• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 4 answers
  • Subscribers 31 subscribers
  • Views 797 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

DLPDLCR160CPEVM: recovery after bad firmware flash

E WilkenResman
Prodigy 70 points
Part Number: DLPDLCR160CPEVM
Other Parts Discussed in Thread: UNIFLASH, DLPC3421, LP-XDS110

Tool/software:

Hello,

I obtained demo code for the DLPDLCR160CPEVM board through the TI secure site. This included source code for the out-of-box firmware, revision E15. Also included was a precompiled .out file in the "Debug" subfolder.

I tried to flash the EVM with this Rev E15 .out file. I used a J-link instead of the XDS110 (J-link is also supported by Uniflash). Through Uniflash, the J-link was able to connect and started the upload. Unfortunately, the flashing process failed almost immediately and the board is no longer responsive.

The green LEDs D1 and D2 are continuously lit after the board is powered via the USB-C connector. There is no reaction to activating the switches, and the projector does not turn on. Also, the J-link is now unable to connect to the board as target in order to erase or reprogram.

Is there any reason why the firmware flash would have failed, and is there anything that can be done to recover the board?

  • 0
    TI__Expert 7035 points

    Hello User,

    The MSPM0 handles the PROJ_ON signal to the DLPC3421 controller. If there is an issue with the MSPM0 coming up the unit would be unresponsive. It seems that the failure to flash the MCU is why this is occurring.

    Please allow us some time to look into this behavior.

    Best,

    John

  • 0 in reply to John McKnight
    TI__Expert 7035 points

    Hello User,

    We have used the standalone Uniflash program with no issues with the XDS110 or LP-XDS110 debug probes.

    For this specific issue, I recommend you create a ticket with the Code Composer Studio Team / Uniflash team. They would be better resources to talk with about the compatibility of a 3rd party debugger with their SW. You should get a faster response on your specific case.

    Best Regards,

    John

  • 0 in reply to John McKnight
    Prodigy 70 points

    Thanks -- I can open a ticket with the Uniflash team, relating to using the J-link with Uniflash. I'm also in the process of getting an LP-XDS110.

    However, the second part of my question relates to the recovery procedure for this specific board, which appears to be bricked. Can you describe how this EVM can be returned to a connectable, programmable state?

  • 0 in reply to E WilkenResman
    TI__Expert 7035 points

    Hello User,

    Please use this message as approval for a RMA request for this EVM.

    We have not ran into this issue. I plan to get this EVM and see if the team can recover the EVM.

  • 0 in reply to John McKnight
    Prodigy 70 points

    Thank you for the RMA authorization and for being willing to take a look at the unit. I received the LP-XDS110 today and would like to try it out with the DLP EVM before sending the EVM back.

    The photo in Figure 3-1 of dlpu131 shows a connection between the LP-XDS and the EVM, with the debug cable extending out from the EVM. Can you confirm that this wiring is right? I only get "low power" i.e. VTref not detected, error messages when the connection is made this way. However, I have measured that the EVM is consistently supplying 3.3V on pin 1.

    If the debug cable crosses over the board, contrary to Figure 3-1, I can read VTref (but there's a different error condition...)

  • 0 in reply to E WilkenResman
    TI__Genius 9640 points

    Hello User,

    Thank you for the information. John will be out this week so please expect a response back early next week. 

    Regards,

    Alex Chan

  • 0 in reply to Alexander Chan
    TI__Expert 7035 points

    Hello  User,

    In Figure 3-1 there is a red colored wire in the ribbon cable where all others are gray. That is on the left side of the EVM header. That red stripe should be on the right  side of the 110 probe header when looking at the figure. Please also insure that that the TGT VDD is set properly. The EVM should be connected to power and generating its own voltage to the MSPM0.

    Best,

    John

  • 0 in reply to John McKnight
    Prodigy 70 points

    Hi John,

    The EVM provides 3.3V on pin 1 of the EVM header J5, as shown on the silkscreen. This is on the right side of the EVM header. I believe the red wire is Pin 1/VTref on the Cortex 10-pin debug cable, and therefore should be on the *right* side of the EVM header in Figure 3-1 in dlpu131 -- i.e., the photo is incorrect.

    I received a second EVM yesterday and had a chance to test this out. When I connect the way shown in Figure 3-1, Uniflash logs a connection failure due to low power. If I reverse the orientation -- red wire on the right side of both the EVM and the LP-XDS110 headers -- I can connect and read memory. I'd appreciate if you can confirm my understanding is correct before I reflash.

    Thank you.

  • 0 in reply to E WilkenResman
    TI__Genius 11565 points

    Hello User,

    Thank you for the information! 

    John will confirm your message and get back to you.

    Regards,

    Lori 

  • 0 in reply to E WilkenResman
    TI__Expert 7035 points

    Hello User,

    I was successful with the setup below. I used it with both UniFlash and Code Composer. We will look into the image on the guide.

    Best,

    John

  • 0 in reply to John McKnight
    Prodigy 70 points

    Thank you for confirming this is the right way to connect the debugger to the EVM.

    My new (working) board connects and can read the MSPM0 memory when wired up as you've just pictured.

    The old (bricked) board prints the following error in Uniflash: "Error connecting to the target: (Error -614 @ 0x0) The target indicates there is an error condition from a previous SWD request. Clear the error the condition, and try the SWD address again."

    Since I'm not aware of a way to clear or reset from this error condition, I will be sending the old board back for RMA as mentioned upthread. Thanks again.