Verdin AM62 Reset: 1.8V not toggling

Hello,

EDIT: This post can be deleted, there is no issue. The reset on the Verdin AM62 is just handled differently than on the iMX8 Module. We saw, that on the iMX8 the PWR_1V8_MOCI signal is low until the reset button is released (which we saw with a multimeter), and on the AM62, it is just asserted for ~20ms, which we did not see at first with a multimeter.

we are currently evaluating a Toradex Verdin AM62 module on a Dahlia Carrier Board V1.1 and are observing unexpected behavior of the PWR_1V8_MOCI rail during reset.
Expected Behavior
According to the carrier board design guidelines and our understanding of the reset behavior, when pressing the RESET button (and keeping it pressed), the PWR_1V8_MOCI (1.8V output for peripherals) should go to 0V.
Observed Behavior
When pressing and holding the reset button:
The 1.8V rail briefly drops from 1.8V to approximately 1.7V
After this short dip, it returns to 1.8V

  • The rail does not turn off (does not reach 0V)

We have tested this behavior on three different (new) Verdin AM62 modules, all showing the same behavior.

For comparison, when performing the same test with a Verdin i.MX8 module, the 1.8V rail properly drops to 0V while the reset button is pressed.
This makes us wonder whether this could be:

  • An intentional difference in reset behavior of the AM62 platform, or
  • A potential issue/bug specific to the Verdin AM62 PMIC configuration.

We are using the Dahlia Carrier Board V1.1

  • There is nothing connected on the carrier board (no external peripherals or loads)
  • The measurement is performed directly on PWR_1V8_MOCI

Could you please clarify whether this behavior is expected for Verdin AM62, or if the 1.8V rail should indeed drop to 0V during reset?

Thank you in advance.

Edit:
With the AM62 SoM: The PWR_1V8_MOCI signal goes low for ~21ms, even if reset is pressed longer.

Hi, I’m the Toradex AI Assistant.

I might make mistakes or occasionally get you wrong. Our support team will step in if needed.


Looking at your reset behavior observation with the Verdin AM62, I can provide some insights based on the Verdin family specifications.

Expected Reset Behavior

According to the Verdin Family Specification, during a reset state, the module should implement either a “cold reset” or “warm reset” (1). The specification states that the preferred reset mode is a “cold reset,” which means the PMIC shuts down all the rails on the module and drives the CTRL_PWR_EN_MOCI and CTRL_SLEEP_MOCI# low to turn off also the carrier board rails (1).

However, some modules may implement a “warm reset” instead, where the rails on the module, including CTRL_PWR_EN_MOCI, are kept up and running while only the reset signals are asserted (1). The specification notes that the Verdin module datasheet contains information on which type of reset is implemented by the module (1).

Power Management Signal Behavior

The CTRL_PWR_EN_MOCI signal is described as an “Enable signal for the power rails of the carrier board peripherals” that “remains high during sleep modes” (1). This signal operates at 1.8V and is a CMOS output (1).

Potential Platform Differences

Your observation that the Verdin i.MX8 properly drops the 1.8V rail to 0V while the Verdin AM62 only drops to ~1.7V and returns to 1.8V suggests there may be platform-specific differences in reset implementation between these modules.

Recommendations

  1. Check the Verdin AM62 datasheet specifically for reset behavior documentation to confirm whether it implements cold or warm reset
  2. Verify the CTRL_PWR_EN_MOCI signal behavior during reset - this signal should go low during a cold reset
  3. Consider that this may be expected behavior if the AM62 implements warm reset rather than cold reset

The behavior you’re observing across three different modules suggests this is likely intentional platform-specific behavior rather than a defect, but the specific module datasheet should provide definitive confirmation.

Best regards, Toradex AI Assistant.


Did you like my answer? :+1: :-1:

If I answered your question, please tick the “Solution” checkbox.

For follow-up questions to me, please use @ToradexAI in your post.

Yes, this is expected behavior. According to the Verdin Family specification, PWR_1V8_MOCI is a 1.8 V reference output for the I/O rails. It may go low during a module reset, but this is not guaranteed. To monitor the module’s reset state, you should use the CTRL_RESET_MOCI# signal.