• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 34 subscribers
  • Views 853 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.

TCA9509: Need more Details

Praveen GD
Guru 57737 points
Part Number: TCA9509
Other Parts Discussed in Thread: TCA9802

Hi Team,

One of my top customer is considering TCA0509 in place of PCA9509/P(NXP). In datasheet of PCA9509/P, there is a statement as follows-'Supports arbitration and clock stretching across the repeater'. This is not mentioned in datasheet of TCA9509. Can you clarify on this if our part can support this as well. 

Also, my customer ideally likes to have some other mechanism to understand the directionality of the signal rather than pulling the SDA to Vol TCA9509 as shown in figure1. Do we have any part similar to TCA9509 which can achieve this.

Regards,

Praveen GD

  • 0
    TI__Guru 62920 points
    Hi Praveen,

    TCA9509 can support clock stretching as well provided that the "low" level being driven by the the device stretching the clock is lower than the "VILC" spec of the device. (Note that this restriction applies to the "A" side with the static voltage offset. Also note there is some mention of this in Section 10.1 of the datasheet, so you may want to reference that for further details.)

    If the customer needs buffering/redriving of the signal but wants to avoid the voltage offset on the low level, they may want to consider the TCA9802 device (or others in this family). It uses a current-based detection mechanism for directionality rather than a voltage-based method, and this allows for the output low voltage to be closer to 0 V. Here is an app note describing this and other benefits in more detail:

    www.ti.com/.../slva878.pdf

    Please let me know if you have any questions.

    Regards,
    Max
  • 0 in reply to Miguel Robertson
    TI__Guru 57737 points
    Hi Ma,

    We have one more query on TCA9509,
    Vol(max) of the device on the Aside port of TCA9509 is 0.18V so please let me know if you see any issue here.
    : level being driven by the device stretching the clock is lower than the "VILC" spec of the device:

    Regards,
    Praveen GD
  • 0 in reply to Praveen GD
    TI__Mastermind 23226 points
    Hello Praveen,
    As long as any device on the A side pulls the bus below VILC min then it will transfer the low to the B side.
    -Francis Houde
  • 0 in reply to fhoude
    TI__Mastermind 23226 points
    Hello Praveen,
    I just want to be explicit about saying that you may have to adjust your pullup resistor to make sure that it is below VILC.
    -Francis Houde
  • 0 in reply to fhoude
    TI__Guru 57737 points

    Hi Francis,

    Can you please explain how to achieve this with Pull-up resistors ?
    As you know, Vol(max) of device on A side is 180 mV and the VILC(min) is 110mV. As per Datasheet, Vol should be less than VILC (110mV).
    Please consider this as priority.

    Regards,
    Praveen GD

  • 0 in reply to Praveen GD
    TI__Mastermind 23226 points

    Hello Praveen,

    You will have to find out what the strength (on resistance or Rdson) of the pull-down FET of your slave or master that is attached to the side with the static voltage offset.  Usually this is done by taking VOL max and dividing that by the defined IOL.  This is essentially the max Rdson of the pull down FET.  Then you just do a simple resistor divider equation where you have the mid voltage as the voltage seen for Vilc.  Then back calculate what the lowest pullup resistor value can be to maintain that Vilc.  It is similar to how Rp(mi) is calculated but instead of using VOL(max) you use Vilc. 

    Here is the app note for the pullup resistor calculations:

    I2C Bus Pullup Resistor Calculations 

    Remember pullup values pertain to both SCL and SDA.

    Let me know if this isn't clear.

    -Francis