Integrator Circuits for PI.

[Orbit]

Simple words, a typical example amplifier cancellation earth effects with a lot of minerals !?sorry for the bad translation !

[mikebg]

You are a good theorist , Mike ... but don't forget that electronics is not only a theory . Electronics is a kind of art

Deemon, the design of an electronic device should start with analysis what is the most suitable block diagram. Do you know what is the most suitable block diagram for PI metal detector? If you have a textbook on signals and systems, the answer is inside.

Of course , the main problem in the classic PI design is the decay curve itself . And the worst thing is that this curve ( and its integral ) can depend on many factors . This fundamental problem which cause DC instability of the whole device ...

The main problem is the name "pulse" of this technology. It is an energy inefficient specific case of wideband technology. The decay curve is simply a large AIR signal because is used the worst search head - selfinductance (monocoil).

[deemon]

Deemon, the design of an electronic device should start with analysis what is the most suitable block diagram. Do you know what is the most suitable block diagram for PI metal detector?

What diagram is the most suitable depends on many different conditions . Electronics has more than one dimension ...

The main problem is the name "pulse" of this technology. It is an energy inefficient specific case of wideband technology. The decay curve is simply a large AIR signal because is used the worst search head - selfinductance (monocoil).

I cannot agree with you , Mike . The exponential shape of the air signal is not the pulse concept flaw , but the property of only one ( the oldest ) implementation . We can make this technology power-efficient , we can make the half-sine flyback pulse with finite duration ( in contrary with old familiar exponent decay ) and thus completely separate transmit and receive intervals , we can even receive during a transmit pulse - and do all these tricks with the same MONOCOIL ( which is not the worst , but the best search head , IMHO ) ... all what we need is a good imagination Electronics is the art , don't forget about it , Mike

[Ferric Toes]

What diagram is the most suitable depends on many different conditions . Electronics has more than one dimension ...

I cannot agree with you , Mike . The exponential shape of the air signal is not the pulse concept flaw , but the property of only one ( the oldest ) implementation . We can make this technology power-efficient , we can make the half-sine flyback pulse with finite duration ( in contrary with old familiar exponent decay ) and thus completely separate transmit and receive intervals , we can even receive during a transmit pulse - and do all these tricks with the same MONOCOIL ( which is not the worst , but the best search head , IMHO ) ... all what we need is a good imagination Electronics is the art , don't forget about it , Mike

I've not had a problem with the exponential self inductance signal (as I prefer to call it). Air signal is a bad description as the same would happen in a vacuum, and provided the TC is <5th of the TC of the target, then cut off is almost as good as instantaneous.

Receiving in the TX on time with a mono coil with conventional rectangular TX pulses is also perfectly possible, as long as certain conditions are observed.

From Mike "the design of an electronic device should start with analysis what is the most suitable block diagram". This was all I needed to see in 1966, a building block to perform an operation that I wanted. Then it was straight to the bench to try it out - and it worked. No long pages of calculations; that comes later when you want to find out how to optimise it. Even so, many things can just be tried on the bench much quicker.

Eric.

[mikebg]

What diagram is the most suitable depends on many different conditions . Electronics has more than one dimension ...
I cannot agree with you , Mike . The exponential shape of the air signal is not the pulse concept flaw , but the property of only one ( the oldest ) implementation . We can make this technology power-efficient , we can make the half-sine flyback pulse with finite duration ( in contrary with old familiar exponent decay ) and thus completely separate transmit and receive intervals , we can even receive during a transmit pulse - and do all these tricks with the same MONOCOIL ( which is not the worst , but the best search head , IMHO ) ... all what we need is a good imagination Electronics is the art , don't forget about it , Mike

Dmitry, I explained this several times in the forum, but I will make again an explanation especially for you.
To increase sensitivity of a metal detector, we need a search head which generates minimal AIR&GND signal.
Below is shown the main part of block diagram valid for all types MDs (metal detectors). The sensitivity of a MD
depends on usable gain of preamp and the power of TX. At ideal induction balance, the sensitivity is limited by GND signal, interference and noise.

This not the case for conventional MDs. The gain of preamp is limited by a large AIR&GND signal existing in its input. We can increase the gain until the AIR&GND signal starts to saturate the output of preamp. If we try to increase TX power, this increases the AIR&GND signal and we should decrease the gain to avoid saturation.

The advantages of Monocoil are: minimal weight, minimal price, easy to build, no zones with back readings. However it
is the worst search head for metal detectors because generates maximal AIR&GND signal.

[mickstv]

Dmitry, I explained this several times in the forum, but I will make again an explanation especially for you.
To increase sensitivity of a metal detector, we need a search head which generates minimal AIR&GND signal.
Below is shown the main part of block diagram valid for all types MDs (metal detectors). The sensitivity of a MD
depends on usable gain of preamp and the power of TX. At ideal induction balance, the sensitivity is limited by GND signal, interference and noise.

