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.

LMZ23610: Possible faulty components - regulators appear faulty in recent manufacturing run

Part Number: LMZ23610

Hello,
We have been using LMZ23610TZE/NOPB in our designs for years.
The recent manufacturing run for our mobile robot found 10/60 units to have faulty LMZ23610TZE/NOPB.

  • 3/10 failures had LMZ23610TZE/NOPB that would not boot up
  • 7/10 failures had LMZ23610TZE/NOPB that would drift, or slowly ramp up past their voltage setpoint
  • Feedback resistors were confirmed to be correct
  • There were no shorts on the board before startup
  • The design was based on this: https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=DBE3875F7AA6310E
  • All boards with the failed regulator had the following two date codes: 08AD85G3 or 08ADE8G3

Do you know if there is a way to check if we received counterfeit or faulty parts?

Thanks,
Kyle

  • Hello

    I will talk with the power module group and see how we can help.

    Thanks

  • Hi Kyle,

    Let me get in touch with the CQE (customer quality engineer) to see if we can run the date codes and possible get a look into the history of those specific lots.

    As for this device, it has been out in the market for some time without much issues related to quality.

    When I get more information I will let you know asap.

    Regarding the failure, are these units being used in a new design or is it the example same existing design that you have had for years? It sounds like you've been using this device in your system for some time so I don't think it would be a component selection issue. 

    Regards,

    Jimmy

  • Hi Jimmy,
    We can send damaged parts for FA if needed
    Cheers!
    Kyle

  • Hi Kyle,

    I think it is best to get a better understanding of use case and where to troubleshoot before sending the parts in for FA.

    In the mean-time if you can answer my questions this will help narrow down where the potential failure can be. 

    Again I'd like to reiterate the questions: "Regarding the failure, are these units being used in a new design or is it the example same existing design that you have had for years? It sounds like you've been using this device in your system for some time so I don't think it would be a component selection issue." 

    You are welcome to send the units back for FA through the customer return link (here) but I'd like to get a better understanding of why these units might be failing given the years of history you've had with this device. 

    Regards,

    Jimmy

  • Hi Jimmy,

    • This design has been active since 2016 and there have been no changes to the design in the last few years.

    • All of these failures were found during manufacturing bringup.
    • There is only the 3.3V LDO and downstream microcontrollers as a load during bringup test since 5V devices are not connected yet
    • In most cases simply replacing the regulator with another one resolves the issue unless the regulator caused overvoltage damage to the downstream PCB components 

    Please let me know if you require more information.

    Cheers!

    Kyle

  • Hi Kyle,

    Additional questions below:

    • How many boards have been built since 2016? Can you ballpark how many fall/pass since the beginning of using this design? 
    • When these boards are tested for pass/fail is it ran through an automated test that could have overvoltaged any of the pins?
    • Do you have a solder reflow profile to help me understand how these units are assembled onto your board? 
    • Can you confirm that the failed units had enough solder paste to make proper contact with all pins to board and enough solder for the large DAP that is used for thermal relief?
    • Is it possible to provide top board view of just the LMZ23610 circuit and the surrounding circuits? I'm curious to see if there are any discoloration of components that might be indicative of thermal stress.

    Thank you for your patience and feedback so far. Knowing all these assembly and testing information will help accelerate this troubleshooting effort. On my end, I've requested the internal CQE team to help look up the date code and check for lot history. When I get more information on that I'll provide further feedback.

    Regards,

    Jimmy 

  • Hi Jimmy,
    Great questions, please see the responses below:

    • How many boards have been built since 2016?
      • I'm not too sure, but since I joined the project in 2018 its around 500 units built. 

    • When these boards are tested for pass/fail is it ran through an automated test that could have overvoltaged any of the pins?
      • I don't believe so, they are a large board at low quantity so they are hand tested with multimeters by staff. The input to the 5V Buck is supplied by a large battery (26V nominal, 30V max) and they are tested without a significant load.

    • Do you have a solder reflow profile to help me understand how these units are assembled onto your board? 
      • I'll have to check with the manufacturer

    • Can you confirm that the failed units had enough solder paste to make proper contact with all pins to board and enough solder for the large DAP that is used for thermal relief?
      • We use epoxy-filled vias on the PCB to prevent solder wicking into the vias and have reasonable confidence in the solder paste adhesion. I would see this being a field issue but not a power-on failure that we are seeing

    • Is it possible to provide top board view of just the LMZ23610 circuit and the surrounding circuits? I'm curious to see if there are any discoloration of components that might be indicative of thermal stress.
      • The visual inspection found all the pins to be connected and the solder was shiny.
      • The components do not have any cracking or visible discoloration or damage.
      • Do you have a more secure way for use to send pictures of the pcb?

    Let me know if you need more information

    -Kyle

  • Hi Kyle,

    For the board image, you can send it directly to my corporate email at j-hua@ti.com.

    As for the input supply, battery supplies can easily spike up to twice the input voltage during hot-plug event. Can you provide the schematic of this design if you haven't already? During the spike it is possible that the input voltage seen from the LMZ23610 is 2X Vbatt ( 52V / 60V) and could potentially be damaging the input pin since it is only rated up to 40V;.

    As such for battery applicatons, it is recommended to use an electrolytic cap with high ESR to help dampen the voltage spike and account for long lead inductances to the input/ground pair. 

    Regards,

    Jimmy  

  • Hi Jimmy,
    I am working with the production team to get failure pictures.
    From what I have seen there is no visible damage.

    One note on the battery, We don't connect a battery for the initial bringup testing.
    Instead, we use a programmable PSU set to 26V @ 0.2A so there is a low chance of the input voltage exceeding VIN Max.

    Please advise the next steps.
    Cheers!

    Kyle

  • Hi Kyle,

    I've received your personal email response and look to continue troubleshooting this issue from there.

    In the mean time I will take a thorough look at the board and check with the internal team to see if they were able to pull up any information on this device's lot/date code.

    If I come up with any more questions I will comment on the direct email you sent with the team. Let's close this thread for now and continue from the direct email chain.

    Regards,

    Jimmy