Announcement

Collapse
No announcement yet.

IGSL

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

  • Originally posted by detecto View Post
    Could some one post here pinout for original minelab musketer TS-1000 coil.
    Thank you
    Regards
    Detecto
    hi detecto
    look at this post.
    http://www.geotech1.com/forums/showthread.php?t=18550
    I think that might help

    (geogle translator use)

    Comment


    • Lately I noticed Ivconic speaking low of his own design:
      Originally posted by ivconic View Post
      When i said "crap" i ment it is just another homebrew in a very long line of similar: TGSL, Wave, IGSL, Volksturm, Anker, SMW, Caesar, Pirate, Golden Mask....etc...etc...etc
      As a proud future owner of this particular gadget I threw my evil eye on this design and my evil ear to youtube videos of it in action, and I think I have some clues of what Ivconic means. There are some details that do not make much sense, but can be cured with minor changes. Just a touch here and there.
      First to go are channel gain stages. They are all the same and they all do fine with very small signal, but they saturate easily, and phases are completely screwed and dependent upon the channel signal level. Hence jerky responses, especially with close by or large targets. I'd say this design (TGSL) was copied from some low gain one and then further "improved" by pumping up gain - it makes more sense at lower gains per stage. I have some solutions to fix it and make it good.
      The other possibility for improvement is an input gain stage that can be easily transformed to a true differential amp and further enhance the "wet grass" wiring solution.

      I must rush now.

      Comment


      • It is not "low" speaking nor spitting on my own design (Qiaozhi was precedent contributor on it too), also not on other homebrews and designers too.
        It was just objective approach with using metaphore (so clumsy done by me, that's because my everyday English is not so good).
        I wanted to be objective as far as i can, that's why i have to be first to criticize my own work.
        IGSL is excellent idea, but not done on perfect way.
        I am sure there are many things to be improved at it.
        At the time i was in a hurry, to finish it and to check it. Later i had other responsibilities and i left it the way it is.
        Posting IGSL project on these forums was best i could do. All the unfinished jobs at it will be done here, because here is the best possible place to do that.
        We have here numerous bright people with good enthusiasm and that's was my main motif to post it here as free project. I knew that sooner or later IGSL will be improved and done on much better way by other people here.
        Same thing happened with TGSL in the past. Such positive feedback and overall experiences pushed TGSL to much better machine than it was at the begining.
        One man band is not good thing. I learned that much more before.
        That's why i will always rather share and try to learn from others.
        I am very glad to see you Davor to pay such interest in IGSL.
        Also am very glad to see that you are already having ideas how to correct its lacks.
        I will be very happy to see better solutions at it.
        You are already right considering channel gain stages. All the problems are starting there.
        Need better optimizing.
        Next issue would be power consumption. 60-70mA is way over i expected. Not bad but could be much better.
        And last but not least is audio.
        I had idea to put two VCO's instead.
        Each block to be fed into VCO.
        But last year and a half i had so much other things to do and i had to leave IGSL as it is.
        Maybe next winter i will have more spare time to continue working on it.
        Meanwhile i will follow this topic with huge attention.
        Cheers!

        Comment


        • MikeBG gave nice TX oscillator with pid controler somewhere on this forum.
          I planned to incude it on next IGSL revision.
          Just wanted to point on that.
          Cheers!

          Comment


          • I had hoped for more depth with this kit.


            I had not realised the absolute critical nature of the coil nulling. Mine was not great, so maybe me.


            Some times I get an oscillating phase shifter - and multiple gating of ths synch det. Stubbon recovery.

            Possibly due to the phase shifter plus inverting buffer - so not gain / phase stable. One channel was ok though maybe partly layout?


            When working its fun, but I never got a coin or musket shot or small brass from more than 4" deep in UK soil. Most stuff in the turf roots.

            Yes plough blades at 8" - but Im not a farmer!


            What is the deepet descriminating kit on here??

            S

            Comment


            • Originally posted by golfnut View Post
              I had hoped for more depth with this kit.


