• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 1 answer
  • Subscribers 36 subscribers
  • Views 679 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.

PCA9306: Multiple Devices Application

Allan Fan
Expert 7090 points
Part Number: PCA9306

Hi expert,

Customer used two PCA9306 in parallel, one with 5V pullup, another one with 3.3V pullup. Is there any consideration need to pay attention? 

So far, this parallel schematic can't work for sure but if customer disable 3.3V, the 5V that path can work. (contrarily, 3.3V pullup that path can work if disable 3.3V that path.) 

High level of SCL_3V3 and SCL_5V0_0 are no problem at 3.3V and 5V.

But customer meet a problem that high level of SDA_3V3 and SDA_5V0_0 always keep at 2.5V, seems as same as Vref2=Vref1+Vth=1.9+0.6=2.5.

Below picture is captured of SDA_3V3 and SCL_3V3:

Regards,

Allan

  • 0
    TI__Guru 72981 points

    Is Vref2 pull up resistor tied to a 5V source?

    Why is R1887 populated but R1888 do not populate?

    -Bobby

  • 0
    Prodigy 40 points

    The high voltage side are 3.3V and 5V, and pull-up resister tied to 3.3V and 5V souce.

    About R1888 NC of SDA_5V0_0, this is because there is internal pull-up of 5V I2C device.  

  • 0 in reply to Mike Hsieh
    TI__Guru 72981 points

    The waveforms seem to show that SCL is 3.3V and SDA is 2.5V. Generally we see this kind of issue when a pull up resistor isn't present while there is a PCA9306 (or I2C switch) on the bus leaking current in from the higher voltage side into the lower voltage side until the voltage is Vgate-Vth. 

    The other possibility is if the gate voltage is higher than expected resulting in the PCA9306 (or I2C switch) trying to regulate itself back to it's cut off region/mode by setting the source to Vg-Vth. My current guess is this is likely the issue.

    Can you get a scopeshot of the EN pin and Vref 1 pin to verify if the voltages are what we would expect them to be?

    -Bobby

  • 0 in reply to BOBBY
    Prodigy 40 points

    Attached please find the EN (pin 8) and Vref1 (pin 2) scopeshot of PCA9306 in parallel of 2V->3.3V and 2V->5V:

    Please help to check

    Thanks

    -Mike

  • 0 in reply to Mike Hsieh
    TI__Guru 53555 points

    Hi Mike,

    Bobby is out of office for the holidays and will return in early January.

    It look like the gate voltage at PCA9306s EN pin is not at the proper level. The nominal voltage at the EN and Vref2 pins will be Vref1 + 0.6V when regulated correctly. Please check the values of the 200k-ohm resistors to ensure there is one more leakage than expected from this supply path. 

    Regards,
    Eric Schott 

  • 0 in reply to Eric Schott1
    Prodigy 40 points

    Hi Eric

     I have double checked 200k-ohm, the resistance is correct. Besides, I have tried to replace with 100k-ohm and 400k-ohm, but the result is not changed.

     Merry Christmas

    -Mike

  • 0 in reply to Mike Hsieh
    TI__Mastermind 26806 points

    Mike,

    Apologies for multiple different engineers jumping in during the holidays. Slight smile

    Can you confirm whether there are pull-up resistors on nets SDA_5V0 and SCL_5V0?

    Thanks,

    Danny

  • 0 in reply to Danny Bacic
    TI__Expert 7090 points

    Danny,

    Before we confirm the pullups on SDA_5V0 and SCL_5V0, I would like to share one information with you that once customer replacing PCA9306 to PCA9616, the issue is gone. The difference I found is PCA9616 integrate EN pullup resistor. Hopefully it can helps to clarify the issue. 

    Regards,

    Allan

  • 0 in reply to Allan Fan
    TI__Mastermind 26806 points

    Allan,

    Pull-up resistors are needed on each segment of the I2C bus for this component. The PCA9306 doesn't have integrated pull-up resistors, and it doesn't have the capability of driving high.

    Are you sure PCA9616 was the component you intended to say? From what I know, NXP's PCA9616 is a different component altogether. However, your comment about integrated pull-up resistors is still relevant, as I agree that makes a difference.

    Best,

    Danny

  • 0 in reply to Danny Bacic
    Prodigy 40 points

    Yes, I confirmed  there are pull-up resistors on nets SDA_5V0 and SCL_5V0.

  • 0 in reply to Mike Hsieh
    TI__Mastermind 26806 points

    Mike,

    Mike Hsieh said:
    About R1888 NC of SDA_5V0_0, this is because there is internal pull-up of 5V I2C device. 

    Could you possibly try populating R1888 to see if it addresses the problem?

    And, while you're looking at this, could you let me know the supply voltage level connected to VREF1 on U12?

    Best,

    Danny