• State Resolved
  • Locked Locked
  • Replies 15 replies
  • Subscribers 82 subscribers
  • Views 3533 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.

Missing PeripheralHeaderIncludes.h for TMDS2MTRPFCKIT

VG
Intellectual 620 points

Other Parts Discussed in Thread: CONTROLSUITE

CCS: v4.2.1

Using ControlSUITE example for kit referenced above.

Got a build error, "could not open source file "PeripheralHeaderIncludes.h"

Where does TI recommend I get this file from?  And the peripheral headers that it references?

  • 0
    Intellectual 620 points

    In project properties, a link existed for the header file via a macro named CONTROLSUITE_LVMOTOR_ROOT.  I could not find a definition for this macro.

    I noticed that PeripheralHeaderIncludes.h is in the ControlSUITE directory.  I created a user macro that points to this directory.

    So now I have a different set of include path directory errors.

    For a kit that is supposed to work out of the box, I am encountering some interesting roadblocks.

  • 0 in reply to VG
    TI__Guru 57255 points

    Hi Vincent,

    First of all, I'm sorry for the issues you are having.  I went ahead and imported this project into my CCSv6.1 installation and the project was able to compile with no issues (in the FLASH build configuration).

    A few questions:
    1) When you imported the project into CCSv4, did you select the option to "Copy project into workspace"?  If you did, you shouldn't.  The copy option directly effects the ability for the path variables to be generated correctly. 
    2) Is it possible for you to try with CCSv6?  In addition to having a sign from me that project was able to compile with CCSv6, it is also a newer version of the software and will therefore be better supported.

    Please let me know how things go.

    (I notice that in your screenshot, DSP2803x_CodeStartBranch.asm [and others] have a yellow exclamation point next to them and are showing that CCS cannot find those files - this means there are more pathing issues which *could* be resolved manually.  However I would recommend reimporting the project since it will likely be less painful)


    Thank you,
    Brett

  • 0 in reply to Brett Larimore
    Intellectual 620 points

    Brett,

    Thank you for your response.

    I imported the project into CCS 6.1.  Then I got a warning about the project being created in a different compilier version.

    I went into project properties and changed the compilier version to 6.4.10.  After a rebuild I got the following warning:

    What does this message mean?  Do I need to rebuild the libraries?  If so, how do I rebuild them?

  • 0 in reply to VG
    TI__Guru 57255 points

    Hi Vincent,

    No, TI owns the source for these libraries.  They were compiled with an older version of a compatible compiler. 

    These libraries work and these warnings can be ignored.
    http://processors.wiki.ti.com/index.php/C28x_Compiler_Error_and_Warning_Messages#Warning:_build_attribute_vendor_section_TI_missing_in_.22.3Clibrary_or_object.3E_:_compatibility_cannot_be_determined


    Thank you,
    Brett

  • 0 in reply to Brett Larimore
    Intellectual 620 points
    Brett,

    Thank very much for your response.

    In regard to the compilier warning: Do you foresee any issues in using v6.4.10 when the project was originally setup in v5.2.3?

    Thanks again,

    -Vince
  • 0 in reply to VG
    TI__Guru 57255 points

    Hi Vince,
    I wouldn't expect any.  The compilers are tested well before releasing.  Most revisions are improvements on the release prior (very few impactful bugs).

    Thank you,
    Brett

  • 0 in reply to Brett Larimore
    Intellectual 620 points

    I tried to run the system and it looks like it is hitting a software breakpoint.  However, I don't see anything on the breakpoints list.  Any idea what is happening here?

  • 0 in reply to VG
    TI__Guru 57255 points

    Hi Vince,

    We are looking into this.  I just had a colleague compile, load and run the various incremental builds onto a controlCARD + Docking Station hardware combination.  He didn't run into any issues.

    I am going to have him try and run the project on the actual kit as well and see what occurs.

    When does the issue occur?  Immediately after hitting the 'run' button?


    Thank you,
    Brett

  • 0 in reply to Brett Larimore
    Intellectual 620 points
    Brett,

    Thank you very much for the update.

    Yes it pretty much does occur right after hitting the "run" button.

    I looked at the linker command file and it looks like 0x3ff59a is part of the bootblock partition. I am not really familiar with the f28035 device so I am not sure how the boot memory space is supposed to be setup.

    It seems like there is no bootloader code on the device. Although, since I am running with a debugger, I am not sure why I would need a bootloader.

    Thanks again and I look forward to your update.

    -Vince
  • 0 in reply to VG
    TI__Guru 57255 points

    Vince,
    After loading, can you try to hit the 'reset' button, next the 'restart' button, and then 'run'?

    Thank you,
    Brett

  • 0 in reply to Brett Larimore
    Intellectual 620 points
    It looks like that solved it. Thank you very much for your help.
  • 0 in reply to VG
    Intellectual 620 points
    One more question....

    Page 19 of the document "Sensorless Field Oriented Control of Multiple Permanent Magnet Motors" states that IsrTicker should increase when EnableFlag is set to 1. IsrTicker seems to increase regardless of what EnableFlag is set to. A global search "ctrl-H" doesn't show the EnableFlag variable being referenced.

    Is there a way to disable motor excitation while the CPU is running?
  • 0 in reply to VG
    TI__Guru 57255 points

    Vince,

    Line 195 in the main source file is where EnableFlag is.  It appears that this got commented out though.  Uncommenting it would allow you to have would you are desiring - however it's probably not as elegant as you'd like.  However, this should allow you to get an idea of how one might do this. 

    There are of course other ways of keeping the ePWMs off until you want them - forcing the ePWM output low in the ePWM module; keeping the ePWM output as a GPIO in the low state; etc.

    (as an FYI: this project/kit is one of the older motor control ones in our portfolio.  It still works, is valid, and the methodology used in this project is nearly identical with what is done in the more recent projects.  That being said, there may be some improvements that likely could be made.  Projects that I might recommend as auxiliary references to this one are the software projects under HVMotorCtrl+PfcKit_v2.1 or DRV830x-HC-C2-KIT)

    Hopefully this helps!


    Thank you,
    Brett

  • 0 in reply to Brett Larimore
    Intellectual 620 points

    "I tried to run the system and it looks like it is hitting a software breakpoint.  However, I don't see anything on the breakpoints list.  Any idea what is happening here?"

    "Vince,
    After loading, can you try to hit the 'reset' button, next the 'restart' button, and then 'run'?

    Thank you,
    Brett"

    After shelving the project for awhile, I tried to load the project into flash and ran into this issue again.  After I load code it shows "(Suspended - SW Breakpoint)" and  "(no symbols are defined for 0x3FF599)"

    CCS version is 6.1.1.00022

    Any ideas as to what is happening here?

  • 0 in reply to VG
    Intellectual 620 points

    Setting up a new target configuration seems to have resolved this issue.