              I had not realised the absolute critical nature of the coil nulling. Mine was not great, so maybe me.


              Some times I get an oscillating phase shifter - and multiple gating of ths synch det. Stubbon recovery.

              Possibly due to the phase shifter plus inverting buffer - so not gain / phase stable. One channel was ok though maybe partly layout?


              When working its fun, but I never got a coin or musket shot or small brass from more than 4" deep in UK soil. Most stuff in the turf roots.

              Yes plough blades at 8" - but Im not a farmer!


              What is the deepet descriminating kit on here??

              S

              It is not that "shallow" ! Cheeese no!
              Something you done wrong there.
              It can detect 2cm coin at 30cm depth in medium soil.
              Larger items up to ~50cm and more in soil.
              Yes, coil nulling is very delicate and critical.
              That's why i offered second IGSL version, adapted to suit original Minelab TS800 and TS1000 coils - to save people from coil making horror!
              Along with mentioned Minelab TS coils; newer Tesoro coils will also fit nicely to IGSL ver2.
              Those with TX=~1mH and RX=~15mH... (Tejon, Vaquero.....etc..)

              Comment


              • "...What is the deepest descriminating kit on here??.."

                Whatever you pick - you will always face same coil nulling problem!
                Same story....

                Comment


                • Thx for the Feedback


                  I maybe try another coil - when I get my shells off Georgi.

                  S

                  Comment


                  • Originally posted by golfnut View Post
                    Thx for the Feedback


                    I maybe try another coil - when I get my shells off Georgi.

                    S

                    I didn't payed much time and space here to elaborate IGSL coil making and nulling because it's been already done very well in TGSL topic in the past.
                    Everything what matters to TGSL coil making is also aplicable to IGSL coil making.
                    IGSL ver1 is sharing the very same coil as TGSL.
                    IGSL ver2 is adapted for some original coils, like i mentioned: Minelab TS800, TS1000 and all newer Tesoro coils with ~1mH and ~15mH specs.
                    Latelly i also discovered more suitable coils for IGSL, i think i mentioned somewhere here; coils for Minelabs FT and GT (16000,17000 and 18000)...
                    Coil making and nulling is horror for me. I am usually spending 4-5 hours to achieve nice nulled coil. Not because it is generally hard task - but mostly because i do that in pretty offhand conditions, as amateur and diyer.

                    Comment


                    • Originally posted by Davor View Post
                      phases are completely screwed and dependent upon the channel signal level
                      Can you explain more about that?

                      I agree the demodulated gain/filter stages are interesting targets for improvements, which is why I would like to make a digital version of those stages where you can really play with idealized filters and find approximate analog equivalents. It may not be so easy to improve though -- those Tesoro tinkerers were good at getting a lot out of a few parts.

                      I have wondered if "pure time delay" filter profiles would help fatten the pulse to give a clearer signal, but you may pick up more unwanted noise also -- so I think there is a trade-off between noise and pulse shape.

                      What I think also is an area for improvement is the section including the final comparators that do the detection logic followed by the "pulse-width filter" (to reduce chatter?) that unfortunately strips some width off the detected pulse as a byproduct. It seems like an ad-hoc afterthought type of design.

                      Of course we know the design can be improved -- or Tesoro would have gone out of business. Improving without adding many more parts or complexity is my goal when tinkering with the TGSL, otherwise might as well work with a different model completely.

                      -SB

                      Comment


                      • First some answers...
                        Originally posted by ivconic View Post
                        MikeBG gave nice TX oscillator with pid controler somewhere on this forum.
                        I planned to incude it on next IGSL revision.
                        Just wanted to point on that.
                        Cheers!
                        No need. If you supply a well noiseless power supply and a symmetric rail to rail driver design, your oscillator will work fine. It will oscillate within the rails voltage, or a bit more, and its noise will be determined by the power supply noise. Building a whole PID control will make your oscillator do the same thing but well under the rail voltage. The only difference between a good and not so god design is in symmetry. The asymmetric ones are prone to PWM troubles.

