• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 34 subscribers
  • Views 1353 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.

SN74CB3Q3384A questions and TCA9548 questions

Michael Huang3
Intellectual 2460 points
Other Parts Discussed in Thread: SN74CB3Q3384A, TCA9548A

Hi team,

1 My customer uses SN74CB3Q3384A to translate 5V to 3.3V I2C bus. I am not sure of it. It seems to work properly...Could you please kindly check?

2 I can only see a MOSFET inside and connect the A and B port. and Rdson is small to 3-8ohm. How can a MOSFET translate 5V to 3.3V logic, which is what I don't know clearly. Could you help kindly show me the theory?  From what I know,  Vax=Vbx, if it is really a MOSFET.

3 My customer uses TCA9548, too. In datasheet, it can translate 5V to 3.3V from 5V-master  to 3.3V slave. My customer's situation is master is 3.3V, slaves are 5V and 3.3V. Is that OK and have no risk?  I also don't know how this part translate, could you please kindly explain it? (it also has only FET inside...)

Best Regards

Michael 

FAE

  • 0
    TI__Guru 62350 points

    Hi Michael ,

    I will move this post into the signals forum initially .

    I feel the CB3Q3384A is not being operated within the recommended conditions . The Vcc should be from 2.3 to 3.6V and the inputs will be up translated from 3.3V to 5V and 2.5V to 3.3V level . There is an internal charge pump which enhances the Vcc voltage level at the gate . The VGS is sufficient to overcome the threshold of the nchannel MOSFET to have rail to rail switching .

    If you desire to have down translation capabilities , then you can consider CBTD3384or CB3T3384 switch .

    1884.scda008_cbxxswitchws.pdf

  • 0
    TI__Genius 16510 points

    Hello Michael,

    I can answer your TCA9548A questions.

    The TCA9548A can be used to translate from any combination of voltages in the range of 1.8V to 5.5V. It does not matter if the master is at 1.8V or at 5.5V. The only thing that you must ensure is that the VCC going to the TCA9548A is the same voltage as the lowest I2C bus voltage that the device will see.

    For example, if your master is 3.3V and your slaves are 5V, then you must connect the VCC pin to a 3.3V power source.

    Another example, if your master is 5V and your slaves are 3.3V, then you must connect VCC pin to 3.3V again, because 3.3V is the lowest voltage source seen by the TCA9548.

    One last example, say the master is 2.5V, and you have Slave A at 5V, and Slave B at 1.8V. You would connect VCC to 1.8V in this case.

    As for how the device performs voltage translation, it uses a MOSFET and biases the gate appropriately to allow for different voltages. Think of the MOSFET as a switch for a simple example. It can either short the master and slave side together, or it can open and leave them disconnected (high impedance). What happens, is that the VCC pin determines what threshold the switch will turn on and off during I2C transactions. So if you have a master and slave side connected in the register, the switch may not actually be closed until a 0V signal is found on one of the sides (master or slave side). When a 0V is seen, the switch will close, and allow current to flow from one side to the other side, and this is how a 0 gets transmitted.

    When the voltage goes back up from 0V, the switch will automatically turn off/open (at a threshold voltage determined by the VCC of the device) and not allow current to flow from one side to the other. At this point, there is a high impedance between both sides, and so the voltages are allowed to go to what ever voltage the pull-up resistor pulls them to, and then they don't interfere with other channels until a 0V is seen again.