• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Subscribers 33 subscribers
  • Views 403 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.

AM2634: Emulation suspend and TSCTR counter stop

Jeeuk Chang
Prodigy 240 points
Part Number: AM2634
Other Parts Discussed in Thread: SYSCONFIG, , TMDSHSECDOCK

Tool/software:

Hello, TI experts.

I set ECAP0 as APWM mode running at 20 kHz with 50% duty in CPU0.

When I suspended CPU0 via JTAG, the XBAROUT4 still made a switching signal output.

Like above pictures, I put the emulation suspend mode.

What is the problem?

  • 0
    TI__Guru*** 135355 points

    Hello Jeeuk,

    Apologies but I just got back from being out of office today and didn't quite get the time to load an equivalent example on my hardware to test the same. I will look to do so tomorrow and investigate this further. From a SysConfig standpoint the code does seem to be generating correctly based on these settings, so I'll have to look into the register settings and lower level drivers next.

    Best Regards,

    Ralph Jacobi

  • 0
    TI__Guru*** 135355 points

    Hi Jeeuk,

    Deep apologies here, I dropped the ball on this request entirely. I noticed that I hadn't gotten back to you today as part of my checks before leaving for a couple weeks, so I have talked with another engineer who will take over this request as far replicating on hardware goes.

    I did look today into the drivers / register settings and they appear to be configured right, so if we are able to replicate the same behavior we'll need to involve deeper experts on SW side as to what could be going awry here.

    Best Regards,

    Ralph Jacobi

  • 0
    Prodigy 240 points

    Hi, Ralph.

    This problem is important only when using the development method with JTAG emulation.

    I bypassed this problem with software works so I am not urgent to solve this problem.

    ------

    In my project, I use two APWM outputs for signal sensing.

    One is 20kHz output and the other one is 20MHz output for SDFM clock.

    (I had no ePWM output pin left so I had to use APWM outputs.)

    When I processed SDFM inputs, I found SDFM FIFO was filled completely whenever JTAG stopped.

    This phenomenon was due to non-stop APWM outputs.

    I checked syscfg setting and eCAP register values but I could not find my mistakes.

    Now I am using direct register modification for APWM setting instead of using syscfg program.

    But I still have the problem.

  • 0 in reply to Jeeuk Chang
    TI__Genius 10463 points

    Hi Jeeuk, 

    I can replicate your setup tomorrow and see if the issue persists on my end.

    Regards,

    Brennan

  • 0 in reply to Brennan Hartigan
    TI__Genius 10463 points

    Jeeuk,

    Are you using a custom application or an example from the AM263x SDK?

    Regards,

    Brennan

  • 0 in reply to Brennan Hartigan
    Prodigy 240 points

    Hi. Brennan.

    We made our application board with AM2634. I am testing with it from this year.

    Regards,

    Jeeuk

  • 0 in reply to Jeeuk Chang
    TI__Genius 10463 points

    Jeeuk,

    Understood, I was asking more in terms of the software you are testing and seeing this issue with.

    Regards,

    Brennan

  • 0 in reply to Brennan Hartigan
    Prodigy 240 points

    Brennan,

    I tested my application in the following environment.

    - CCS : 12.8.1

    - Compiler : TI Clang v3.2.2.LTS

    - SysConfig 1.21.0

    - MCU+ SDK for AM263x 9.2.0.55

    - Configuration : Release

    Regards,

    Jeeuk

  • 0 in reply to Jeeuk Chang
    Prodigy 240 points

    Brennan,

    I changed my environment today.

    - CCS : 12.8.1

    - Compiler : TI Clang v4.0.1.LTS

    - SysConfig 1.22.0

    - MCU+ SDK for AM263x 10.0.0.35

    - Configuration : Release

    I tested that setting again with AM263x SDK and got the same result.

    I attached modified sysconfig file used with "empty_am263x-cc_r5fss0-0_nortos_ti-arm-clang" project

    1541.syscfg.zip

    You can see a square signal on AM2634 A10 pin or TMDSHSECDOCK HSEC_SPI1_CLK 79 pin.

    The signal do not stop regardless of JTAG suspension.

    Regards,

    Jeeuk

  • +1 in reply to Jeeuk Chang
    TI__Intellectual 2026 points

    Hi Jeeuk,

    Thanks for holding on, this is possible due to the fact the application may not have enabled Halt Controls for the ECAP. Could you please take a look at the register in Controlss_ctrl space for something like ECAPx_HALTEN ? is it enabled for the CPU that you are currently using for debugging?

    If not, try writing to this register to enable Halt, please note that you would need to unlock this space before writing to the register to make the writes happen.

    The current version of SDK is not supporting this configuration in auto generated files from the sysconfig. this will be addressed in the future releases. 

    Let me know if this helps!

    Thanks and regards,

    Madhava 

    [p.s. this is similar to the epwm halt en configurations if you are using any]

  • 0 in reply to Mihira Madhava Bollapragada
    Prodigy 240 points

    Hi Madhava,

    I checked CONTROLSS_CTRL_ECAPx_HALTEN registers and understood more configuration is needed.

    SOC_controlModuleUnlockMMR(SOC_DOMAIN_ID_MAIN, CONTROLSS_CTRL_PARTITION0);
    *(volatile uint32_t *)(0x502F05D0U) = 0x0001U; // CONTROLSS_CTRL_ECAP0_HALTEN
    SOC_controlModuleLockMMR(SOC_DOMAIN_ID_MAIN, CONTROLSS_CTRL_PARTITION0);

    I got the normal result with the above command added.

    Best Regards,

    Jeeuk

  • 0 in reply to Jeeuk Chang
    TI__Intellectual 2026 points

    Happy to see this got resolved! 

    thanks and regards,

    Madhava