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.

TS5MP646: (TMUX646 Pin-to-pin upgrade supporting 6-GHz bandwidth available) could be used for i2c switch?

Part Number: TMUX646TS5MP646
Other Parts Discussed in Thread: TS5MP646

Hi team,

Could you help check if TS5MP646 be used for i2c switch? Customer only use one lane, so they want to know if the other could be used as i2c switch. Could you help suggest?

Thanks,

Will

  • Hi Will,

    The customer can use any of the other channels for I2C SCL and SDA lines. Please make sure the customer has the required pull-up resistors on input and output per their driver strength and I2C bus requirement. 

    Thank you,

    Regards,

    Sandesh

  • Hi Sandesh,

    Thanks for your feedback. Customer has following questions about the i2c application:

    1. Should they use difference lanes for I2C SCL and SDA, I mean use D2P for SCL, use D3P for SDA to avoid cross noise within lane signal(D2P, D2N).

    2. Is there any risk that the I2C signal effect the close MIPI signal?

    The following is there schematic, could you help check if this is OK?

  • Hi Will,

    1. We don't spec the cross talk noise on 1 channel  we only have cross noise between different channels - please see the figure below:

    Typically the closer the switch the worse the cross noise is - but at I2C frequencies you will have a very low cross talk noise so realistically you can place SDA and SCL on the same channel and it won't negatively affect the system too much. However since you have open channels you will most likely get the best performance by separating the I2C signals - but the improvement may not be much better than using the same channel. Either way the application should work fine. Just ensure the proper pull-ups are installed on both sides of the switch.

    2. There is cross talk noise - however I2C frequencies are going to be typically < -80dB  (< 0.01% of I2C signal will show up on MIPI pathway typically). So the I2C won't have a large effect on the MIPI signal being passed through the switch. That being said the MIPI signa may cause noise on the I2C channel - we have typical cross talk spec'd at -40dB at 1.25GHz which means about 1% of the MIPI signal (if running at 1.25GHz) can leak into other channels. If the clock frequency is different please refer to the first graph I listed to find the approximate typical cross talk noise. However these are still relatively low magnitudes that are leaking through the switch so the application should still work but there will be some leaking of signals into other channels but this is generally low - but it increases with an increase of frequency.

    Otherwise schematic looks good - just remember to add pull-ups for I2C signals on both sides of the switch.  

    If you have any other questions please let me know!

    Best,

    Parker Dodson

  • Hi Sandesh,

    Thanks for your spport. I have another question about TS5MP646.

    In the SPEC 9.2.2 part of TS5MP646, it is mentioned that "TI recommends that unused I/O signal pins be connected to ground through a 50 Ω shunt to prevent signal reflections and maintain device performance" .When Channel3 is used as IIC Bus, the 50Ω resistor of Channel3 will be removed. Will this have an impact on MIPI SI ?

    Looking forward to your reply. Thank you.

  • Hi Baron,

    Since Channel 3 is being used for I2C it does not need termination resistors as it is used. The 50 Ohm terminations are only for I/O pins - this is to help prevent reflections on the transmission lines due to impedance mismatches of the floating I/O pins. 

    The concern with MIPI integrity shouldn't be too large of a concern as the I2C line is ensuring channel 3 is being used and therefore not floating. The SI of the MIPI signal will suffer a bit from the crosstalk of the I2C signal - but as I pointed out above this should be less than 0.01% of the I2C signal will "leak" into the MIPI signal. 

    This application in particular shouldn't suffer major SI issues for the MIPI signal.

    Please let me know if you have any other questions! 

    Best,

    Parker Dodson

  • Hi Parker,

    Thanks for your comment.

    Now we reserve Channel3 for IIC, after we get PCBA, we will test MIPI SI to see the data quality.

    Thanks.