• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 22 subscribers
  • Views 336 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TXS0108E: TXS0108ENMER application question

Alice Fan
Intellectual 2150 points
Part Number: TXS0108E
Other Parts Discussed in Thread: TXU0304, SN74CBTLV3126, SN74CBTLV3245A, TCA39306

Tool/software:

Hi team,

Would like to ask if the TXS0108ENMER is suitable for SPI / eSPI / I2C communication?
Do TI have any other recommended materials for SPI/eSPI/I2C level shift? SPI and eSPI can reach up to 66M, and I2C can reach up to 1M

Current Applications
SPI application is 16Mbps, 1.8V <-> 3.3V;
eSPI application is 20Mbps, 1.8V <-> 1.8V;
400k for I2C applications, 1.8V <-> 3.3V;
The datasheet mentions that it is bi-directional.

From the datasheet, it should be able to support the current design, but there is a problem. The clock data of SPI and eSPI will be delayed, causing reading problems.
I2C seems to have an error in data bidirectional communication

Thanks!

Best regards,

Alice Fan

  • 0
    Guru 317310 points

    For SPI, a TXS translator might work, but the best solution is a unidirectional translator like the TXU0304.

    For eSPI, you have the same voltage on both sides. Do you want buffering? If yes, Why?

    For I²C, the TXS might work.

    Please note that the TXS does not work well with long traces/cables; it is likely that these oscillations are caused by that.

  • 0 in reply to Clemens Ladisch
    Prodigy 30 points

    Hi Clemens,

    For eSPI, you have the same voltage on both sides. Do you want buffering? If yes, Why?

    -> That's because Host and Device have different power domain, and we found that device would has a ~0.3v leakage from eSPI bus that might cause Host CPU unstable during power up in previous project, so we would like to use a level shift to do power domain isolation. 

    TXS does not work well with long traces/cables; 

    -> Do you have the data for how long it can support?

    thanks~

    Tim

  • 0 in reply to Tim Chang
    Guru 317310 points

    1. To disconnect two buses, use a simple switch like the SN74CBTLV3126 or SN74CBTLV3245A.

    2. The limit for the capacitive load is about 150 pF. Note that the oscilloscope probe might affect this.

  • 0 in reply to Clemens Ladisch
    Prodigy 30 points

    Hi Clemens,

    1.  SN74CBTLV3126 or SN74CBTLV3245A

    -> eSPI data pin can be configured as bi-directional, do these two part can handle bi-directional communication?

    2. So if we want to use a I2C level shift with a long trace, best solution should be TCA9617. 

    Sorry there has one more question, do you have any suggestion part for I3C buffer? need to support 1v or1.8v i3c voltage level.

    Thanks~

  • 0 in reply to Tim Chang
    Guru 317310 points

    1. These are analog, bidirectional switches; a closed switch behaves like a small resistor.

    2. Yes, a translator without edge accelerator works better with long traces. For I3C, use the TCA39306.

  • 0 in reply to Clemens Ladisch
    Prodigy 30 points

    Thank you very much, i have no further questions for this topic.