This not the case for conventional MDs. The gain of preamp is limited by a large AIR&GND signal existing in its input. We can increase the gain until the AIR&GND signal starts to saturate the output of preamp. If we try to increase TX power, this increases the AIR&GND signal and we should decrease the gain to avoid saturation.

The advantages of Monocoil are: minimal weight, minimal price, easy to build, no zones with back readings. However it
is the worst search head for metal detectors because generates maximal AIR&GND signal.

This thread is about..... Integrator Circuits for PI.


Not about how good induction balance detectors are, maybe you would like to start your own thread.

[mickstv]

Hi Eric, I tried the first circuit from your first post and compared it to the standard method if integrating. I noticed with my long sample periods the output was a-lot smoother using the track/hold integrator. More testing needed.

[mikebg]

From Mike "the design of an electronic device should start with analysis what is the most suitable block diagram". This was all I needed to see in 1966, a building block to perform an operation that I wanted. Then it was straight to the bench to try it out - and it worked. No long pages of calculations; that comes later when you want to find out how to optimise it. Even so, many things can just be tried on the bench much quicker. Eric.

Eric, the name of this block is DEMODULATOR (see my previous post) and this not an integrator, but specific case of sample - hold circuit. The DC output of this demodulator represents the "color" of target, but it is calculated by only two samples. This is not the best design solution.

[mikebg]

This thread is about..... Integrator Circuits for PI.


Not about how good induction balance detectors are, maybe you would like to start your own thread.

Mick, induction balanced search head is the most suitable search head for all types metal detectors including PI and BFO. Last year I started a thread for BFO design and soon will continue with analysis of serching oscillator for deep targets.

[Qiaozhi]

Mick, induction balanced search head is the most suitable search head for all types metal detectors including PI and BFO. Last year I started a thread for BFO design and soon will continue with analysis of serching oscillator for deep targets.

Are you able to post some photos and design data for a metal detector that you have actually designed and built?
Or does this only exist on paper and/or in a SPICE simulation?

[deemon]

I've not had a problem with the exponential self inductance signal (as I prefer to call it). Air signal is a bad description as the same would happen in a vacuum, and provided the TC is <5th of the TC of the target, then cut off is almost as good as instantaneous.

Of course , from the target's "point of view" - this exponential pulse almost as good as an ideal pulse . But for receiver electronics it isn't I mean that annoying exponentially decaying coil discharge "tail" in classic PI devices .... although it's must decay faster than target's response , it can be a problem when we push up amplification . And this tail don't allow us to sample closely to the flyback beginning , as you know ..... this is why I decided to completely discard it , changing the exponential flyback pulse shape to half-sine .

[mikebg]

Are you able to post some photos and design data for a metal detector that you have actually designed and built?
Or does this only exist on paper and/or in a SPICE simulation?

George, if you have not designed and built yet a BFO type metal detector with induction balanced search head, you will soon get that chance. I will do an analysis of deep searching oscillator without to use SPICE. The oscillator will respond only to targets buried deep in bad soil; it will ignore shallow and surface targets. For experiments you need poor soil, but at first can be used a plastic bowl filled with salt water. See you in that thread!

[Qiaozhi]

George, if you have not designed and built yet a BFO type metal detector with induction balanced search head, you will soon get that chance. I will do an analysis of deep searching oscillator without to use SPICE. The oscillator will respond only to targets buried deep in bad soil; it will ignore shallow and surface targets. For experiments you need poor soil, but at first can be used a plastic bowl filled with salt water. See you in that thread!

Mike - OK, I look forward to seeing the practical results you obtain from your calculations.
I guess you don't happen to have any real world results to show at the moment?

[WM6]

Hi mikebg,

this is your solution, ready to use, from our China friends:

IB BFO capable even to discriminate Ferro - non Ferro:

http://www.alibaba.com/product-gs/60...ch_Golden.html

[sinclairuser]

Hi mikebg,

this is your solution, ready to use, from our China friends:

IB BFO capable even to discriminate Ferro - non Ferro:

http://www.alibaba.com/product-gs/60...ch_Golden.html

hi all if you live in the uk, i would discourage you from ordering MDs from here, apart from the fact that 3-5 days is a total lie(if in the uk it is).
Also check the link ive posted this MD is clearly designed to deceive, this site is littered fisher f2/f4 lookalikes garrett rippoffs and thats the point.
Dont spend money on products like this, because if customs are on the ball you wont even get your MD, it will be confiscated and destroyed!.
So while this site maybe ok for some supplies, there is stuff on there that wont get through customs.
http://www.alibaba.com/product-gs/84...G_GPX4500.html