LSF0102: Questions of LSF0102

The device that is driving the lines low must be able to sink the current through all pull-up resistors. What is that 1.8 V device? What is its drive strength?

Hi Clemens,

The 1.8V side is an open-drain configuration of the QUALCOMM CPU, so the drive current flows through the pull-up resistors.
We have tried to change the pull-up resistors on the 4.2V and 1.8V sides to 1K, but this did not improve the I2C signal voltage on the 4.2V side.

Many Thanks,

Jimmy

What model is that CPU? What is the drive strength of its I²C pins?

Hi Jimmy,

In addition to Clemens' response, please provide the non-functioning waveform for further debugging. You may also see the TXS0102 that contains internal one shot circuitry to help accelerate the rising edge to the desired output voltage.

Regards,

Jack 

When the I²C bus is idle, no current should flow. Are there any other devices connected to the I²C lines that could pull them low? Please check for shorts to GND.

Hi Jimmy,

Clemens' is correct here- please check if there is are any external pulldowns that are not shown on the schematic or if there are any leakage paths.

Regards,

Jack 

Hi Jack,

The I2C bus was only connected to the TLC59208FIPWR, so there were no other devices to pull the bus low. The I2C bus NACK waveform on the 4.2V side is shown below:

However, I have replaced the LSF0102DCTR with dual MOSFETs, and it has the correct waveform. Therefore, I suspect that the port B internal sink current of the LSF0102DCTR is not able to pull up to high voltage.

Many Thanks,

Jimmy