• State TI Thinks Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 334 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.

LM5176: LM5176 Boot1 & SW1 pin getting damaging

Shankar Kanki
Prodigy 180 points
Part Number: LM5176

Hi Brigitte & TI Team,

We are facing an LM5176 IC damaging issue while testing above 5A load with input voltage 33V & 30V output voltage.
Issue Details are as below:
We used the same values as mentioned in the LM5176 evaluation board, only updated feedback resistance so that the output voltage would be 30V.
It was working well for VIN 33V, and Vout 30V till 5A load. But above 5A load LM5176 IC getting damaging i.e system stopped working, output voltage not getting 30V ( only showing 4 or 3 V).

Note:
After analysis, it is observed that the impedance between Boot1 & SW1 is very low ( ~ 100 Ohm) & sometime high side mosfet (which is connected to HDRV1, SW1 & Boot1) also gets damaged.
This high side mOSFET has 60V max operating VDS voltage.

As of now, there are 5-6 qty LM5176 were damaged in the same manner.

We need your quick support on the same.



  • 0
    TI__Guru 53781 points

    Hi Shankar,

    thank you for using the E2E forum.

    Please check the voltage rating of the output caps on the EVM. The are not rated for that high output voltage.

    So I assume they got damaged and worst case also triggered a damage of other components.

    Best regards,

     Stefan

  • 0 in reply to Stefan Schauer
    Prodigy 180 points

    Hi Stefan,

    Forgot to update, the output capacitors and other lower rating devices has been replaced with high voltage rate.

    e.g output caps replaced with higher voltage.

    We checked these caps but it was not damaged 

  • 0 in reply to Shankar Kanki
    TI__Guru 53781 points

    HI Shankar,

    thank you for the update.

    Can you then check on the SW1 and SW2 for under- and overshoots?

    Some signal levels seems to be outside of the allowed range and damaging some components.
    This might be the SW with esp. in this case having undershoots - please check them against the values given in the abs max table of the Datasheet.

    Best regards,

     Stefan

  • 0 in reply to Stefan Schauer
    Prodigy 180 points

    Hi Stefan,
    We had checked SW1 waveforms but it is not outside of the allowed range.
    Hence we could not understand why LM5176 ICs are getting damaging.

  • 0 in reply to Shankar Kanki
    TI__Guru 53781 points

    HI Shankar,
    did you also check SW2?

    Can you share the measured waveform?

    What settings did you use in the scope: Bandwidth of the scope? Any bandwidth filter enabled?

    Best regards,

     Stefan

  • 0 in reply to Stefan Schauer
    Prodigy 180 points

    Hi Stefan,
    We have used 100MHz bandwidth and not bandwidth filter enabled.
    please find the below waveforms for SW

  • 0 in reply to Shankar Kanki
    TI__Guru 53781 points

    HI Shankar,

    what was the operation condition for that measurement?
    Is this SW1 or SW2?

    Thanks,

     Stefan

  • 0 in reply to Stefan Schauer
    Prodigy 180 points

    Hi Stefan,

    The operation condition was 33Vin, 30V output, load 1 or 2A.

    This was the sw1 waveform.

  • 0 in reply to Shankar Kanki
    TI__Guru 53781 points

    HI Shankar,

    do you see any change when you go closer to the failing condition?

    Just to confirm: the replaced cap have the same capacitance then the original ones. When using MLCC they have a similar DC Bias value?

    Best regards,

     Stefan

  • 0 in reply to Stefan Schauer
    Prodigy 180 points

    Hi Stefan,
    We haven't seen any change when we were close to failing but one observation is that the switching was alternate meaning marked red color switching was not happening when LM failed.


    Important:
    Considering ongoing discussion & resolution we are losing time & our demo timeline with our valuable customer.
    Requesting TI Team, Please look into this issue with the TI internal team and expert and provide us immediate solution on the same.
    We count on you for this issue.

  • 0 in reply to Shankar Kanki
    Prodigy 180 points

    Hi Stefan,
    Any update on this issue?

  • 0 in reply to Shankar Kanki
    TI__Guru 53781 points

    Hi Shankar,

    this forum is maintained by the applications team - so you get already support from the experts.

    It may also accelerate the process it you could provide answers to all questions.

    To get the full picture and have the full alignment:

    - Can you please provide a detailed list of the components you have updated:

    Designator Value  Partnumber DC bias for MLCC capacitor

    - I assume you have checked the input voltage being stable in the full operation range - please confirm

    - I am not sure if i understood your last comment right:
    "We haven't seen any change when we were close to failing but one observation is that the switching was alternate meaning marked red color switching was not happening when LM failed."
    -> does this mean that the switching frequency gets reduced to 50% on the SW.
    -> this happens in BuckBoost Mode but if you are running with VIN 33V and Vout 30V the device should already be in BuckBoost mode
    -> I understood that failing was a damage of the device - have you observed any other things?

    -> how fast was the device failing once entering the critical operation condition?

    -> Have you observed an increase of the temperature of any components - please check and provide a thermal image

    Please send a scope plot showing this signals
    (each line with one scope plot / showing 3-5 switching cycles and another one with ~ 10ms time:


    - VIN / VOUT / COMP

    - VIN / VOUT / COMP   (VIN and VOUT ac coupled)

    - SW1 and SW2 ( showing 3-5 cycles and also zoomed in to the falling edge for both with ~ 10ns/div)

    You can also add a Zener Diode parallel to the Boot Cap to prevent an overcharge of the BOOT input.

    Based on your description the Boot logic seems to get damaged - the Zener diode can prevent that.
    This over charge normally happen due to over or undershoots on the SW node which could not be proven by the scope plots you sent before but this could still be a measurement issue. With a 100MHz probe a few ns spike can not be detected. You can also try to further zoom in to the falling edge of the SW to get more resolution - the scope plots above do not show the horizontal resolution.

    Best regards,

     Stefan