AM263P4: AM263P4: Clarification of Errata i2508 and use of RC_CLK as DCC reference when Limp Mode is disabled

Part Number: AM263P4

Hello,

We are using the AM263P4 and would like to clarify the intended use of the internal RC oscillator with the DCC module.

Errata i2508 (RC OSC Usage in Safety System) states that the internal 10 MHz RC oscillator serves as the safety monitor for the XTAL clock and recommends using an external high-accuracy reference clock with the DCC when monitoring the XTAL or PLL clocks.

However, the TRM (Section 6.4.6 Limp Mode) states:

"Switch to Limp mode feature is not enabled by default. The feature needs to be explicitly enabled by TOP_RCM.LIMP_MODE_EN.XTALCLK_LOSS_EN."

and also describes the dedicated coarse clock loss logic that continuously checks the XTAL clock against RC_CLK.

Our application does not require ppm-level clock accuracy verification. We only want to detect gross failures of the system clock (e.g. PLL failure or a significant clock frequency error) using the DCC. For this purpose, the accuracy of the internal RC oscillator is sufficient.

Our question is whether Errata i2508 is only applicable when Limp Mode is enabled, or whether it also affects systems where Limp Mode remains disabled.

Our current understanding is:

  • The coarse XTAL-loss detection logic may continuously compare XTAL against RC_CLK.
  • However, the automatic switch to RC_CLK only occurs when Limp Mode is explicitly enabled.
  • Therefore, if Limp Mode is disabled (default configuration), RC_CLK should still be available as an independent reference clock for the DCC.

Could you please clarify whether this interpretation is correct?

Specifically:

  1. Can the internal RC_CLK be used as the DCC reference clock when Limp Mode is disabled?
  2. Does Errata i2508 only apply when the Limp Mode feature is enabled?
  3. Is there any hardware interaction between the coarse XTAL-loss detection logic and the DCC that prevents RC_CLK from being considered an independent DCC reference, even when Limp Mode is disabled?

Thank you for your clarification.

  • Hi,

    Thank you for your query!!

    Unfortunately, the Subject Matter Expert is Out of office.
    Please expect delay in response. Apologies for the inconvenience.


    Thanks & Regards,
    Rijohn

  • Hello,

    Apologies for the late response.

    Can the internal RC_CLK be used as the DCC reference clock when Limp Mode is disabled?

    Yes, they are 2 different settings. Limp mode is the fallback of all the clocks to use RC Clock as the source when there is a failure in the XTAL. This can be enabled to be activated with either of these 3 events: XTAL clock loss, PLL clock loss or DCC0 instance error using the TOP_RCM.LIMP_MODE_EN MMR. As mentioned in the errata, RC_CLK cannot be used to verify the accuracy of XTAL but can only be used for the detection of presence/absence of it

    Our application does not require ppm-level clock accuracy verification. We only want to detect gross failures of the system clock (e.g. PLL failure or a significant clock frequency error) using the DCC. For this purpose, the accuracy of the internal RC oscillator is sufficient.

    As you mentioned this, you could use the DCC.

    • The coarse XTAL-loss detection logic may continuously compare XTAL against RC_CLK.
    • However, the automatic switch to RC_CLK only occurs when Limp Mode is explicitly enabled.
    • Therefore, if Limp Mode is disabled (default configuration), RC_CLK should still be available as an independent reference clock for the DCC.

    This understanding is inaccurate. Limp mode and use of RC_CLK as independent source for DCC is unrelated. 

    Is there any hardware interaction between the coarse XTAL-loss detection logic and the DCC that prevents RC_CLK from being considered an independent DCC reference, even when Limp Mode is disabled?

    Coarse XTAL-loss detection logic is an independent hardware logic that ALSO uses RC_CLK to monitor the absence of XTAL. You could either enable this, or the RC_CLK as reference clock of DCC. Advantage of DCC over the XTAL-loss detection logic is the ability to measure the accuracy, but that would only be possible using the external clock with higher accuracy as reference as stated in the errata. Hope this is clear.

    Regards,

    Sahana