STUSB4761 - low level input voltage for I2C not compliante to spec?

[MKN]

Hi,
the maximum I2C input low voltage is stated as 0,35V for the STUSB4761.

Is this a typo?

Because in NXP I2C spec this level is stated as 0,3VDD 

We are using a PCA9516 I2C Multiplexer (0,5V Low) and observe massive connection failures with STUSB4761.

Regards

Michael

[TDK]

Nothing would indicate it's a typo. 0.35 V is less than 0.3*VDD so is compatible with the NXP I2C spec.

> We are using a PCA9516 I2C Multiplexer (0,5V Low) and observe massive connection failures with STUSB4761.

Likely will need to debug this further. Look at the SDA/SCL signals on the line to see if they are valid or not. Ensure you have appropriate external pullups present.

[MKN]

0.35 V is less than 0.3*VDD

That is the problem.
I2C Spec allows up to 1.5V for a valid Low_In (0.3*Vdd @ 5V )

STUSB fails if I2C low level is above 0.4V. (No ACK)
Furthermore it will detect a logical high above 1.2V.
1.2V is less than 1.5V.

Please explain how this could be compatible to NXP I2C spec.

I replaced the PCA9515 against a Nfet Switch / Level Shifter on SDA and the failure is gone.

[TDK]

> That is the problem.

You're right, I misread.

To the original question, it's unlikely to be a typo, especially as you've confirmed the issue. It appears this is done to prevent latch-up.

[MKN]

 >this is done to prevent latch-up.

Where is this explanation from?
Can't find this in NXP I2C Spec

STUSB4761 Datasheet:
2.3.4 I²C interface
An I2C interface can be used for non-volatile-memory customization.

This interface ist not electrical compliant to I2C Spec. and incompatible with common I2C Multiplexer.
A very annoying fact for a I2C component without address pins.

Thank you for comfirmation.

[NBALL]

Hello

Correct, please refer to table 16 for level triggering which need to be use. There is no typo .

Best regards

Nathalie