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.

PCA9534: I2C Expander is not working on 5Volt

Part Number: PCA9534
Other Parts Discussed in Thread: PCA9306

hi team,

we are using this I2C expander in our project. The main processor is working on 3.3 v, It is working fine with 3.3v but when we switch to 5 v it is not responding. we have also used level shifter on i2c line.

When we run I2C expander on 5v then It doesn't works and SCL and SDA levels are also falls to 2.6V . We need to run it on 5 v. Please provide solutions.

Please find attachments for schematic.sch_501-1-00787_a1.pdf

Thank you,

Subham patil.

  • Hi Shubham,

    The connections for the PCA9534s look good. Since your seeing problems at SDA and SCL, I suspect the problem may be related to the PCA9306 level shifter.
    It looks like in your schematic, you've "Do Not Stuff"ed the pull-up resistors R150 and R151 for the 5V use case. Are these populated when you move down to 3.3V?

    What else is connected to the SDA and SCL lines? It looks like there might be something trying to pull the lines down through a 4.7k-Ohm to create the 2.6V.

    Regards,
    Eric
  • Hi,

    When we are  making Pull up R128 and R128 to 500 ohms then the level goes to 4.4V but still it is not working so Please provide solutions for this .

    Regards,

    Shubham

  • Hi Shabham,

    The PCA9306 needs different pull-up values on either side of the device when operating as a level translator. In the case where VCC_PE is 5V, SCL2 and SDA2 are pulled up to this voltage by R128 and R129 (4.7k-Ohm). In the schematic, SCL1 and SDA1 are pulled up to VCC_QCA_3V3 by R150 and R151 ("Do Not Stuff").

    With these two resistors not populated, SCL1 and SDA1 could be floating (depending on what else is connected to these lines). This could cause the PCA9306 device to pull SCL2 and SDA2 down to the 2.6V you're seeing.

    On the device you're testing, are R150 and R151 included on the board? Are there any other devices on QC_I2C0_CLK and QC_I2C0_DATA?

    Regards,

    Eric

  • Hi Eric,

    SDA1 & SCL1 also has been pulled up by the 4.7K & those Pull ups are connected on the Board.
    As I mentioned,
    Same circuitry has been worked when VCC_PE is on 3.3V. When we are making R128 & R129 then Voltage level increased to 4.3V.

    But still I2C lines are not working.

    Please provide your feedback at the earliest.

    Thanks
    Shubham
  • Hi Shubham,

    Could you take a few voltage measurements to help better diagnose the problem?

    Please measure the voltages at the following points when no data is attempting to be transmitted:
    PCA9306:
    EN,
    SCL1,
    SCL2,
    Vref1,
    Vref2.

    PCA7534:
    INT#,
    SCL,
    Vcc.

    My suspicion is that there's a voltage divider being created with the pull-up and pull-down resistors that creates the static voltage on the SCL and SDA lines. These readings will help identify if this is occurring around these devices.

    Regards,
    Eric
  • Hi Eric,

    Below is the Inline response:

    Please measure the voltages at the following points when no data is attempting to be transmitted:
    PCA9306:
    EN: 4.3V
    SCL1, : 3.3V
    SCL2, : 2.7V
    Vref1,: 3.3V
    Vref2.: 5V (after 200K it is 4.3V)

    PCA7534:
    INT#,: Pulled High
    SCL, : 2.7V
    Vcc.: 5V

    I am still in doubt that how the Voltage devider being created as this same circuit is giving the correct data while running on the 3.3V.

    Thanks
    Shubham
  • Hi Shubham,

    I agree a voltage divider is unlikely with these values. I had suspected maybe some components were switched when the mode was changed to 5V to cause this, but these values suggest otherwise.


    Is this the only board you're testing, or do all of the boards exhibit the same behavior?
    It is not characteristic of this device to allow SCL2 and SDA2 to fall below SCL1 and SDA1 when idle. It's possible the device was somehow damaged and the lines in question are being weakly shorted to GND internally to the device. This would explain the voltage divider-like effect. To test this, could you GND the EN pin and Vref2 pin (possibly from the R126 pad) to disable the device and see if the lines are still sitting at 2.7V?

    Could you also test VCC_PE when it enters R128 and R129 to make sure this supply is accurate?
    Is there anything else connected to this I2C bus other than the IO expanders and the level translator?

    Regards,
    Eric
  • Hi 

    On the same I2C Bus 6 slave is connected all are working

  • Hi Shubham,

    Are these slave devices connected to the I2C_PORT_SCL/SDA nets or the QC_I2C0_SCL/SDA nets?

    Regards,
    Eric
  • Hi Shubham,

    I wanted to follow up: were you able to find a solution to the problem? Or are the I2C lines still being pulled below 5v?

    Regards,
    Eric