• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 35 subscribers
  • Views 883 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.

PCA9534A: Device is showing up at wrong I2C address ...

Edwin Reich
Prodigy 90 points
Part Number: PCA9534A
Other Parts Discussed in Thread: PCF8574

We have grounded the A0,A1,A2 pins of this chip, and hence I expected to find the device at address 0x38, but when I run an i2c scanner it doesn’t find anything at 0x38.  It does find a device at 0x48, which isn’t any of the other devices on the I2C bus.

We have another PCA9534A that our tech wired up and it responds to the address correctly.  Is it possible that we have a bad batch of parts that we manufactured with OR could it be a different line of chips OR ...? So far they all are at the wrong address and with the wrong address, it does not function right with the rest of the commands.

Any help is appreciated.

  • 0
    TI__Guru 53305 points

    Hi Edwin,

    Thanks for bringing your question to E2E. I've never heard of an entire batch of devices responding to the wrong address. It's interesting to hear that it occurs for bits that are not configurable. Please help with some more info so we might identify where the issue originated. Based on the device comparison table on the PCA9534A datasheet, there doesn't appear to be a device we have with the 0x48 address possibility, though it's possible this is a different device. Though, I suspect this would be distinguishable based on device markings. 

    • Do the devices respond to the incorrect address (0x48) when addressed directly? Or is does this response only occur during the scan? Are other devices being detected correctly in the same scan?
    • Do all of the affected devices come from the same shipment? If so, what is the lot number on the shipping label? If not, please share the markings on the top of one of the units. A picture may be easiest as it would also show relative position and size of the markings. 

    • Does the separate device tested by the tech come from the same lot/shipment as the affected devices? If not, could an affected device be removed and tested in the same environment as the correct device?

    Let me know if you find out any more information in the meantime. 

    Regards,
    Eric Schott

  • 0
    Guru 314230 points

    The PCA9534A has I²C address 0111xxx, which indeed corresponds to 0x38 in your circuit.

    Section 5 of the datasheet shows that there is no expander with address 1001xxx. i2cdevices.org/addresses/0x48 or learn.adafruit.com/i2c-addresses/the-list?view=all#0x48-2630920-8 do not show anything obvious (the PCF8574 entry is wrong).

    Are the 0x48 chips marked in the same way? Did you get them from an authorized TI distributor?

  • 0 in reply to Eric Schott1
    Prodigy 90 points

    The devices do respond to the incorrect address (0x48) when addressed directly.

    All other devices are detected correctly in the same scan.

    I'm looking into whether all the devices are from the same shipment.

    The separate device is from a different shipment.

    We are looking at removing the part from an affected device to determine its behavior in the test environment.

    We are also placing a known good part onto an affected device to see if that resolves the issue.

  • 0 in reply to Edwin Reich
    Prodigy 90 points

    So, it turns out the 0x48 address was a red herring.  One of the other chips in the system responds to a custom address and a broadcast address of 0x48.

    It turns out it doesn't respond to any address in the system despite our being able to see the I2C address at the chip with a scope.

  • 0 in reply to Edwin Reich
    TI__Guru 68810 points

    Edwin,

    Are you able to post the o-scope shot you're seeing?

    Edwin Reich said:
    it doesn't respond to any address in the system despite our being able to see the I2C address at the chip with a scope.

    Do you also have a schematic you can provide for review?

    -Bobby