• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Subscribers 21 subscribers
  • Views 349 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

TXS0102: TXS0102

TXS0102: TXS0102

Tom Li4
Prodigy 140 points
Part Number: TXS0102
Other Parts Discussed in Thread: TCA9800, P82B96, , TCA9509, TCA9517

Hi Support Team,

We are having an issue in long-distance I2C applicaTion: it is the VIL(Low-level input voltage): By using a I2C buffer P82P96 we have extended

the distance, but the issue is I2C communication is not reliable. The following image explains our issue. and looking for advise.

  • +1
    Guru 243220 points

    The TXS is a passive switch and does not buffer the signal; it would make the VOL worse.

    Try using a buffer like the TCA9800.

  • +1
    TI__Guru* 96591 points

    Hello,

    Do you have any scope shots of the issue?

    If your VOL is too high, the TXS translator won't improve the VOL value since it passes that through (note that 0.4V VOL is only guaranteed if the input voltage is 0.15V).

    I would first recommend to increase the pull-up resistor value to 10k or 20k to see if it reduces your VOL -- and if you don't have a scope shot of the issue, I would recommend to get one to see exactly why your communications are unreliable. It may be unrelated to this.

  • 0 in reply to Emrys Maier
    Prodigy 140 points

    Hi Emrys,

    Yes I do have scope screentshot:  CH1 is signal generator (simulate MCU's I2C, connected to local P82B96's Sx/Sy ), CH2(BLUE) is the output of local PB2B96 (Tx/Rx), Ch3 (pink) is the one(THE ISSUE) coming out of REMOTE P82B96(Sx/Sy) (this meets its specification (Sx/Sy: 0.8-1V (@3mA, 0.67~0.79@200uA). I did increased PULLUP to 10k (and higher). tried to decrease the current. The funny thing is , the VOL of Sx Sy becase slightly higer when increasing pullup). Anyway, eventhough figure out a way to make P82b96'S VOL lower, its minimum still 0.67V(as per the spec.). Why this P82B96 make the VOL that high? and is there any suggestion/solution to solve our issue?

  • 0 in reply to Tom Li4
    Prodigy 140 points

    Because of this 0.67~0.79V Vol of P82B96, and our sensor's input Vol rage requirement is =<0.8V, it is too marginal. I am strongly suspecting this is the reaosn of unreliable I2C communication issue. So our first job is to find a way to decrease the Vol of P82B96 (as said, increase pullup do not work, meanwhile have to keep a minimum current flowing through the open drain). So we are looking for help how to decrease the Vol of PB2B96, or find a proper level translator. The VOH is not a problem now.

  • 0 in reply to Tom Li4
    TI__Guru* 96591 points

    Hey Tom,

    I will ask the experts on the PB2B96 to respond - since you opened this thread on the TXS0102, it came to my team instead. It may take up to 24 hours for them to respond.

  • 0 in reply to Emrys Maier
    Prodigy 140 points

    Thank you Emrys.

    I was planning to use TXS0102, untill you helped and given me the answer which can not be found on the datasheet.

    Tom

  • 0 in reply to Tom Li4
    TI__Guru 60112 points

    Emrys,

    Thanks for directing this thread to us and supporting it thus far!

    Tom,

    I've notified the applications expert of this thread and they'll be responding accordingly. Thank you for your patience.

    Regards,

    Eric Hackett

  • 0 in reply to Eric Hackett
    Prodigy 140 points

    Thank you Eric,. Looking forward to it.

    Tom

  • +1 in reply to Tom Li4
    TI__Mastermind 49110 points

    Hi Tom,

    Judging from your block diagram, since you are using 5V on the P82B96 and 3.3V on the I2C responder/device side; the best device to sit inbetween the two sides to redrive the signal would be the TCA9509. Have B side face the P82B96 and A side face the I2C responder/devices. You may need to remove the pull up resistor on A side or weaken it since A side has an internal current source already. 

    Thanks.

    -Bobby

  • +1 in reply to BOBBY
    Prodigy 140 points

    Hi Bobby,

    Thank you very much for suggesting TCA9509, it looks like promising solution. We have further questions before moving forward to TCA9509:

    1. What is the determinative standard when saying B side allowing 400pF bus capacitance? 2. If our I2C bus needs longer distance, say 50m -100 m long, can we add additional TCA9505 in serial , to achieve the long distance purpose? 3. regarding A side VILC(input low contention level), the datasheet said the device(slave)'s VOL must belowe VILC min. For our applciation (B-SIDE 5V interfacing BUS, A-side 3V3 interfacing sensor SLAVE),  does this mean the sensor's VOL must below 0.11V? If this is TRUE, then TCA9509 brings a new problem: our sensor has spec saying VOL max =0.4V( it did not say Typicle neither minimum of VOL). In that case, can be parallel an external resistor pull-down to TCA9509's A side in order to decrease VOL?

  • +1 in reply to Tom Li4
    TI__Mastermind 49110 points

    "1. What is the determinative standard when saying B side allowing 400pF bus capacitance? "

    The 400pF value was chosen due to the I2C standard of 400pF max cap allowed on an I2C bus for standard mode and fast mode. 

    "2. If our I2C bus needs longer distance, say 50m -100 m long, can we add additional TCA9505 in serial , to achieve the long distance purpose?"
    Not using the TCA9509 as the cable driver. The P82B96 should be able to drive that capacitive loading but the TCA9509 likely wouldn't be able to. You can use the P82B96 for the cable ends and then use the TCA9509 to redrive the signal to a lower level for your onboard I2C devices.

    "3. regarding A side VILC(input low contention level), the datasheet said the device(slave)'s VOL must belowe VILC min. For our applciation (B-SIDE 5V interfacing BUS, A-side 3V3 interfacing sensor SLAVE),  does this mean the sensor's VOL must below 0.11V?"
    Correct. 

    "If this is TRUE, then TCA9509 brings a new problem: our sensor has spec saying VOL max =0.4V( it did not say Typicle neither minimum of VOL). "

    VoL max of 0.4V is an I2C standard spec at a IOL of 3mA. The key word here is at 3mA. Typically designers will provide more margin for these max specs. At 1mA (A side's current source without external pull ups), your max value assuming linearity (an NFET in an open drain bus would operate in linear region of operation) would be 0.133V. We typically don't see issues with meeting this spec if the device is I2C compliant and has built in margin AND there are no series resistors or external pull ups on the bus.  

    "In that case, can be parallel an external resistor pull-down to TCA9509's A side in order to decrease VOL?"

    I'm not sure how you would do this while maintaining bidirectionality. Do you have a split I/O for SDA (incredibly rare to have)? If so, then yes, you could place a unidirectional driver to redrive the signal. 

    If you're worried about the TCA9509 spec. Then you can try the TCA9517 and have B side face the 3.3V side and A side face the P82B96. 

    -Bobby