Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

  • State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 65 subscribers
  • Views 246 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.

LP55231: I2C Pins Shorted to VCC/GND

Stephen B
Prodigy 20 points
Part Number: LP55231
Other Parts Discussed in Thread: LP5009

Hi all,

We've been experiencing an issue with LP55231 chips, on at least 5% of boards now. In two instances, SCL has been found shorted to GND. In another instance, SDA and SCL were found shorted to VCC. In all instances, these shorts were resolved by desoldering the LP55231 package from the boards. These shorts occurred after hours of normal operation, so unlikely to be solder bridging. We would be interested in hearing whether other customers have experienced similar issues, and/or whether there are some suggested mitigation techniques that could prevent these pins from shorting internally.

The cause of the shorts is still unknown. An oscilloscope capture upon normal power-up shows VCC (channel 1, yellow), SDA (channel 2, magenta), SCL (channel 3, cyan), and GND (channel 4, green). There is about 0.4V overshoot on the SDA line above VCC.

In the datasheet, there are no absolute maximum ratings listed for EN/SCL/SDA pins, although the recommended maximum rating for logic I/O is listed as 0V to VCC. For D1-D9, the absolute maximum rating is -0.3V to VCC+0.3V. What is the absolute maximum rating for EN/SCL/SDA pins? What would happen if we exceeded those absolute maximum ratings?

Thank you!

  • 0
    TI__Mastermind 18025 points

    Hi Stephen,

    Could you please share the schematic of the board?

    As for the maximum of EN/SCL/SDA, it should be the same as VCC. But I am not sure what happen if you exceeded those absolute maximum rating, I will discuss about this with our design and reply you later.

    BR, Jared

  • 0 in reply to Jared Zhou
    TI__Expert 6440 points

    Hi,

    The absolute maximum rating for EN/SCL/SDA pins is -0.3 - 6V, and the recommended operation voltage for these pins are 0-Vdd.

    For the problem you met, you can take below forum as reference:

    (+) LP55231: I2C stalls after re-enable the EN pin - Power management forum - Power management - TI E2E support forums

    And I noticed that you use 3.3V power, so you may not need the charge pump function for lighting up LEDs, so I would like to recommend you LP5009, which is cheaper and easier than LP55231.

    Thanks!

    Ives Li

  • 0 in reply to Jared Zhou
    Prodigy 20 points

    Hi Jared,

    I've attached an image of the schematic block with LP55231. Note that SCL and SDA pull-ups are on a different page, these are 4.7 kOhm for SCL and 1 kOhm for SDA. The enable signal comes from an IO expander chip and is pulled-down with a 47 kOhm resistor (as pictured).

    The SDA and SCL signals are originating from an FPGA sitting approx. 15 inches away, and perhaps this long trace contributes to the overshoot we've seen in the oscilloscope capture. If the absolute maximum ratings for those pins is -0.3V to 6.0V as Ives Li mentioned (Zener protected pins?) then the overshoot wouldn't have exceeded those specs. There is some current limiting from the pull-ups anyways. So unfortunately it is still unclear to me how these pins are getting damaged.

    Thanks for the response!

  • 0 in reply to Ives Li
    Prodigy 20 points

    Hi Ives,

    Thanks for the response! I have some relevant information in my response to Jared that you may be interested in as well. Regarding the post you referenced, I had looked at that and thought we may encounter that issue on occasion, but that should be a recoverable problem -- strobing the clock, resetting the chip, or power-cycling the entire board should resolve that issue. The issue we are facing is permanent damage to LP55231, measured as a short either to GND or VCC on those I2C pins. I'm not sure how we are damaging these pins though.

    Regarding LP5009, I will check out your suggestion, thanks. We do have some diodes that have typical 3.3Vf @ 20 mA, so some overhead was nice, but we are running them at a lower current to reduce the brightness anyways.

    Best,

    Stephen

  • 0 in reply to Stephen B
    TI__Expert 6440 points

    Hi Stephen，

    Thanks for your information！

    Today Jared and me are out of office，we will reply you later.

    Thanks for your understand！

    BR，Ives Li

  • 0 in reply to Ives Li
    Prodigy 20 points

    Hi Ives and Jared,

    I wanted to update the post with some recent findings. On another board, we had an intermittent short on the SCL pin and visible corrosion near the LP55231 package. We are operating in a high humidity environment, and it's possible this is contributing to electrochemical migration (ECM) / dendrite formation. After encapsulating the package with a silicon coating, these intermittent shorts went away. We haven't seen another permanent short between SCL and GND, but our findings suggest ECM due to high humidity is the culprit here. LP55231 is one of two QFNs on the board, so the relatively tight spacing between pins is maybe why we have only seen this happen to LP55231.

    Best,

    Stephen

  • 0 in reply to Stephen B
    TI__Expert 6440 points

    Hi Stephen,

    Thanks for your new finding results!

    If there is anything we could help you, please let me know.

    Thanks!

    BR,

    Ives Li