• State
  • Replies 1 reply
  • Subscribers 31 subscribers
  • Views 42 views
  • Users 0 members are here
Support feedback
Options
Options
Related

AM263P4-Q1: Assertion occurs in Drivers_open() when ADC external reference is configured

makoto yamashita
Prodigy 90 points

Part Number: AM263P4-Q1
Other Parts Discussed in Thread: SYSCONFIG, TMDSCNCD263P

■ Configuration  
●AM263Px MCU+ SDK: 11.00.00.19  
●CCS: 20.3.0.00014

■ ADC Reference Configuration
The SysConfig settings are configured to use an external reference.

image.png


■ Issue
When the application starts, an assertion occurs in Drivers_open() in ti_drivers_open_close.c.
The assertion occurs at the following location:

image.png

Even if the assertion for adcInstance:0 is skipped, the same assertion occurs for adcInstance:1 and adcInstance:3.
It appears that the assertion is triggered because SOC_getAdcReferenceStatus() returns FALSE.


■ Questions  
●What kind of function is SOC_getAdcReferenceStatus()?
  What are the possible causes for it to return FALSE?  

●Please advise on the recommended configuration or countermeasures to avoid this issue.

Requested response date: October 19, 2026.

  • 0
    TI__Prodigy 380 points

    Hi Makoto-san,


    Thank you for the screenshots. Your SysConfig setup is correct for an external reference: both internal reference buffers are disabled.


    What SOC_getAdcReferenceStatus() does
    It reads the status of the ADC reference monitor for the given ADC's reference group (the OV/UV "good" bits in the control subsystem). It returns TRUE only when the voltage on that group's VREFHI is inside the monitor's valid window. Drivers_open() checks one instance per monitor, which is why you see instances 0, 1 and 3 (monitors 0, 1 and 2). This check runs for both internal and external reference configurations.


    Possible causes of FALSE
    Since all three monitors fail, the cause is most likely common to every group:


    1. External reference not yet valid when Drivers_open() runs. For example, the reference IC is powered later than the MCU, or VREFHI has large decoupling capacitance and has not settled within the short wait before the assert.


    2. Wrong reference voltage. The ADC reference must be 1.8 V. A reference intended for another device (e.g. 2.5 V or 3.3 V) will fail the monitor and should not be applied to VREFHI.


    3. Reference not reaching VREFHI. On the TMDSCNCD263P control card, check that SW9.1/SW9.2 select the external path and that SW7 selects a valid 1.8 V source. On a custom board, check the VREFHI/VREFLO connections for each group.


    Recommended actions
    Measure VREFHI for each group at the moment the assert fires and confirm it is 1.8 V and stable.
    As a diagnostic, temporarily increase the wait before the asserts in ti_drivers_open_close.c (e.g. 1 ms). If the asserts then pass, the issue is power sequencing or settling time. Make sure the external reference is stable before Drivers_open() is called.
    Please do not simply disable or skip the assert as a fix. If the monitor reports FALSE, the ADC is converting against an invalid reference and the results cannot be trusted.
    If VREFHI is 1.8 V and stable but the monitors still report FALSE, please share the measured voltage, your reference circuit, and the board you are using, and we will investigate further.


    Regards,

    K.B