Announcement

Collapse
No announcement yet.

Photo of GP Extreme Badness?

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

  • #31
    Originally posted by bklein View Post
    A BSS84 is a SOT23 P channel. Those two fets are SOT89 N channel. Their drains are going to the coil. So are you describing other parts?
    OK what you are describing are the first two their gates go to a BC857 which has a 100R between emitter and base what I was describing are the next along the chain hope that helps.

    PM me.

    Regards, Ian.
    Last edited by IBGold; 03-19-2012, 03:39 AM. Reason: Addition

    Comment


    • #32
      Yes, I get it now. Takes me awhile sometimes.

      Comment


      • #33
        Hi Bklein,

        Was the scope ground connected to coil ground?(black wire)
        If so then you have a problem with the flyback voltage regulation cct (snubber). Start by checking the voltage on the large electrolytic cap, it should be 180~190. You will have to trace the cct around there to figure what is meant to be going on. The voltage on this cap is recycled to keep it at its predetermined level. Don't hook up your new fets yet or they will fry.

        Mario, When they went to dvt, they were able to shorten the smaller pulses but still get the same bang. This allowed the cycle time to be reduced equaling more samples / second improving signal to noise ratio.

        Cheers Mick

        Comment


        • #34
          Originally posted by Mechanic View Post
          Hi Bklein,

          Was the scope ground connected to coil ground?(black wire)
          If so then you have a problem with the flyback voltage regulation cct (snubber). Start by checking the voltage on the large electrolytic cap, it should be 180~190. You will have to trace the cct around there to figure what is meant to be going on. The voltage on this cap is recycled to keep it at its predetermined level. Don't hook up your new fets yet or they will fry.

          Mario, When they went to dvt, they were able to shorten the smaller pulses but still get the same bang. This allowed the cycle time to be reduced equaling more samples / second improving signal to noise ratio.

          Cheers Mick
          They seemed to have also increased sensitivity by using two 797s. Fortunately NE5534s are pin compatible with 797s, so I can replace that top NE5534 and/or components without much hassle.

          It's too bad the Minelab people switched to using noisy 4066s from the much more quiet 4053s. They actually went one step backward since the SD2000s already had the 4053s!

          Comment


          • #35
            My scope ground was the blue wire that contacts the panel EMI shield. So are you saying if the snubber was working the pulses would not go to 208V? What should they peak out at? Would they be essentially clipped to the level on the big cap, or lower?

            What should the signals on the right fet look like? If the snubber is bad it would be the cause of the left fet shorting. The right one survived. So I'm wondering if my first effort should be to check out the snubber before scratching paint off all the other RX parts and testing them. I don't understand the functional purpose of the two fets - if they are parallel circuits or each has its own purpose.

            I wasn't aware there are noise differences between the 4066 and 4053. Is this noise switching noise transients? Would it make a difference in the PI clone projects?

            Comment


            • #36
              The 4066s are indeed more noisy -I've considered replacing them at a later time.

              Is there any way you can zoom in to those 3 short pulses? how long are they?

              Comment


              • #37
                Hi Mario,

                Which 2000 schematic are you lookin at? Mine says 4066!

                Bklein,

                With a DD coil on it is likely that only 1 fet has blown as the DD receive fet would not be seeing the voltages that the mono(tx coil) is seeing. These are 2 parallel ccts, one for receiving the signal from the mono coil the other for the dd coil.

                The voltage on the cap might be just a tiny bit lower than the flyback voltage observed at the coil(diode body voltage drop) But it should be less than 200v. The flyback voltage should reach the voltage of the cap and then flat top for 3~4us on the long pulse and 1.5~2 us on the short pulses.

                The 4053 has less oscillations when it is turned on and I suspect that the capacitance(input to output) would be lower, blocking higher frequencies from making it through when the switch is off.....

                Cheers Mick

                Comment


                • #38
                  I checked all the diodes and fets etc. with a Fluke 87VI ohms and diode check and didn't find any shorts or something to wonder about. I then came back with the scope and saw symmetrical signals in the two RX channels. I realized that the 208V excursions were due to my lengthy scope ground. You can kinda see it in the previous photo but here's one that proves clamping is active.

                  Also I have attached a photo that shows damping fet action, and finally a better one of the fet gate timing.

