Announcement

Collapse
No announcement yet.

Announcement

Collapse
No announcement yet.

Designing a highly sensitive digital PI detector

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

  • Designing a highly sensitive digital PI detector

    Hi! I’m very new to this whole metal detector thing. My experience with metal detectors begins and ends with a toy that came with a magazine I got when I was about ten. From memory, it could barely detect a car from about an inch away.

    I have an application that may benefit from a metal detector - I’m trying to detect a metallic particle within a cylindrical volume approximately 50 mm in diameter and 50 mm deep. The volume is filled with a mildly conductive saline solution (roughly 1 S/m). The coil must be static, placed at one end of the cylinder and cannot be moved during the measurement. I’m able to sample for several seconds if needed to help minimise noise. I could also perform a calibration on a known sample before measurement if required.

    Probably the most challenging aspect is that the particles I need to detect can be as small as 0.2mm³. Most particles are ferrous, and I don’t need any information other than the presence or absence of metal.

    The coil geometry is also limited - the diameter can’t be much wider than the detection cylinder. The coil geometry and conductive medium are pushing me towards a PI detector with a mono coil (hence posting in this section), but I’ve heard they aren’t as good as VLF for small targets.

    I’m a lot better at the digital/software side than analog (every idiot can count to 1). As schematics with more than a handful of op-amps scare me, my instinct is to do as much digitally as possible. I plan to have a minimal analog frontend - just enough to get the signal into an ADC without introducing too much noise.

    Does this sound like something that’s possible? I’d really appreciate any tips/pointers/information that may help. I’ve been binge-reading anything I can find for the last couple of weeks, but I’m sure there’s plenty of useful information I’ve missed.

    Thanks in advance!

  • #2
    Hi Nogtail,
    An object with a volume of 0.2mm^3 is not so small (1x 1x 0.2mm), so detecting such target in two inches, while challenging, is not an impossible task. You could use a (gold prospecting) PI metal detector, preferably in static or low-motion mode (if there is slight or no movement of the targets during measurement). The fact that the search coil has a small diameter (two inches) will be beneficial for detection, allowing you to make a highly sensitive coil (It is clear that the measuring chamber should not be made of metal).
    ​For iron particles, you could also combine it with a constant magnetic field, which would give you a certain concentration of pieces in the area of ​​the magnet. If these targets move along with the fluid, you could measure/detect the change in the magnetic field at the farthest point of the measurement cell.

    There are, of course, thousands more ways, but they depend on the specific situation.

    Here in the forum, there are excellent specialists, and I'm sure they will help you with specific recommendations and schemes.

    Comment


    • #3
      Hi Nogtail,
      For your first post it would be better to be more modest. Your words "any idiot can count to 1" about analog solutions are deeply wrong - successful analog solutions also need serious in-depth knowledge, as well as successful software solutions. If analog solutions are possible for any idiot - why do you avoid them if they are so easy?​​

      Comment


      • #4
        Hi Nogtail,
        The problem with recognition of a metal particles with volume 0.2mm^3 in the volume with 50mm diameter and 50mm high is very serious. My experience in area of high sensitive pulse MD for pin-pointer units - many tests with coils with 1 or 2 inch diameter coils gives results for 2inch distance only for metal parts with 1grain weight (appr. 0.056g) - far from your aim! Using of magnet for concentrate of the particles near to sense coil is not useful idea - you search for not only ferrous metal particles. I'm not sure that high sensitive PI MD will do the work.

        Comment


        • #5
          Originally posted by Detectorist#1 View Post
          Hi Nogtail,
          For your first post it would be better to be more modest. Your words "any idiot can count to 1" about analog solutions are deeply wrong - successful analog solutions also need serious in-depth knowledge, as well as successful software solutions. If analog solutions are possible for any idiot - why do you avoid them if they are so easy?​​
          Sorry, it was intended as quite the opposite. "Every idiot can count to one" is a quote from Bob Widlar about how comparatively easy digital is:

          Click image for larger version

Name:	Bob Widlar Digital.jpg
Views:	73
Size:	422.7 KB
ID:	448307

          Comment


          • #6
            Hi Nogtail,
            I accept your apology. Bob Widlar is my god in analog technology.

            Comment


            • #7
              Regardless of the tech used, you will need to calibrate on a clean sample, maybe each time. If a PI is used it will need to be aggressive, maybe sampling down to 1us. But I would lean toward a high frequency VLF, maybe 100kHz. I think it would be easier to achieve.

              Yes, Widlar was quite the character. I like the pic.

              Comment


              • #8
                In my opinion.... a target of size 1x1x0.2 mm made of medium and high conductive material... is too small a target for a PI detector...and moreover if such a target is made of lead - low conductor... then the PI detector will not detect it at all....
                My testing of ATX with a 3x7" coil from Infinium..../ATX is one of the really sensitive PI detectors / shows where the detection limits are for very small targets..
                Such small targets can be detected to a certain extent by a VLF detector with a frequency above 40khz and small coil, best from 70khz and more khz with a small coil... ​

                airtest. ATX with a small coil 3x7" DD Infinium ..test detector without GB..
                Attached Files

                Comment


                • #9
                  Hi Nogtail,
                  Try to find the scheme of Falcon MD8 or MD20 pin-pointers on geotech1.com site. They are VLF type and the working frequency of these pin-pointers is 300KHz. These pin-pointers are advertised as capable to find very little metal particles ( almost invisible). It is good that your task not need EFE and GB features.

                  Comment


                  • #10
                    Hi Nogtail,
                    Falkon MD20 is "Energy theft" type, not VLF type. Nevertheless, try to use this scheme for your purpose.
                    Attached Files

                    Comment

                    Working...
                    X