Announcement

Collapse
No announcement yet.

PI limit to early sampling

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • PI limit to early sampling

    Hello everybody

    That's my first message on this forum so let me introduce myself briefly : I'm French (so my english is poor, sorry in advance if I'm not very clear), 58 years old, electronics engineer with a long experience of analog electronics design in transmission and avionics equipments.

    I have a recent interest in metal detectors particularly in Pulse Induction type but not for professional use just for fun. I have read some technical papers on this forum and on the net on this topic and I think I begin to understand how a PI detector works but I have many questions to go deeper ( in my understanding and maybe later deeper to find items buried ! )

    The first one is the following : Why, on every schematics I have seen so far, the first sampling after Tx off is so late, more than 10 us ? The designers wait for the signal to be very low then amplify it very hard and try to extract useful information buried in noise. Why don't sample sooner and amplify less to improve S/N ?

    Is it because just after Tx off, the signal level difference between "no-target" and "target" is too low to be detected ?

    I have seen also that to be able to sample early, it is necessary to have a "fast coil" (low parasitics capacitance). May I have some explanations about that because I don't understand why ?

    Thank you in advance for an answer from the experts board surfing on this forum !

    Philippe

  • #2
    The Tx coil runs a 1-3 amperes pulse and this current does not disappear instantly.

    Comment


    • #3
      Originally posted by Phiphi View Post
      I have read some technical papers on this forum and on the net on this topic and I think I begin to understand how a PI detector works but I have many questions to go deeper ( in my understanding and maybe later deeper to find items buried ! )
      I assume you are aware of ITMD?

      https://www.amazon.fr/s/ref=nb_sb_no...metal+detector

      Comment


      • #4
        I know this book but I haven't already bought it ...
        Do you mean that the answers to my questions are inside ?

        Comment


        • #5
          Originally posted by Phiphi View Post
          I know this book but I haven't already bought it ...
          Do you mean that the answers to my questions are inside ?
          Your question is "why not sample during the turn-off transient". It's because the transient is orders of magnitude stronger than the target's response and you'd get a horrible signal/noise ratio.

          Comment


          • #6
            Originally posted by Phiphi View Post
            I know this book but I haven't already bought it ...
            Do you mean that the answers to my questions are inside ?
            I think it would answer a lot of your questions.
            You can use the "Look Inside" feature on Amazon to view some of the contents.

            Comment


            • #7
              Originally posted by Phiphi View Post
              The first one is the following : Why, on every schematics I have seen so far, the first sampling after Tx off is so late, more than 10 us ? The designers wait for the signal to be very low then amplify it very hard and try to extract useful information buried in noise. Why don't sample sooner and amplify less to improve S/N ?
              A couple of reasons. As you try to push much below 8-10us, ground starts to become detectable. Also, as you push further up the decay curve, you have to run less gain to avoid saturation, and lose SNR in the process. Finally, the earlier you try to sample up the decay curve, the more sensitive everything gets to component variation (including the coil) and temperature shifts. In short, everything gets exponentially harder.

              If you want to sample early (even down to 0us), the easiest thing to do is build an IB coil.

              I have seen also that to be able to sample early, it is necessary to have a "fast coil" (low parasitics capacitance). May I have some explanations about that because I don't understand why ?
              The turn-off flyback & decay are due to the coil inductance and parasitics. The speed of the coil determines how fast its flyback is removed, leaving only the target signal.

              Comment


              • #8
                Thank you all for your answers. It's clearer for me now.

                What I understand is more or less what I was saying in my first post : between Tx off and, let's say 10 us after, it is impossible to distinguish no-target response to target response because the part due to coil is very strong. And the fastest is the coil, the sooner it will be possible to sample. Is this analysis seems correct for you ?

                If yes, one of the difficulty is to find the right sample time to maximize the difference between no-target response and target response. Am I right ?

                Comment


                • #9
                  Originally posted by Phiphi View Post
                  Thank you all for your answers. It's clearer for me now.

                  What I understand is more or less what I was saying in my first post : between Tx off and, let's say 10 us after, it is impossible to distinguish no-target response to target response because the part due to coil is very strong. And the fastest is the coil, the sooner it will be possible to sample. Is this analysis seems correct for you ?
                  Yes.

                  If yes, one of the difficulty is to find the right sample time to maximize the difference between no-target response and target response. Am I right ?
                  Yes, and it varies for different targets. The eddy response to a magnetic field step function is of the form t*e^(-t/tau), which has a peak at a time dependent on the target tau.

                  Comment


                  • #10
                    Originally posted by Carl-NC View Post
                    The eddy response to a magnetic field step function is of the form t*e^(-t/tau), which has a peak at a time dependent on the target tau.
                    Which can be seen very clearly in this simulation:
                    http://www.geotech1.com/forums/showt...228#post162228

                    Comment


                    • #11
                      Originally posted by Qiaozhi View Post
                      Which can be seen very clearly in this simulation:
                      http://www.geotech1.com/forums/showt...228#post162228

                      Here is an LTSpice simulation of equation 11 in John Alldred's paper:

                      .......... (Eq.11)
                      Thank you, interesting. Where can I find this John Alldred's paper ?

                      Comment


                      • #12
                        Don't search, I have found it in the referenced thread in post #687.

                        Comment


                        • #13
                          After a short but successful search on the Web, I have found these links on a french site : http://www.mantametaldetectors.com

                          [2] Lettre de F.B.Johnson en réponse à Éric Foster
                          http://www.mantametaldetectors.com/wpcontent/
                          uploads/2016/01/Johnson-letter001.jpg
                          [3] A method of detecting a mass of non-ferrous metal located
                          at depth in the earth. J.H.Westcoot Ph.D. Imperial College.
                          22rh febrary 1955
                          http://www.mantametaldetectors.com/wpcontent/
                          uploads/2016/01/Westcott002.pdf
                          [4] A pulsed bomb locator by F.B.Johson, M.A.
                          25th January 1956
                          http://www.mantametaldetectors.com/wpcontent/
                          uploads/2016/01/Johnson-FB005.pdf
                          [5] Von Claus COLANI « Ein neuartiges metallsuchgerat nach
                          dem pulsverfahren fur grofbe gelandeflachen mit
                          elektronisher objektanalyse und –auswahl »
                          08th june 1968
                          http://www.mantametaldetectors.com/wpcontent/
                          uploads/2016/01/Colani-1968_German.pdf
                          [6] E. J Foster « Futher developments of the pulsed induction
                          metal detector » 1968
                          http://www.mantametaldetectors.com/wpcontent/
                          uploads/2016/01/FOSTER.pdf
                          [7] George Payne patent US4507612 A / 1979
                          http://www.mantametaldetectors.com/wpcontent/
                          uploads/2016/01/US4507612.pdf

                          The site is in french but the papers linked are all in english but one from Colani in german.

                          I'm sure many of us will be interested ...


                          Comment


                          • #14
                            Originally posted by Phiphi View Post
                            [4] A pulsed bomb locator by F.B.Johson, M.A.
                            25th January 1956
                            The Pulsed Bomb Locator paper has been discussed on the forum here ->
                            http://www.geotech1.com/forums/showt...ohnson-s-Paper

                            Comment


                            • #15
                              OK. I hadn't seen it, there is so many topics on this forum !

                              I think this paper is very interesting because it gives formula useful to calculate the parameters of a PI system. This paper is just 60 years old !
                              Now, with modern electronics, we are more in the us range than in the ms range as in 1956 ...

                              Comment

                              Working...
                              X