Cart New Account Login

HomeAbout usProductsSupportForumsBlogCustomer Service

search inside this forum
USB UNIVERSAL DEAD, 3x times!
Denis A. Mar 30, 2016 at 09:40 AM (09:40 hours)
Staff: Gerardo R.

  • Hi,
    Our USB-ML-UNIVERSAL is dead again. From what I read on the forum, it happens to other people too. We had 3x programmers dead now in few years. The main MCU of the programmer is getting very hot. I would like to buy a better protected programmer (Cycone from PE???) or add extra protections inside this programmer.

    I'm suspecting two problems: I *think* there are no ESD protections inside the programmer itself. I read in another thread that there are ESD protections inside the Cyclone familly.

    So first:
    - I would like to add these ESD protections inside the Universal programmer.

    Second:
    - I'm looking for a complete optical isolation of the JTAG port provided by an extra PCB.

    Does anybody has any other clue how to protect this programmer? Any off the shelf optical isolator to propose?
    Thanks,
    Denis Alain




    Comments

  • I'd like to further understand the kind of target that you're connecting the Multilink that keeps killing it. There's nothing we can really do to add ESD isolation inside the Multilink programmer, but I can provide you with the debug header schematics so you can design your own or procure a third party solution.

    • Hi Gerardo,
      The programmer is connected to a Kinetis MK21FX512VLQ12 on the JTAG port + reset. The programmer do not power the board. There is a 4.7K pull-up on TMS connected to +3.3V and 1x 100pF on TCLK. TDO and TDI are directly connected. The reset pin has 100nF.

      I'm interested to any comments about this circuit or any additionnal protections we can add, since I would like to correct this issue in production.
      Thanks,
      Denis

      • There's nothing that seems out of the ordinary with that configuration. Is there anything worth noting about your VCC and ground pins? Are the pins on the debug header connected to power planes or by a trace? Any bypass capacitors nearby? If the target is self powered, we do siphon off some current from the VCC pin to drive the debug pins at the same voltage.

  • Hi Gerardo,
    Sorry for my late reply. Both VCC and ground pins are directly connected on the plane (no traces) and there are many caps around the connector, since the MCU is just few mm away.

    The last time the debugger blowed off, I suspect that it was when I connected the 24VDC on the board (we are generating all powers from 24VDC). This connector is very far from the MCU itself, so an ESD event is still possible, but...

    I will modify the test bench to switch the AC input of the 24VDC external power supply. So it will provide a smooth rising/falling on the 24VDC itself instead of connecting a live 24VDC on the DUT (which is not a normal use case anyway).

    Can I have your proposed debug header schematics you talked in your first reply?Thanks for your help,
    Denis

    • I have sent you a partial schematic of the input circuitry. I did notice that we placed 5.6V Zener diodes to protect against ESD on the RESET and VCC lines. That of course doesn't protect you from ESD events on any other debug line though. You may be willing to salvage the units by replacing the tri-state buffer.

Add comment


   Want to comment? Please login or create a new PEMicro account.







© 2017 P&E Microcomputer Systems Inc.
Website Terms of Use and Sales Agreement