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.

PCA9546A: Vpass MIN MAX explanation

Part Number: PCA9546A
Other Parts Discussed in Thread: PCA9548A

Hi Guys

Our customer is using our device for I2C level shifter. In our datasheet, Vpass:

"In order for the PCA9546A to act as a voltage translator, the Vpass voltage must be equal to or lower than the lowest bus voltage. For example, if the
main bus is running at 5 V and the downstream buses are 3.3 V and 2.7 V, Vpass must be equal to or below 2.7 V to effectively clamp the downstream bus voltages."

Why do we have to limit Vpass lower than lowest bus voltage? 

What is it about the Internal detailed logic circuit?Could you explain why the Vpass is lower than lowest bus voltage?

Thanks

-Pengfei 

  • Hey Pengfei,

    I think there may be a misunderstanding as to what Vpass is and how to choose Vcc based on Vpass.

    "Vpass voltage must be equal to or lower than the lowest bus voltage" Vpass is defined in our datasheet as the lowest pull up voltage on the channels (main channel included) but can be chosen as something lower.

    "Why do we have to limit Vpass lower than lowest bus voltage? " You do not need to choose a lower voltage than Vpass but choosing a lower Vpass forces you to choose a lower Vcc for the device.

    "What is it about the Internal detailed logic circuit?"

    This device uses a pass FET architecture. In order to pass a low you need the Vgs to be greater than Vth.

    "Could you explain why the Vpass is lower than lowest bus voltage?"

    If you choose a lower Vpass then the Max Vcc for the device goes lower. The Vpass is used as a definition solely for the purpose of using figure 11 in the datasheet.  For example, if 2.7V is the lowest voltage on the bus then Vpass is 2.7V or lower. Chart 11 states that if I used 2.7V as the Vpass then my Vcc for PCA9548A MUST be 4V or lower (I could use Vcc as 3.3V for example) but I cannot exceed 4V (I'm using the typical curve in green).

    Think about this in the sense of a passFET where Vcc feeds into the gate.

    If Vth is 1.3V @ Vcc=4V and my pull up voltage on one side of the bus is 5V for example but 2.7V on the other side then Vgs is 4V-2.7V which is 1.3V which is NOT greater than Vth so the two sides are high impedance (open switch) in this case my device keeps 2.7V seperated from 5V.

    Now lets say I choose Vcc=5V for my Vcc of PCA9548A and I still have a 2.7V bus and a 5V bus. This means the Vgs is 5V-2.7V which is 2.3V. 2.3V is greater than the Vth which in my last example was 1.3V. This means the impedance between the 5V bus and the 2.7V bus is no longer high impedance but instead low impedance (closed switch) so the two buses are no longer isolated.

    Now in this last example, My lowest bus voltage will be 2.7V like before but instead I choose Vcc=3.3V because Vpass =2.7V or lower, I pick Vcc to be 4V or less (3.3V is less in this case). So the Vgs in this case is 3.3V-2.7V which is 0.6V. 0.6V is less than 1.3V (The Vth we decided to stick with). The two buses are seperated because the FET is acting in the cut off region.

    I prove this in simulation here:

    I hope this makes sense to you. If you have further questions, I'd be happy to address them.

    Thanks,

    -Bobby

  • Thanks Bobby. Got it. You mean that when "1" is transferred between master and slave. Our pass FET will work on cut off region not Variable resistance zone. Right?
    Thanks
    -Pengfei
  • Hey Pengfei,

    That is correct. If Vcc is chosen correctly then when "1" is sent from the main channel to a secondary channel, the pass fet will seperate the two pull up voltages correctly.

    Thanks,
    -Bobby
  • Got it. Thanks Bobby.

    -Pengfei