i3c using FRDM-MCXA156

[ting8999]

Subject: Integration issue on NXP I3C Mixed Bus: I3C Set Dynamic Address (SETDASA) failed for Magnetometer

Hi everyone,

I am currently working on an I3C Set Dynamic Address from Static Address (DAA) implementation using an NXP MCU (MCXA156) with the MCUXpresso SDK. I have two sensors connected to the same I3C Bus (I3C0):

• P3T1755: Temperature sensor (I3C compliant, onboard).

• Magnetometer : Supports I2C/I3C/SPI.

【Issue Description】 During system initialization, I am attempting to assign dynamic addresses using the I3C CCC_SETDASA (0x87) command:

• The P3T1755 responds with ACK and successfully switches from its static address to the assigned dynamic address (0x08).

• However, when I send the SETDASA command to the IST8216 (with busType set to kI3C_TypeI3CSdr, attempting to assign 0x09), the hardware returns a NACK, and the address assignment fails.

【Terminal Output Log】

Plaintext

P3T1755 Before DAA -> 0x48 ACK
IST8216 Before DAA -> 0x18 ACK (Ping via I2C successful)
P3T1755 After DAA -> 0x08 ACK (P3T1755 successfully switched to dynamic address)
IST8216 After DAA -> 0x09 NACK (IST8216 failed to recognize the new address)
IST8216 set dynamic address failed.

【My Analysis and Questions】 Regarding the hardware constraints, the user manual (Section 2.8 I3C interface) states:

"It has a temperature register to store the digital temperature reading that can be read by a controller via the 2-wire serial I3C (up to 12.5 MHz) or I2C (up to 3.4 MHz) interface. On the FRDM-MCXA156 board, the U5 sensor can work in one of the following modes:

• I2C mode (default mode): Populate resistors R57 and R55 (populated by default). The static I2C address (7-bit) of the U5 sensor is 0x90.

• I3C mode: Unpopulate (remove) resistors R57 and R55. The U5 sensor is assigned with an I3C address dynamically."

Although I have not removed resistors R57 and R55, the P3T1755 is still functioning correctly in I3C mode.

1. Does the presence of these resistors (R57/R55) have no impact on I3C functionality in this specific onboard setup, or could it be interfering with the I3C bus signal integrity?

Any guidance on whether this "Mixed Bus" integration approach is correct or if there is a way to force I3C DAA on the magnetometer would be greatly appreciated.

Best regards,

[Harry_Zhang]

Hi @ting8999 

The issue is most likely not related to the I3C bus configuration itself.
P3T1755 works because it is a true I3C target and supports SETDASA, so it can switch to a dynamic address.
However, IST8216 appears to be an I2C device and does not support I3C CCC commands such as SETDASA. Therefore, it will ACK at its static I2C address but NACK any attempt to assign a dynamic address.
In a mixed I3C bus, such devices should be treated as legacy I2C devices and accessed via their static address, not through I3C dynamic addressing.

BR

Harry