Announcement

Collapse
No announcement yet.

THE PERFECT GROUND BALANCE

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

  • #16
    Originally posted by Carl-NC View Post
    Bipolar pulsing.
    Bi-polar TX pulses.

    Ok, let's look at the TEM TX pulses in the picture below. About 2.5A positive pulse and then 2.5A negative pulse. I added the blue 0A trace to make it easy to see.

    Compare with monopolar pulses from a traditional PI. The pulses look much the same, but they are only positive. Again the blue trace shows the 0A level.

    Does the TEM TX qualify?

    Tinkerer
    Attached Files

    Comment


    • #17
      With PI signal GB is a trivial thing and it is a moment when PI pulse collapses. Your stop-watch starts at that moment. Unfortunately you can't start sampling at that very moment due to ringing etc.

      Comment


      • #18
        Originally posted by Davor View Post
        With PI signal GB is a trivial thing and it is a moment when PI pulse collapses. Your stop-watch starts at that moment. Unfortunately you can't start sampling at that very moment due to ringing etc.
        Could you expand a little on that?
        I have no problem at all sampling anywhere along the whole TEM TX wave form. Anywhere, anytime.

        Tinkerer

        Comment


        • #19
          OK, the whole idea is in orthogonality.

          As RF engineers we are mostly inclined to think of periodic signals, but PI is a different animal, and I'll have to draw some parallels.

          For some reason I can not fathom someone christened a 90° sampling signal a "ground balance" and it stuck. It is related to ground influence because oscillation conditions of a free running oscillator are changing as coil nears the ground, but it is still a 90° signal.
          Target responses are shifted from Tx signal by a degree dictated by target tau, and the idea is that multiplying a Rx signal with the one orthogonal against Tx you'll have only target response, and the remaining non-ballanced Tx->Rx signal will cancel out, thus broadening the dynamic range of the targets' response.

          How that relates to PI? Not at all. PI Tx is in great deal orthogonal against PI Rx because Rx starts when Tx stops. Theoretically they do not co-exist. IB PI schemes simply shorten the transition period by means of isolating Tx from Rx, thus making Tx decay ringing less intense, but for all practical purposes sampling "T zero" begins at the moment Tx signal collapsed. Additional delay is due to technology limits, but GB point is Tx collapse.
          In practical terms that would be a moment when you can distinguish Rx signal as not being affected by Tx artefacts.

          In case you gate a PI pulse by a microcontroller, your sample is constrained by the microcontroller clock period, and PI pulse duration. In case of inductance change, say ground proximity, this time variation of PI pulse could be a problem, so it would be better to work this out in some feedback fashion, e.g. detecting a Tx voltage collapse and supplying this as indication of GB, thus triggering counters.

          So perfect timing would require Tx signal voltage near-zero-reaching criterion.

          Comment


          • #20
            Originally posted by Tinkerer View Post
            Bi-polar TX pulses.

            Ok, let's look at the TEM TX pulses in the picture below. About 2.5A positive pulse and then 2.5A negative pulse. I added the blue 0A trace to make it easy to see.

            Compare with monopolar pulses from a traditional PI. The pulses look much the same, but they are only positive. Again the blue trace shows the 0A level.

            Does the TEM TX qualify?

            Tinkerer
            It does and it doesn't. It's bipolar but not symmetrical. The positive pulse rises slowly then falls sharply wheres the negative pulse starts with fast rise and finishes slowly. In addition the pulses are not independent ie. starting from a point of zero coil current flow or with a settled target. So it won't be as simple as inverting the negative pulse samples and averaging them with the positive ones. However I tend to think the information is still there it just needs a slightly more complicated method of extraction.

