Announcement

Collapse
No announcement yet.

vanquish Tiny PCB

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

  • Skippy
    replied
    A few thoughts and observations:
    1 ) Mr Foma's 'schematic' for the Eqx is VERY rough and approximate, and contains many errors. It serves as a guide, upon which can be built on, by doing a proper reverse-engineering job.
    2 ) Notice the PCB has many unfilled transistor pads, SOT-23. Pairs of these are paralled up - this is clearly a low-noise discrete front-end, like a long-tail pair, surrounded by a low-noise opamp to complete the feedback loop. As they are missing, ML are clearly satisfied that just the opamp alone is adequate for noise.
    3 ) The PCB silk-screen shows the RX output is a differential pair, like on the Eqx. So is safe to assume the second opamp in the IC package is used as an inverter to create the complementary signal.
    4 ) Regarding 'redesigning it with through-hole part' . NO. It's not the 1980's, do it like Minelab have done it. Short tracks, small components all make for better performance, eg. less noise, less EMI pickup, more effective power-supply decoupling etc. The parts may not even be available in leaded packages. And obviously, a small magnetic footprint make the circuit less visible to the coil, so less disturbance of the induction-balance.
    5 ) Notice the small inductor on the Eqx PCB is missing on this board. I thought it was in series with the RX coil input, possibly to improve high-frequency performance, possibly just filtering.
    6 ) The tracks etc need some close attention. Clearly the ground-plane has that thin finger track pattern to give electrical shielding, without a huge magnetic footprint that will be seen by the coil. The X-ray does show a shadow of a continuous square shape, looking like a solid ground-plane. This seems unlikely, but it may be there.
    7 ) The PCB can only be reverse-engineered when it's in the hands of someone skilled. Doing it from crap photos is just stupid. GOOD photos of the PCB will be needed if anyone on the Forum is going to help you.

    Leave a comment:


  • sinclairuser
    replied
    i think you are missing a connection,https://www.geotech1.com/forums/atta...7&d=1664707625
    clearly shows a 3rd connection on the centre pad(shield or buck), but this is missing from https://www.geotech1.com/forums/atta...1&d=1664707887
    since no one noticed i thought i would.

    Leave a comment:


  • waltr
    replied
    Originally posted by Altra View Post
    I actually captured the data with my Rigol logic analyser. It would be a lot of work to decode and and duplicate the hand shake, specially when we don't know the exact device. Most of the new encryption schemes use random keys and are nearly impossible to crack.
    Didn't think about it being an encryption with a key input and response out. This would be very hard to crack.

    Leave a comment:


  • Altra
    replied
    Originally posted by waltr View Post
    It may be possible to capture the transfers on the DATA pin during detector power up. But this requires a working unit.
    I actually captured the data with my Rigol logic analyser. It would be a lot of work to decode and and duplicate the hand shake, specially when we don't know the exact device. Most of the new encryption schemes use random keys and are nearly impossible to crack.

    Leave a comment:


  • waltr
    replied
    Originally posted by Altra View Post
    It's called "one wire serial communications"

    Maybe this is the 36c4p?

    https://www.microchip.com/en-us/product/ATsha204a

    Yes, very likely this is the 3-pin chip used or another that is similar.

    Notes from the data sheet:
    Chip markings are only manufacturing codes and DO NOT identify the chip.
    The DATA pin is Bi-directional, Wakeup & commands from Host, data to host.

    It may be possible to capture the transfers on the DATA pin during detector power up. But this requires a working unit.

    Leave a comment:


  • h9361
    replied
    Thank you.
    Do you know which other models of ML is not protected?

    Leave a comment:


  • Altra
    replied
    Etrac is not protected

    Leave a comment:


  • h9361
    replied
    Thank you dear Altra.
    Do you think E-Track also works like this?
    Because the Detech company are generating Coil for E-Track.

    Leave a comment:


  • Altra
    replied

    Leave a comment:


  • h9361
    replied
    I tried to find its datasheet, but i couldn't.

    Leave a comment:


  • Altra
    replied
    Originally posted by h9361 View Post
    OK, thank you.
    But 36C4P is a mystery to me.
    one pin is connected to + 3.3V, another to ground, the third (data) can be an input, but there is no output, or it can be an output but there is no input!
    It's called "one wire serial communications"

    Maybe this is the 36c4p?

    Leave a comment:


  • h9361
    replied
    OK, thank you.
    But 36C4P is a mystery to me.
    one pin is connected to + 3.3V, another to ground, the third (data) can be an input, but there is no output, or it can be an output but there is no input!

    Leave a comment:


  • kt315
    replied
    yup, Altra, no chance. i meant that above.

    Leave a comment:


  • Altra
    replied
    Originally posted by h9361 View Post
    Dear kt315
    Do you think 36c4p is diode ?
    [ATTACH]57963[/ATTACH]
    The 36c4p is a encrypted anti-cloning device. Minelab use these in all their new detectors. Without this device the detector will shut off.

    The only way to make coils for the Equinox, Vanquish, GPZ7000 and GPX6000 is to have a licensing agreement with Minelab.

    Leave a comment:


  • kt315
    replied
    pin DATA is digital signal, in my opinion. some kinda one line of the interface. while i really do not know
    what Foma wanted to say.

    Leave a comment:

Working...
X