• State Resolved
  • Locked Locked
  • Replies 14 replies
  • Subscribers 31 subscribers
  • Views 769 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.

AM2632-Q1: ECAP Signal monitoring Unit not working as expected

Jeffrey Hanzlik
Prodigy 90 points
Part Number: AM2632-Q1
Other Parts Discussed in Thread: AM2632, , AM2634, AM2634-Q1

Tool/software:

I am following the information from the TRU section 7.4.6.7 I am following the configuration requirements.  I have interrupts enabled for the eCAP device.  I have the signal unit configured to do High Pulse Width detection.  I am not getting interrupts for either min or max events.  

Is there a demo application that would configure the eCAP and signal monitoring unit.  For reference I am using the eCAP to verify PWM signals that are inputs to the AM2632.

  • 0
    TI__Guru*** 135355 points

    Hello Jeffrey,

    I am not aware of such an example - can you post what you've done configuration wise for review?

    Best Regards,

    Ralph Jacobi

  • 0 in reply to Ralph Jacobi
    Prodigy 90 points

    Jacob above is the SysCfg for the eCAP.  From Section 7.4.6.7.1 of the spruj7f the TRU for the AM263x it outlines the settings required for the MUNIT to operate.  I think I have the settings correct.  The code I have configures the interrupt.  I have set a breakpoint in the ISR and verified the results of the function ECAP_getInterruptSource.  I have not seen it signal the MUNIT events.  The min/max values I choose for the configuration are outside the source signal.  I have verified this by reading the capture registers using  ECAP_getEventTimeStamp for each cap event 1 -4.  

  • 0 in reply to Jeffrey Hanzlik
    Prodigy 90 points

    Jacob,

    Can you confirm the version of eCAP present in the AM263x shipped on the EVM TMDCNDC263.  The reason I am asking is that I have added reads of the eCAP register MUNIT_1_CTL after I called the ECAP_enableSignalMonitoringUnit().  The result is 0.  This would imply that the Signal Monitoring Unit is not supported.

    Also, could you confirm if the Signal Monitoring Unit is enabled in the AM2632-Q1.

  • 0 in reply to Jeffrey Hanzlik
    TI__Guru*** 135355 points

    Hello Jeffrey,

    Can you provide the full part number and lot code details for both the AM263 on your EVM as well your AM2632-Q1 unit? Either with images of the packages with markings clear (preferred) or by writing them out in the same format as our errata shows the part numbers: 

    Best Regards,

    Ralph Jacobi

  • 0 in reply to Ralph Jacobi
    Prodigy 90 points

    Ralph,

    The picture of AM2634 on the EVM.  I did decode the markings using the latest data sheet.  It does have 10 ECAP devices.  I will leave it to you to further interpret the markings.

    Some additional notes.  I did inspect the code in the file ti_drivers.open_close.c.  The function Drivers_ecapOpen does include the ECAP_ISR_SOURCE_MUNIT_1_ERROR_EVT1 or other MUNIT related flags.  I have added a call to the function ECAP_enableInterrupt() in my code to include those flags.  One other observation the code generated to configure the MUNIT is after the ECAP is configured to run.  As per the documentation in the spruj42d AM263x Register Addendum section 3.8.10.1 all configuration is supposed to be done before the ECAP is setup to run.  I had also corrected that in my code by stopping the ECAP adding the MUNIT config settings and then re-starting it.

  • 0 in reply to Jeffrey Hanzlik
    TI__Guru*** 135355 points

    Hello Jeffrey,

    Been working this with the team and its being investigated.

    Just to follow up, you mention AM2632-Q1 as well - are you only testing with the EVM or with your own hardware too?

    Best Regards,

    Ralph Jacobi

  • 0 in reply to Ralph Jacobi
    Prodigy 90 points

    Ralph,

    We are currently testing only on the EVM.  Our board will not be ready for ~8weeks.

    Jeff

  • 0 in reply to Jeffrey Hanzlik
    TI__Guru*** 135355 points

    Hello Jeff,

    Our engineering team has re-created this issue and are further investigating root cause. Anticipate having another update on progress mid-week next week.

    Best Regards,

    Ralph Jacobi

  • 0 in reply to Ralph Jacobi
    Prodigy 90 points

    Ralph,

    Just checking to see where we are on this issue.

    Jeff

  • 0 in reply to Jeffrey Hanzlik
    TI__Guru*** 135355 points

    Hi Jeffrey,

    I'll provide full details tomorrow but the very quick answer is we incorrectly published features ahead of the device that has them enabled, so there will be a future orderable part that will have the Signal Monitoring Unit functionality available, expected early next year.

    I'll provide more background / info tomorrow.

    Best Regards,

    Ralph Jacobi

  • +1 in reply to Jeffrey Hanzlik
    TI__Guru*** 135355 points

    Hi Jeffrey,

    Sorry for getting to this later today than I intended but wanted to close the loop here before the weekend officially starts.

    Okay so what has been identified is that our SDK & TRM have pre-emptively released features that are coming on a device variant (same GPN, different orderable) that is due to come out in 1Q25. The collateral was put together in preparation for that, but we didn't properly apply filters to keep it behind closed doors until the proper orderable was released.

    The list of features impacted by this are the following:

    • EPWM Multi Compare Shadow Registers (XCMP)
    • EPWM Pulse Edge Monitoring
    • ECAP Pulse Edge Monitoring
    • Diode Emulation Logic (DEL) Function + XBAR

    In 1Q next year, the orderable with these features included will be available and our device datasheet will be updated to reflect what the new orderable part provides from a feature standpoint. Reviewing the content of the TRM, we may need to make some adjustments there too. That's something I'll discuss as part of this whole issue.

    In the meantime if you are indeed of development samples with these features enabled to test them out, we can look into providing a few samples for such purposes. Ideally if you have a TI sales rep then we can work it through them, but I believe there should be other ways as well since we can open a private conversation to get company/shipping details.

    Best Regards,

    Ralph Jacobi

  • 0 in reply to Ralph Jacobi
    Prodigy 90 points

    Ralph,

    Thank you for providing the information.  I will share this with our hardware engineer.

    Just for clarification this variation will carry the same safety certs as the AM2632-Q1 ?  We are building an ECU for an automotive application.

    Jeff

  • 0 in reply to Jeffrey Hanzlik
    TI__Guru*** 135355 points

    Hi Jeff,

    Yes it would carry the same safety certs. It will just be another orderable part number, though I will have to see if we will have it for AM2632-Q1 as well - I know the initial discussion was more for AM2634-Q1.

    Best Regards,

    Ralph Jacobi

  • 0 in reply to Ralph Jacobi
    Prodigy 90 points

    Ralph,

    Thank you for the insight about which variant would have the feature.

    Jeff