                  I'm thinking maybe its ok to install the new parts??
                  Attached Files

                  Comment


                  • #39
                    Hi Bklein,

                    Time to put the new fets in I think.... Good luck!

                    Cheers Mick

                    Comment


                    • #40
                      Originally posted by Mechanic View Post
                      Hi Mario,

                      Which 2000 schematic are you lookin at? Mine says 4066!

                      Bklein,

                      With a DD coil on it is likely that only 1 fet has blown as the DD receive fet would not be seeing the voltages that the mono(tx coil) is seeing. These are 2 parallel ccts, one for receiving the signal from the mono coil the other for the dd coil.

                      The voltage on the cap might be just a tiny bit lower than the flyback voltage observed at the coil(diode body voltage drop) But it should be less than 200v. The flyback voltage should reach the voltage of the cap and then flat top for 3~4us on the long pulse and 1.5~2 us on the short pulses.

                      The 4053 has less oscillations when it is turned on and I suspect that the capacitance(input to output) would be lower, blocking higher frequencies from making it through when the switch is off.....

                      Cheers Mick
                      My bad. It looks like they use the 4053s later in the audio cct.


                      I checked all the diodes and fets etc. with a Fluke 87VI ohms and diode check and didn't find any shorts or something to wonder about. I then came back with the scope and saw symmetrical signals in the two RX channels. I realized that the 208V excursions were due to my lengthy scope ground. You can kinda see it in the previous photo but here's one that proves clamping is active.

                      Also I have attached a photo that shows damping fet action, and finally a better one of the fet gate timing.

                      I'm thinking maybe its ok to install the new parts??
                      Looks like you'll be digging up holes soon enough. There's lots of gold up them California mountains.

                      Let us know once you get that unit working. Also, one last photo of the board lay out would be nice.

                      If you can't find a fet with low enough capacitance numbers, try using 2SK216 -the same one used in the 2100s. It's more than good enough for the job.

                      Comment


                      • #41
                        I bought BSS87's from Digikey. Not 90pf but 120pf - will they be ok?
                        The 2sk216 seems harder to get and different package.

                        Comment


                        • #42
                          HI Bklein,

                          The bss87's will be fine

                          Cheers Mick

                          Comment


                          • #43
                            Originally posted by Mechanic View Post
                            Hi Bklein,

                            Time to put the new fets in I think.... Good luck!

                            Cheers Mick
                            AWWWW MANNNNNN.....
                            It worked out of the case, I put it back inside, cleaned everything up, and it died again.
                            Same problem - BC857C and left fet bad.

                            I tested the battery system by itself and they do have a soft-start in the design and the output voltages are accurate; without overshoot.


                            Barry

                            Comment


                            • #44
                              There is a definite intermittant at the coil connection to the unit on the yellow wire (shield) pin 1.
                              As far as I can tell it is due to the connectors themselves - I read on Woody's site that he saw this issue too - said the plating on the contact was worn down to brass. Mine probably are like that.

                              What are recommended part numbers/sources for the plugs?

                              Comment


                              • #45
                                Hi Bklein,

                                Bugger!

                                For some reason on the gp's they stopped bridging pin 1 and 5 inside the detector. On the 2000 pin 1 and 5 are bridged and pin 3 and 4 are bridged also. Check on the coil plug and see if these pins are bridged on that. If so I would be inclined to bridge these pins on the detector side too. It is added protection and less prone to burnouts due to bad pin connections. It should also be quieter if the coil plug is bumped while detecting. My only thought as to why this was done was to stop circulating ground current in the coil plug. All good and well if decent bulky connectors are used. Perhaps try cleaning the plugs first, but I would be inclined to bridge the pins anyway.

                                Also use a cotton tip with some contact cleaner and clean the terminals in the plug on the detector and use a toothpick with contact cleaner to clean the on the coil too. Clean until the toothpick and cotton tip come out clean. If your really trick you can wind some cotton onto the toothpick to aid in the cleaning.

                                Cheers Mick

                                Ps Just checked on a Minelab 11"DD that I purchased last year and the pins are bridged on the coil plug so bridge the pins on the detector side too.
                                Pss Also make sure you spray the board with a protective coating to keep the moisture out, eg plasticoat

                                Comment

                                Working...
                                X