment with magnetic field in order to make response from meteorites different from those of sand. Regards Mike.

Gday MIke,

I have given your comments considerable thought--sleepless night & searching for related information to your question which is a good one.

Your comments on the soil there having the same permeability as the meteorites your looking for i found very interesting. You said that the metereorites have very weak magnetism with relative permeability faintly above 1.
Of course i am not sure on what Class of Meteorite your after.

I may be wrong--again, but it is my understanding that the meteorites Magnetic Susceptibilty--MS is "Frequency Dependant--FD" & the MS is increased with the correct inducing Frequency. This may be an avenue to explore because the MS of the surrounding Matrix would surely differ than the meteorites correct frequency for higher MS.

The frequencies i have read about are :
465Hz, 825Hz, 4650Hz & (19000Hz but i'm not sure you want to go there.)

825Hz seems to give some good results & i will attach a couple of graph's to look at.

If this is right maybe you could try some of the different frequencies & make them variable +/- 100Hz to experiment.
I think the first thing to do is run some frequency tests on your Matrix conditions to get some MS figures & then isolate the meteorites for some comparitive testing.

This may help, or maybe not, let me know?

[mikebg]

I meant with XL (w*L) the reactance of the coil of course.
I will write only X in the future.
Aziz

Dear Aziz,
X is not "reactance of the coil". This is reactance of the target seen by sensor. If the sensor has TX and RX windings, they both "see" the target. If our target has a timeconstant, that means it has a "coil" connected to a "resistor" but how our sensor sees their reactance? It is not reactance of the eddy coil.
The reactance X of single eddy current target is
Oops, I can write formulas. To be continued in other formats.

[mikebg]

Demagnetization by air gap

TNX B^C,
I need help from you and your old technical Bible, because designing meteorite locator, I have problems with the magnetic properties of soil and meteorites.
I'm experienced in QRP radio. In our "Q code", the term QRP means radio communication with extremely weak signals covered with noises. There exists and such one odd hobby.
I have build a meteorite locator using as platform a Bulgarian Mine Detector MT66 (produced in 60-es). My machine can locate small stones sticked into soil if they have magnetizm or conductivity, very slight different from emvironment. Now I have a useless collection of near tausend stones, none of them is meteorite.
It is known that iron meteorites have very weak magnetizm – relative permeability is faintly above 1. Unfortunately the dry soil here has near the same permeability. The problem is how to iluminate environment with magnetic field in order to make response from meteorites different from those of sand. My radio colleagues think that should be difference in demagnetisation. The hysteresis cycle of soil and stones should be different because of a phenomenon named “air gap” or “nonmagnetic gap”. There is more air gaps between grit in sand than between magnetic domains in the stone. Please, browse your Bible for information about demagnetisation. I stop here because this is unimportant information for other participians, and because I should write a posting for Aziz. Hi again uses a wrong term.
Regards Mike.

[B^C]

Dear Bruce,
In front of the whole world (because this is an excellent worlds forum), I beg your pardon for my bad words comcerning your knowledge, education and your old Bible with technical articles. Too soon the fate has showed that I'm wrong and willed me to look for help from physicists and from old technology Bibles for solving problems, which an amateur (ham radio) designer, like me can't solve using all WEB resources and whole community of amateurs.
I will explain these problems and my requests for help in other postigs.
Submissive yours Mike.

Gday Mike,

No apology needed, i did not take offence at your comments, i actually thought your comment was amusing, i'm Australian & a thick skinned son of a ***** with a good sense of humour

Experienced ham radio designers are "very" intelligent people & more so than others recognise, i am as intelligent as a rock.
No need to beg my pardon, i am but one man & nobody in particular. We can all learn something here on this forum, i just wish others would add there bit as well & i see some others have responded.

So, no need to apologise in the future, i question everything until it is proven so it can seem a little abrupt at times, it's just the nature of the beast.

[Aziz]

Aziz,

In SD200 the demodulators is not Qadrature Demodulators
This is like deferential integrator for good SNR...
If you want to make Digital PI (true), first you need very fast ADC
I make before tests with 12 bits, 400 K samples per sec(my avatar is PCB board) - results is good but not the best...

Hi TheWizard,

I do not need really an ADC for this. Digital processing is too much expensive at the moment and it is also an issue of power consumption.
This can be realized very simple and very effective.

Before speculating or promising too much, a frequency domain analysis including noise analysis must be made. And I will change my simulating software for this purpose particularly.
Aziz