[ABANDONED] J-Link v11 not working, but v8 is

This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

  • [ABANDONED] J-Link v11 not working, but v8 is

    Hello,
    i upgraded form the J-Link EDU v8 to the v11 a few months ago. Debugging my prior project (STM32L4P5CE) worked (and still works) just fine with the v11.
    Trying the same with a STM32F301K6 (with or without opt. bytes) however does not work. I tried using VSCode+Cortex-Debug and also STM-CubeIDE. Both show the same behavior: The v8 version works, the v11 does not.
    When using J-Link-Commander directly i can connect to the chip with the v11 and it identifies the Cortex-M4 but it is somewhat inconsistent. Sometimes it has problems connecting in general, other times it is not able to see all the entries in the "CoreSight components"-category, the v11 also sometimes can not connect to the DAP initially while the v8 can.

    I tried:
    • uninstalling and reinstalling the entire Software-Package (now v7.88b)
    • updating the J-Link firmware
    • reassining the Segger drivers
    • Switching the entire PC-sided-system from Win11 to Ubuntu, also changing the hardware
    • checked every J-Link configuration that i could find in the Wiki and the open Internet


    At this point i am more then confused:
    -The v11 Probe is not just entirely broken as it works just fine with my other MCUs.
    -The Target is also not just defective, as it works perfectly fine with my v8-Probe.

    I have attatched screenshots of the Commander-Feedback using the v8 and two different results using the v11.
    I really would appreciate some help or guidance here.

    Best regards,
    Lukas
    Images

    • v8_justfine.png

      57.81 kB, 507×930, viewed 213 times

    • v11_alsoBroken.png

      47.36 kB, 595×603, viewed 196 times

    • v11_conUnderReset_noCoreSight.png

      56.77 kB, 503×913, viewed 243 times

  • If you reduce the SWD speed to let’s say 2 MHz or even 1 MHz, does it become stable?

    I do not know the exact parameters out of my head but it is very possible that the V8 does not do exactly 4 MHz but more something in the league of 3.2 - 3.9 MHz while the V11 should be able to do the 4 MHz more exactly.

    Maybe the 4 MHz are too much for the HW design you are using and the V8 is just lucky to be somewhat slower than configured?

    Just a guess at this point…
    Please read the forum rules before posting.

    Keep in mind, this is *not* a support forum.
    Our engineers will try to answer your questions between their projects if possible but this can be delayed by longer periods of time.
    Should you be entitled to support you can contact us via our support system: segger.com/ticket/
    Or you can contact us via e-mail.

  • Hello again,
    so as a quick followup:

    Does the speed have an effect?
    Yes, definetely. I have to throttle the v11 down to < 800 kHz and the Commander can connect, halt, go and read mem.
    It however can neither flash the device nor debug it in VSCode&Cortex-Debug or STCubeIDE. The v8 still does that happily.

    I have also given a identical target to a friend who owns another v11. I have made sure that it is one that i tried before. It just works for him as it should.

    I have also done some more "extensive" connection testing:

    works just fine:
    -STM32L4P5CE
    -RP2040_0 (not on v8, no multidrop debugging)
    -TM4C123GH6PM
    -STM32F091RC
    -NRF52832 AA
    (i still have some more, but it seemd to me like the general setup is not at fault)

    only works on v8:
    -STM32F301K6

    The tests above were done with the same cables, so that is not an option.

    I also noticed that it seems to be getting worse. While i used to get inconsistent results on 4MHz, now i do not get any traces of a connection.

    At this point i had insufficient grounding in mind. I have reviewed the target hardware layout (done my my UNI-Project-Partner). I noticed that the debug-connector (standard 20Pin) is connected to GND by one singular teeny tiny trace. That theory is however crashed by the fact that it works fine for my friend. I also have handsoldered some more GND-Connections to the Connector (not pretty, but i am getting desperate), no change.
    Is there a possibility that this is still the problem?

    Is there any other ideas? Any help/suggestions is appreciated.

    Best regards,
    Lukas