            Comment


            • #21
              Not bipolar at all, in the sense of the magnetic field. Below are 4 pulse methods; the top two are not bipolar, the bottom two are. The current min/max values are irrelevant; the top pair of pulses are identical as far as target response goes. Same with the bottom pair of pulses.
              Attached Files

              Comment


              • #22
                TEM phase shift

                Originally posted by Midas View Post
                It does and it doesn't. It's bipolar but not symmetrical. The positive pulse rises slowly then falls sharply wheres the negative pulse starts with fast rise and finishes slowly. In addition the pulses are not independent ie. starting from a point of zero coil current flow or with a settled target. So it won't be as simple as inverting the negative pulse samples and averaging them with the positive ones. However I tend to think the information is still there it just needs a slightly more complicated method of extraction.
                Target extraction:

                Below is the unfiltered TEM RX signal.

                The red trace is NO TARGET
                The green trace is the GOLD TARGET
                The blue trace is the FERRITE TARGET

                We can observe the variation of target amplitude as well as the phase shift.

                Does the TEM qualify as a PI-VLF hybrid?

                Tinkerer

                Comment


                • #23
                  TEM phase shift picture

                  Originally posted by Tinkerer View Post
                  Target extraction:

                  Below is the unfiltered TEM RX signal.

                  The red trace is NO TARGET
                  The green trace is the GOLD TARGET
                  The blue trace is the FERRITE TARGET

                  We can observe the variation of target amplitude as well as the phase shift.

                  Does the TEM qualify as a PI-VLF hybrid?

                  Tinkerer
                  Sorry, forgot to add picture.
                  Attached Files

                  Comment


                  • #24
                    Bi-polar or not

                    Originally posted by Carl-NC View Post
                    Not bipolar at all, in the sense of the magnetic field. Below are 4 pulse methods; the top two are not bipolar, the bottom two are. The current min/max values are irrelevant; the top pair of pulses are identical as far as target response goes. Same with the bottom pair of pulses.
                    Looking at your drawings, I would agree. A pulse that goes from +V to -V would be monopolar because the current only goes in one direction.

                    However, the strange thing with the TEM method, is that the current in the coil does go in both directions.
                    Or does it not?
                    Am I mistaken?

                    The Flyback capacitor absorbs the Flyback and then sends the current through the coil back to the battery, or in this case the capacitor bank.

                    I added the LTSpice simulation so you can have a closer look at it.

                    Tinkerer
                    Attached Files

                    Comment


                    • #25
                      Yes, but the current return is not at the same slew rate as the flyback. Ergo trying to match a flyback sample with a recharge sample will give radically different results. You need a positive flyback matched with a negative flyback.

                      Comment


                      • #26
                        Also due to the charging cycle that begins right after pulse, this design has a shifted zero that has to be compensated. True bipolar pulses use no current in pause between them, and that is a beauty of it. Instead they use high voltage charged capacitors to supply the charge.

                        I just love that concept.

                        My exciter goes to that direction but with unipolar pulses. I intend to explore it a bit further.

                        Comment


                        • #27
                          With bipolar pulsing target response is bipolar as well, but Earth field response is always unipolar. So subtracting the responses cancels Earth field, and adds target responses.
                          Attached Files

                          Comment


                          • #28
                            The differential front end ( and coil ) does it all in the same pulse ..... the earth field shows up as a common mode voltage which is rejected by the diff amp CMMR in the front end. Target signals are differential voltage changes.

                            this is one of the first posts on Geotech showing the waveform ....
                            http://www.geotech1.com/forums/showp...0&postcount=48

                            there is a trick to the sample and timing though.

                            moodz.
                            Last edited by moodz; 03-05-2012, 12:20 AM. Reason: typo 2

                            Comment


                            • #29
                              Originally posted by moodz View Post
                              The differential front end ( and coil ) does it all in the same pulse ..... the earth field shows up as a common mode voltage which is rejected by the diff amp CMMR in the front end.
                              Is that true? I haven't tested an autotransformer coil for Earth field, but I would expect it to need bipolar pulsing as well.

                              Comment


                              • #30
                                Originally posted by Carl-NC View Post
                                Is that true? I haven't tested an autotransformer coil for Earth field, but I would expect it to need bipolar pulsing as well.



                                On the diff front end I build it never cancelled out Earth field effect, the only way was with using a second sample, so it was really no different than a standard mono frontend.

                                Comment

                                Working...
                                X