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.

BQ79600EVM: Different Versions of TMS570LS12 ?? - "SPI-example" does not run on the "new" boards.

Part Number: BQ79600EVM
Other Parts Discussed in Thread: LAUNCHXL2-TMS57012, CODECOMPOSER, TMS570LS1224, USB2ANY, CCSTUDIO

Hi,

hopefully this is just a setting I need to change ...

The "SPI-example"-code was changed for my needs and works fine on the TMS that was ordered back in July 2020 (LAUNCHXL2-TMS57012).
This TMS came with a "black and red, nicely labeled" Box and had short 40pin-connectors soldered to them.

Two additional boards were bought several months later and they don't execute the code :-/
The additional boards were shipped in blank white boxes, without labels and have extremely long 40pin-connectors
(at least double the length of the "normal" pins)

Both new boards required a Firmware-Update when first tried to upload code
and both got different Firmware Versions?? (3.00.00.14 and 3.00.00.15)

I can upload the code but I cannot debug it like with the "old" one.
I don't even reach the point after initialization with the first printf.
On one board I can "Pause-Execuion" and he is at some interrupt failure.

Is this a target-issue?
I could change to different target-boards in CodeComposer, but all fail to run/debug the code (but test-connection is all right)
Putting the previous board, I can change, load and debug the code without any issues.

How can I find out if this is a TMS570LS12-24; -25; -27 .. ? (If this is even the issue)
In any case the "new" LAUNCHXL2-TMS57012xx-boards are different (HW and SW) and don't work with the sample code :-/

Best Regards
Collin

  • Hi Collin,

    Yes the mcu code was wrote for TMS570LS1224 in the LAUNCHXL2-TMS5012. I have forwarded this thread to TMS570 team rather than bq79600 team (which I support) to receive feedback from the TMS570 team regarding which one you may have. This seems to be more of an issue regarding the MCU/CCS.

    In terms of other things to check ensure:

    • If you are plugging in the MCU directly to the 600 board, you do need to switch 2 resistors (remove R7 and R9 and populate R8 and R10) and follow user guide suggestions for SPI. The board is configured to use UART when using the MCU. Notice these resistors are only when using the headers to plug in the MCU, these resistors don’t have any effect when using the USB2ANY.
    • I’ve had issues to communicate when using longer wires to connect the MCU to the 600. You get a much cleaner connection but plugging in the MCU directly to the 600 EVM.

    Regards,

    Taylor

  • Hi Taylor,

    thanks for the suggestions. Sure all three boards are modified to use SPI, and all of them work.
    I can attach all 600 boards to the "old" board and each time I modified something small to see if uploading works.
    And everything works fine each time.

    I'm using the same wires, same length, same setup, even the same spot on the desk, when I use the "new" boards.

    I cannot even reach line 116 of sys_mai.c:   printf("\n\n\nBeginning Program\n"); nor can I suspend the execution to check the code.

    the same code on the "old" TMS looks like this:

    Hope this help to solve the issues.
    Best Regards

  • Hello,

    The R7, R8, R9 and R10 are populated. R10 is a external pull-up resistor to nRST signal. R8 and dR9 are used to calculate the threshold which triggers the nPORRST. 

    I'd like to know the reason for removing R7 and R9.

  • Hi QJ,

    The R7, R9 vs R8, R10 topic is related to the BQ79600 boards.
    R7 and R9 connect to M_Rx and M_TX and R8 and R10 connect to M_SOMI and M_SIMO.

    but this is not the issue right now ;)

    I have the same behavior if I try to run the UART-Example.
    I cannot run or debug on the "new" boards.

    They are physically different and I don't know what to target or what to change, because code-composer doesn't give me an error.
    All connection-test are OK and successful.

    I was able to flash the program, erasing the demo that was on there by default. So I know that everything went well.
    Also, I updated the firmware (now all are on 3.00.00.15) then I encountered, that my "old" board always received COM8 and COM9
    while the new boards connect to COM10 and COM11.

    "msinfo32" gave these differences (old vs new) see below pic.

    PNP Device ID USB\VID_0451&PID_BEF3&MI_00\D&106F64B8&0&0000
    PNP Device ID USB\VID_0451&PID_BEF3&MI_00\D&12A151C8&0&0000

    All "new" ones have the same IDs and don't run the code at all, not even the 'printf' in the beginning 
    (but code composer says running, see last post)

    I also made pictures of the labels:
    2019 Version works perfectly ... 2020 Versions not at all.

    the chips on the Launchpad also have different Labels:

    small chip 2019-Version: YF TM4C129 ENCPDTI3 78AYPTW G4 <- working
    small chip 2020-Version: YF TM4C129 ENCPDTI3 8CCELZW G4 <- not working 

    large chip 2019-Version: TMS570LS 1224CPGEQQ1 YFC-66ADNKW G4 <- working
    large chip 2020-Version: M46 L852CPGET YFC-84CC5NW G4 <- not working


    Would be great if you could tell me what checkbox I need to change in code-composer to get both up and running.
    Best regards, Collin

  • Hello Collin,

    Thanks for your clarification. 

    Can you please try a bigger stack size and heap size?

  • Hi QJ,

    unfortunately, there are a lot of broken links on the page you recommended.

    ---
    This Page Does Not Exist
    The Processors Wiki was removed on January 15, 2021
    ---

    could you be more specific for this ti-example-code?

    or maybe @Taylor, @Leslie, @BQ796xx-Team someone could provide an updated sample?
    maybe one that also targets a higher version of the ARM compiler (> 18.1.2)

    Thanks in advance, I will continue to try to get this running.
    Best Regards 

  • Hello Collin,

    How about the two new launchpads? Have you make them work?

  • QJ Wang said:

    How about the two new launchpads? Have you make them work?



    Hello QJ,
    unfortunately there was no progress on the RM46 boards ...
    And also no new Error-Messages or any more information from the CCStudio.
    It just continues to flash and run (but it never starts running).

    Any idea where the data from the Wiki went?
    If you have more suggestions, I'm happy to try them.

    I don't understand why they are shipped under the same name if they are so different.
    I cannot even buy new ones, because no-one can guarantee that what I order are "real" TMS570LS1224s :-/
  • Hi Collin,

    I am sorry for the problem which may be caused by the launchpad.

    large chip 2020-Version: M46 L852CPGET YFC-84CC5NW G4 <- not working

    I just noticed that the new board you are using is RM46 rather than TMS570LS1224. RM46 is a little endian device, but TMS570LS1224 is a big endian device. To use TMS570LS12x SW on RM46 device, you need to do some minor code change.

  • Hi Collin,

    Can you share the "SPI-example" for BQ79600, so I can help you modify the code for RM46 device? thanks

  • sure, I've already sent the file via private message. would be great if you could help.

  • Hi Collin,

    I am sorry. I am not able to download the file. Please send to my email. Pleas find my email address in the private message. Thanks

  • I checked your project. DMA is not used in SPI transfer. The code can be used for RM46 without modification except for the project property:

    1. Device variant

    2. Device endianness

    3. Runtime support library