Announcement

Collapse
No announcement yet.

felezjoo PI(the best pulse induction metal detector that I made until now)

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

  • Originally posted by green View Post
    No target simulation. Change R11 to vary target time constant. L/R, 4uH/.1 ohm=40usecTC about the same as an aluminum soda can. Change L1(C par)to see coil resonance effect. Change R5 to damp coil with different C par. Playing with C par in spice, I'm guessing coil capacitance is a big reason coils act different. Adjusting R damping(R5) should solve the problem.

    Perfect thank you.

    Comment


    • Originally posted by green View Post
      Do you have a scope? I would be interested in seeing amplifier out, similar to reply 1207 with the two different amplifiers(LF357 and LME49990). It might be easier to answer your question with the scope pictures.
      I have scope but i can't take the output signal of LME49990 because it don't work at all.
      I soldered two LME49990 on adaptor to DIP but don't work, maybe ics are fault. I have another one if it will work i will connect the scope.
      Anyway will inform....

      Comment


      • Originally posted by green View Post
        No target simulation. Change R11 to vary target time constant. L/R, 4uH/.1 ohm=40usecTC about the same as an aluminum soda can. Change L1(C par)to see coil resonance effect. Change R5 to damp coil with different C par. Playing with C par in spice, I'm guessing coil capacitance is a big reason coils act different. Adjusting R damping(R5) should solve the problem.
        Good guessing!

        Comment


        • Originally posted by Geo View Post
          I have scope but i can't take the output signal of LME49990 because it don't work at all.
          I soldered two LME49990 on adaptor to DIP but don't work, maybe ics are fault. I have another one if it will work i will connect the scope.
          Anyway will inform....
          If the output stays above 10 volts during sample times it won't work(A-D full scale). A high enough bias current could cause this. Shunting R13(18k, + input of amplifier) resistor to lower the output to around 9.5 volts during the sample period should allow it to function. The spec. bias current isn't high enough but it's something to check. Actually the A-D input volts(junction R7 R8 needs to be less than AVCC), component tolerances could cause you to be closer to the over scale point where bias current might cause a problem. Just guessing what might cause the problem.

          Comment


          • Originally posted by green View Post
            If the output stays above 10 volts during sample times it won't work(A-D full scale). A high enough bias current could cause this. Shunting R13(18k, + input of amplifier) resistor to lower the output to around 9.5 volts during the sample period should allow it to function. The spec. bias current isn't high enough but it's something to check. Actually the A-D input volts(junction R7 R8 needs to be less than AVCC), component tolerances could cause you to be closer to the over scale point where bias current might cause a problem. Just guessing what might cause the problem.
            Very good analisys!
            If it's the case, it could explain also critical differences of Felezjoo circuit with some LF357: datasheet report 30 to 200 pA input bias current range.
            Give to a trimmer a chance!

            Comment


            • Finally i made some tests....
              3 new LME49990 are failed.... When i supply them they consumption 50ma!!! and don't give any output signal.
              Anyway let them...
              I connected the scope at out of front end IC and i took photos with different ics. I took 2 photos for every ic, one without any metal near the coil and one with a soda can 5cm near to coil.

              Here is photos from LF357
              Attached Files

              Comment


              • Below is a AD797 and finally is a LF356.
                It is clear that the LF357 is the ""winner"""
                Attached Files

                Comment


                • Can you do the same with LM318?

                  Comment


                  • Try CA3130, works better than fake LF357

                    Comment


                    • I tried it.
                      It has about the same results as original LF357 but it gives many random beeps.
                      Anyway i can check it with scope.....

                      Comment


                      • Originally posted by ivconic View Post
                        Can you do the same with LM318?
                        Ivica i had one piece. If will find it then no problem to check it at scope
                        At any case will inform

                        Comment


                        • Originally posted by Geo View Post
                          Finally i made some tests....
                          3 new LME49990 are failed.... When i supply them they consumption 50ma!!! and don't give any output signal.
                          Anyway let them...
                          I connected the scope at out of front end IC and i took photos with different ics. I took 2 photos for every ic, one without any metal near the coil and one with a soda can 5cm near to coil.

                          Here is photos from LF357
                          Thanks for the tests and the scope pictures.

                          Comparing your no target picture with a spice no target analysis, it looks like your circuit is over damped. Including a spice analysis with a lower coil capacitance to give a decay similar to yours. Changed damping resistor(R5 spice, R12 felezjoo)to critical damp. Wondering if adjusting the damping changes the detection distance. If I had a Felezjoo circuit I would do a test. First, record the detection distance for an aluminum soda can(bottom end toward coil, 40usec TC) and a piece cut from the side of a can(1x1 inch, 5usec TC). Second, adjust the damping resistor to see how close to the spice picture I could get. Third, repeat the target distance measurements. If you or anyone tries the test I would be interested in the results.

                          Thinking need to keep the critical damped no target decay below 9.5volts, underdamped(overshoot) might cause a problem
                          Attached Files
                          Last edited by green; 12-16-2016, 02:20 PM. Reason: added sentence

                          Comment


                          • Originally posted by nae View Post
                            Try CA3130, works better than fake LF357
                            Yesterday i compared the 3130 with the 357. Both have about the same depth.
                            I think that 3130 gave a bit better waveform than the LF357 but at soda can the LF357 detected it at 2cm more.
                            At coin 1 euro i did n't see any diffference.

                            Ivica,,,, sorry but i did n't find a LM318 for test...

                            Comment


                            • Originally posted by green View Post
                              Thanks for the tests and the scope pictures.

                              Comparing your no target picture with a spice no target analysis, it looks like your circuit is over damped. Including a spice analysis with a lower coil capacitance to give a decay similar to yours. Changed damping resistor(R5 spice, R12 felezjoo)to critical damp. Wondering if adjusting the damping changes the detection distance. If I had a Felezjoo circuit I would do a test. First, record the detection distance for an aluminum soda can(bottom end toward coil, 40usec TC) and a piece cut from the side of a can(1x1 inch, 5usec TC). Second, adjust the damping resistor to see how close to the spice picture I could get. Third, repeat the target distance measurements. If you or anyone tries the test I would be interested in the results.

                              Thinking need to keep the critical damped no target decay below 9.5volts, underdamped(overshoot) might cause a problem
                              Thanks Green, i will try to make a test with diff damper.
                              At begining of this project i changed the damper but i did not see any diff, but now there is other coil and better frond end ic.

                              Comment


                              • Originally posted by Geo View Post
                                Yesterday i compared the 3130 with the 357. Both have about the same depth.
                                I think that 3130 gave a bit better waveform than the LF357 but at soda can the LF357 detected it at 2cm more.
                                At coin 1 euro i did n't see any diffference.

                                Ivica,,,, sorry but i did n't find a LM318 for test...
                                Hi Geo. Did u tried with LF357H (metal case)?

                                Comment

                                Working...
                                X