• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 34 subscribers
  • Views 755 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.

TCA4311A: VOL of I/O

Antony Lin
Genius 13480 points
Part Number: TCA4311A
Other Parts Discussed in Thread: P82B96, TCA9517, TCA4311

Hi,

 In TCA4311A datasheet, the VOL is defined as below.

Here we have some related questions.

  1. Is this VOL spec allied to both sides (IN and OUT) of the chip?
  2. When the application scenario is implemnted as below from the datasheet, Can still refer the VOL spec to be <0.4V still?  Or the votlage level could be higher?

Thanks!

Antoyn

  • 0
    TI__Guru 73331 points

    Hey Antony,

    "Is this VOL spec allied to both sides (IN and OUT) of the chip?"

    This device uses a dynamic voltage offset, so if VoL on the "out" side is 400mV then the "in" side of the device will be shifted up by about 100mV (max of about 175mV). Now it seems in your diagram that you have a long distance bus/cable. This means there will be losses across the cable, mainly the cable may add parasitic resistance which will cause the receiving 4311 a larger VoL.

    For reference, the dynamic offset I am talking about:

    "When the application scenario is implemnted as below from the datasheet, Can still refer the VOL spec to be <0.4V still?  Or the votlage level could be higher?"

    The VoL spec was under specific conditions. If for example, the driving side pulls low with VoL of 200mV and the offset voltage is 175mV then the outside will be 375mV and if there is a 25mV drop across the cable then the total will be 400mV the receiving device sees. The inside of the receiving unit (assuming 175mV shift) will see 575mV.

    Now if you were pulling low from the driving side at 400mV and we assumed the same conditions then VoL for the receiving slave would be 775mV.

    I actually do not like the idea of using TCA4311A for long distance communication. It has a rise time accelerator which can make cable communication bad. I have also seen cases where pairing two TCA4311As together sometimes does not play well together.

    Can you tell me more about your application and the needs?

    1) Largest VoL acceptable?

    2) Max clock frequency?

    3) Vcc?

    4) distance of cable?

    We can likely find a device which may be better suited for the application.

    Some thoughts that come to mind are P82B96 and TCA9517.

    Thanks,

    -Bobby

  • 0 in reply to BOBBY
    TI__Genius 13480 points

    Hi Bobby,

    The application scenario looks like below which uses only one TCA4311 chip to support I2C hot-plug feature.  Will you suggest if TCA4311A a good optin for this case or not?  THanks!

    1) Largest VoL acceptable?

    • 0.3*VCC as SMBUS spec.

    2) Max clock frequency?

    • 100K Hz

    3) Vcc?

    • 3.3V

    4) distance of cable?

    • 280mm

    Antony

  • 0 in reply to Antony Lin
    TI__Guru 73331 points
    Hey Antony,

    TCA4311A would be better to use than TCA4311 (non A) as it does exactly the same thing as the non A and loosens the timing of the non A.

    From what you have provided this is currently the only hot insertion device we have so we do not have any other solutions to replace this currently.

    Thanks,
    -Bobby
  • 0 in reply to BOBBY
    TI__Genius 13480 points

    Hi Bobby,

    Thanks.  So, how do you think TCA4311A can meet customer's needs based on the information provided earlier?

    THanks!

    Antony

  • 0 in reply to Antony Lin
    TI__Guru 73331 points

    Hey Antony,

    At 100kHz and 3.3V, this should not be a problem. 0.3*Vcc should be easy to meet as well.

    The cable distance mixed in with the rise time accelerator of the device means there could be some overshoot when we try to pass highs. Though to me, this distance is pretty low and at 100kHz shouldn't be too much of a concern. The customer could add some DO NOT POPULATE series resistors and a DO NOT POPULATE capacitor to GND so we can try to dampen any overshoots later if we do see any.

    Thanks,

    -Bobby