• State Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 2 answers
  • Subscribers 34 subscribers
  • Views 2150 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.

PCA9306: PCA9306

user4860691
Intellectual 310 points
Part Number: PCA9306

I have I2C slave that is pulling the I2C down when its powered off. So I'm thinking about using PCA9306 to activate and dis-activate the I2C side of the slave.

Questions:

Can the VREF1 and VREF2 be both 3.3V  

Can switch EN between 0 and 3.3V to enable and disable the PCA9306 and isolate I2C in the Slave side

  • 0
    TI__Guru 64110 points

    Hello,

    "Can the VREF1 and VREF2 be both 3.3V "

    -Yes they can be. This is a question that pops up occasionally, a more complete answer can be found here below:

    PCA9306: Supplying same voltage to VREF1 and VREF2 - Interface forum - Interface - TI E2E support fo...

    e2e.ti.com
    Part Number: PCA9306 Hi, I have a question about PCA9306. Can I supply same voltage against VREF1 and VREF2? The level translation are not required and I'm

    EDIT:

    "Can switch EN between 0 and 3.3V to enable and disable the PCA9306 and isolate I2C in the Slave side"

    -By supplying 0V to the enable pin, the B side will not be pulled low when SDA/SCL on A is pulled low. The B side will however still be pulled high because of the pull up resistors.


    Applying 3.3V on EN will turn on the device and allow communication again.

    However, if you are asking if using the PCA9306 as a buffer. No, this device does not buffer the SDA/SCL bus capacitance.

    -Bobby

  • 0 in reply to BOBBY
    Intellectual 310 points
    Are we still need a pull-up resistor in both sides (SCL2 & SDA2) when we tie VREF1 & VREF2
    Thank you very much for your prompt response
  • 0 in reply to user4860691
    TI__Guru 64110 points
    Yes, they are required on both sides. I2C operates on an open drain topology. This means we can actively drive a signal low by sinking current but we cannot actively drive it high. The purpose of the resistors is to passively pull our SDA/SCL bus high again after the master/slave releases the line.

    -Bobby
  • 0 in reply to BOBBY
    Intellectual 310 points
    Thank you Sir
  • 0 in reply to user4860691
    Intellectual 310 points

    I like to ask about your opinion using this approach, the EN pin is tied to the microcontroller. 

  • 0 in reply to user4860691
    TI__Guru 64110 points

    1) Vref1 should have a resistor on it in the case EN is powered, then Vref1 is powered, then Vref2 is powered last. If this occurs, the FET between Vref1 and Vref2 will be turned on for a short amount of time as Vref2 ramps up. During this time, the impedance between these two points will be similar to a short and a large amount of current will travel across these two points in this short time. This is similar to the case described above, the device may not break the first time this happens but over time it will.

    ^this should not be a problem if you have tied Vref1 to Vref2, as they should ramp up at the same rate.

    Thanks,

    -Bobby

    P.S. Please also see fhoude's diagram involving the FET tied to Vcc of the slave.

  • 0 in reply to user4860691
    TI__Mastermind 21745 points

    How does this work?  If the fet is a P-channel then it only works with uC control is low, thus PCA9306 is OFF.  Why are you trying to do it this way?  What is your application.  I am very interested in seeing what you are trying to do.  

    If you are trying to turn off the supply and disable the PCA9306, here is a way of doing it with a P-channel FET.  The logic for the micro will have to be negated, but this should work.  Are you trying to save money?  What are you design constraints?   

    If the drawing you have is an N-channel FET it won't work.  If you need explanation let me know.  I can explain in more detail.

    -Francis Houde

  • 0 in reply to fhoude
    Intellectual 310 points
    Thank you for your help,

    The slave pulling down I2C when its off so I'm trying to keep the slave side of I2C separated from the rest of the circuit. also the slave has a short lifetime I'm trying to cut its power as well. I can't just cut the power off the slave without cutting the its I2C communication.
    about your above circuit it will work perfect I was just drawing the same approach and yes the FET has to be a P-channel.
    So in this case I believe I don't need a 200K pull-up of the EN&VREF2

    Thank you again
  • 0 in reply to user4860691
    TI__Mastermind 21745 points
    You don't need the 200k in series with EN/VREF, but in the circuit above, I would at the very least have a place holder for the discharge resistor between gate to source. Some FETs have them built in. You want to make sure it gets turned off even if the IO is left floating.
    -Francis Houde
  • 0 in reply to fhoude
    Intellectual 310 points
    Thank you, I really appreciate your help & guidance