Announcement

Collapse
No announcement yet.

Problem with Surf Pi Pro - SURF XX PCB

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

  • Problem with Surf Pi Pro - SURF XX PCB

    I have problem with Surf Pi Pro that I have constructed .
    It is working very badly, low sensitivity - few cm with strong magnet,
    or only by touching with large iron object.

    All voltages are there...
    12V +5V -5V
    Ajusted voltage on pin 6 on NE5534 to 0 V ... tried -0,5V and + 0.5V
    Used working coils of various inductivity... from 220 to 385 uH

    I have no start beep sound that I have on working surf pi pro.
    I am using all components as always except two capacitors
    C5 and C6 both 220nF in original schematics ...
    on my PCB they are 330nF

    Using 12V battery 1.3 Ah ...
    Tried changing all ICs with no luck.

    Please help.

  • #2
    Sorry to hear it.
    Definitely something wrong with it.
    I worried that at some point you will need oscilloscope to see on what stage signal has lost.

    But first try to replace "DELAY" pot with 50K pot (or add 20K-30K resistor in line) and try to detect with different longer delay settings.
    PI PRO has a very big gain and very short delay so integrator can be saturated easily and detector stop detecting especially if coil (and cable) is slow.
    Even foil shield in coil can create overload of integrator if foil used for shield instead of graphite.

    To know if integrator overloaded or not you can measure integrator output by multimeter.
    LM358 pin7 has to be + 1-3 volts to ground.
    Also check is it changing in presence of target. Sure keep coil away from any metal.
    If you see more than 3.75 volts and it stays unchanged so it is overload.

    Comment


    • #3
      Originally posted by Waikiki_Sweep View Post
      Sorry to hear it.
      Definitely something wrong with it.
      I worried that at some point you will need oscilloscope to see on what stage signal has lost.

      But first try to replace "DELAY" pot with 50K pot (or add 20K-30K resistor in line) and try to detect with different longer delay settings.
      PI PRO has a very big gain and very short delay so integrator can be saturated easily and detector stop detecting especially if coil (and cable) is slow.
      Even foil shield in coil can create overload of integrator if foil used for shield instead of graphite.

      To know if integrator overloaded or not you can measure integrator output by multimeter.
      LM358 pin7 has to be + 1-3 volts to ground.
      Also check is it changing in presence of target. Sure keep coil away from any metal.
      If you see more than 3.75 volts and it stays unchanged so it is overload.
      Thank you for your help.
      I had already tried 15k resistor in place of R25 and delay pot 50k
      About same results... very small advance
      I don't have any shield on my coils. All are working more or less with another working Surf Pi Pro.

      Maximum voltage that I have on pin 7 of LM358 is 0,20 V
      Minimum is 0 ... with no metal close

      My concern is no sound on startup - all my surf pi pro detectors had that beep

      When I measure on pin 11 of CD4066 I have loud signal in speaker (piezo).
      Trimmer R2 5k is on 3k5

      Dumping resistor is heating up... probably .. not serious but "hotter"
      Last edited by sdesic; 01-27-2014, 06:52 PM. Reason: added something

      Comment


      • #4
        If you have 0.2 volts max on integrator output then signal delivered to integrator is 100-1000 times weaker than it should be.
        Do you have any changes in voltage on preamp output that you have set to zero earlier if you move some big metal target near coil?

        Comment


        • #5
          Originally posted by Waikiki_Sweep View Post
          If you have 0.2 volts max on integrator output then signal delivered to integrator is 100-1000 times weaker than it should be.
          Do you have any changes in voltage on preamp output that you have set to zero earlier if you move some big metal target near coil?
          Measured 0 V on pin 6 of 5534 with no metal close
          and with large iron -0,2 - +0,2 V
          and with strong magnet up to -1V and + 0,5 while waving it over coil

          Comment


          • #6
            Measure current consumption of your detector. Fact that it responds to magnet only suggest that TX pulse is missing due to some reason. Combined with damping resistor heating up, maybe MOSFET is faulty, or something wrong around 3904, misplaced or faulty components etc. Use piezo to test 555 output, 3904 c and e, MOSFET gate, drain...steady signal must be audible on 4093 outputs.

            Comment


            • #7
              I have just tried my working detector and have NE5534 (pin 6) output -0.1V change for a quarter ($0.25) and -0.5V change for a pliers if I put it straight on the coil.
              Can you get similar results?
              If preamp works OK in your detector but signal does not go to integrator. Than something wrong with CD4066 switches or timing CD40193.
              Did you inserted those chips correctly? Sometimes it happens to turn it on 180 degrees.

              Comment


              • #8
                Also, yes. Current consumption can show a lot.
                Usually for Surf PI it is 55-75 mA where schematic consuming 20-25 mA and coil rest of it.
                If you disconnect coil you have to see approx 30-50 mA less current consumption than with coil connected.

                Comment


                • #9
                  Originally posted by Tepco View Post
                  Measure current consumption of your detector. Fact that it responds to magnet only suggest that TX pulse is missing due to some reason. Combined with damping resistor heating up, maybe MOSFET is faulty, or something wrong around 3904, misplaced or faulty components etc. Use piezo to test 555 output, 3904 c and e, MOSFET gate, drain...steady signal must be audible on 4093 outputs.
                  Results for current consumption is 52mA for detector without metal close by and 50mA (less ? ) with magnet close by
                  and signal in speaker.
                  Pin 3 of TLC555 and gnd with piezo speaker gives loud beep
                  Emitter of 2N3906 gives loud signal too, collector ... crackle or no sound from piezo connected to GND and pins ...
                  Gate of IRF9640 same loud signal and drain too...

                  What pins are output from CD4093 ? pins 10 and 11 or 3 and 4 ?

                  Comment


                  • #10
                    Originally posted by Waikiki_Sweep View Post
                    I have just tried my working detector and have NE5534 (pin 6) output -0.1V change for a quarter ($0.25) and -0.5V change for a pliers if I put it straight on the coil.
                    Can you get similar results?
                    If preamp works OK in your detector but signal does not go to integrator. Than something wrong with CD4066 switches or timing CD40193.
                    Did you inserted those chips correctly? Sometimes it happens to turn it on 180 degrees.
                    I get -0,520V for quarter 0,25$ year 1999 left on coil which is 380uH
                    and -1,150V for pliers touching coil

                    starting with -0,002V
                    both 14pin ICs are "looking" to left as in parts layout schematics

                    Comment


                    • #11
                      Originally posted by Waikiki_Sweep View Post
                      Also, yes. Current consumption can show a lot.
                      Usually for Surf PI it is 55-75 mA where schematic consuming 20-25 mA and coil rest of it.
                      If you disconnect coil you have to see approx 30-50 mA less current consumption than with coil connected.
                      Current consumption with no coil is 17mA

                      Comment


                      • #12
                        Changed IRF9640 with new .. same results
                        Will try to change 2N3906 also an report back

                        Comment


                        • #13
                          Problem solved ... it was 2N3906 or 2N and IRF together ...

                          Thank you Waikiki Sweep and Tepco.

                          Comment

                          Working...
                          X