This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

PCA9518: SMBus waveform abnormal for PCA9518

Part Number: PCA9518

Hi,

We use the TI PCA9518 for the SMBus expand function, and now, we found the SMBus waveform is abnormal, please refer to below test validation:

1. We test the PCH to PCA9518 SCL and SDL waveform, the waveform is oscillating abnormally, the waveform as below:

2. We test the PCA9518 to RAM, the SDA is abnormally:

3. We modify the design as below:

1) We push up the VPP2_5 with 3.3Kohm resistor on the circuit SDA/SCL to RAM

2) Parallel the 3pF and 10pF capactance;

There are no any improvement for the waveform;

4. We test the NXP PCA9518PW instand of TI PCA9518, the waveform is normal;

Per check the compara with TI and NXP PCA9518, thare are no more any difference between these two device except the below 

So please kindly help to check how to solve this issue, thanks.

The NXP PCA9518PW for your evaluation.

PCA9518(NXP).pdf

  • The PCA9518 cannot be used with rise-time accelerators.

    Please show the schematic, including all I²C devices.

  • Hi Zhang,

    As Clemen's pointed out, the waveform appears to have potential contention with a rise time accelerator. Are you able to provide a block diagram of your I2C bus?

    It might also be worth looking at the EXP pins to verify if they are seeing any noise or crosstalk. (less likely since the waveform would look different if they did)

    -Bobby

  • Hi Bobby& Clemens,

    Thanks for your kindly support.

    Please refer to the schematic diagram and Layout:

    Because SMBus waveform is normal for NXP PCA9518, so we consider if there are any design difference between TI and NXP? 

  • Hi Zhang,

    I still believe that the issue is related to a rise time accelerator like Clemen's pointed out. TI's PCA9518 can overshoot a little bit when it regulates it's VoL to the offset voltage. I think the NXP part with the scopeshot you provided is also overshooting but looks like it's not overshooting enough to trigger the rise time accelerator on the bus. I suspect you have a rise time accelerator in one of your other blocks like the DDR block.

    You may want to try to make the pull up resistor weaker and add like 50pF of cap on the channel that is seeing the issue to see if you can dampen the overshooting.

    -Bobby