                        Originally posted by simonbaker View Post
                        I agree the demodulated gain/filter stages are interesting targets for improvements, which is why I would like to make a digital version of those stages where you can really play with idealized filters and find approximate analog equivalents. It may not be so easy to improve though -- those Tesoro tinkerers were good at getting a lot out of a few parts.
                        You'd most likely end up building some 16bit++ ADC and some incredibly complicated DSP to do something that can be achieved in analog world just as well. The only real advantage of digital world against the analog is FIR filtering. It makes for sharp time response. And some phase stuff too. For everything else you'll find some equally good or better solutions in analog world.
                        If you really wish to do everything in digital world, check about the quadrature signal processing, and the Costas loop. IMHO it is the direct path to discrimination.

                        Now some explanations...

                        Comment


                        • Now about the gain stages. There are two opamps, first in non-inverting, and the second one in inverting configuration. Together they are inverting an input signal.
                          Business area of a comparator that follows are positive signals, so the expected input would be negative, and (without further elaborating IB theory) it is band pass filtered to ~10Hz peak. Bandpass means also motion compensation.
                          Input signals can vary in a wild span of several decades, and someone mentioned 6th power law response viz the target distance. Anyway, it is a lot to squeeze into a nice "beep".

                          To go along with the requirements of the task, I prepared a signal source that mimics a coil swinging response (10Hz) right after the bilateral switches (featuring phase detectors) in case of the expected beep, and stepped values are in octave relationship: 0.2mV, 0.4mV ... 200mV. Altogether 3 decades.
                          Attached Files

                          Comment


                          • Next is the existing solution and clues to what's wrong with it. First please note that the response of a target is squeezed to under 0.6V (diode, eh) and the comparators really work with that much, and the sensitivity control is really designed to work with this much. It is the same with TGSL.
                            Furthermore, the non-response part is saturating, and also widening in a process, which leaves active part ever shrinking with rising levels.
                            Picture speaks for itself...
                            Attached Files

                            Comment


                            • Now comes the solution that does not change much in the existing design. Actually it will require some changes further on, but lets play with this for the time being.
                              First of, the D2 polarity is wrong. When reversed the output swing is allowed to go beyond the diode barrier voltage on positive side, while the previously offending saturation is kept well under 0.6V. That's a start.
                              Next I wish to enable as much gain as possible, while maintaining a log relationship at the output. To do that I reverse a diode in a first stage. To avoid floating with rising signal (this is a motion compensation design) I need to introduce an additional diode in counter-polarity to the existing one. I checked it with various offsets and it always works good. Now the response signal shape looks good, and the output swing is ~6V, with no saturation, no shifting left or right, no shrinking of active response, and the best of all - it is logarithmically compressed. See how octave increase maps into linear output?

                              I'd say that whoever designed this stage had an inspiration, but somewhere in a process he/she made a small mistake.

                              And a completely unimportant one: 100k in series with the first stage is doing nothing but increasing noise. At least with LF347. Just get rid of it.

                              Please feel free to play with these designs. I'd say they are pretty elaborate, but for some unexplainable quirk in a process of implementation they just got wrong.
                              Attached Files

                              Comment


                              • Originally posted by Davor View Post
                                You'd most likely end up building some 16bit++ ADC and some incredibly complicated DSP to do something that can be achieved in analog world just as well. The only real advantage of digital world against the analog is FIR filtering. It makes for sharp time response. And some phase stuff too. For everything else you'll find some equally good or better solutions in analog world.
                                If you really wish to do everything in digital world, check about the quadrature signal processing, and the Costas loop. IMHO it is the direct path to discrimination.

                                Now some explanations...
                                I happen to like analog anyway - so I'm thinking a digital model would be handy to decide what analog filter you want. Could actually use a laptop with sound card a la Aziz's experiments and just probe into the TGSL and do the tinkering in software.

                                But If I live long enough I'll probably try some digital version also, trying to use the lowest end, low power processor possible.

                                -SB

                                Comment

                                Working...
                                X