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.

LM5157: VCC Pin internally shorted to ground debug

Part Number: LM5157
Other Parts Discussed in Thread: STRIKE

Hello TI Engineers,

I was using my LM5157 SEPIC circuit when all of a sudden it stopped working. The schematic is shown below and the layout is very similar to the LM5157 SEPIC Evaluation Board. The LM5157 circuit is a subcircuit on a a main mixed signal design board. I have used the board for a couple of days and have been using it regularly for testing. The Input Voltage, VLOAD is 9-15V. The output, SEPIC_OUT is set to 12V. The LOAD is a 12V 2-wire fan and a number of LEDs and 2 Relays that draw a maximum current of just over 1A. 

The first thing I noticed was the LEDs and FAN stopped working. I probed the SEPIC output test point and saw no Voltage in the Multimeter. There were no shorts on the LOAD side. There was a high current draw on the power supply. The LM5157 is enable is controlled by a PNP transistor Qmc2 which is controlled by a microcontroller PIN, FAB_ENABLE. VLOAD is always connected to the LM5157 circuit/ I first removed the LOAD connected to the LM5157 and Enabled it. There was still no voltage on the output (SEPIC_OUT).

EN pin is getting the correct voltage when Qmc2 is ON. Then I probed the the VCC pin and measured 0V. Measuring the other good boards I have, the LM5157 VCC pin is measuring 5V when enabled. I removed R1, 4.99Ω to see whether that was causing the short, but the VCC to Ground short remained after removing the resistor. 

Would you happen to know what might cause only VCC to get shorted to ground? The Input, Output, BIAS do not show any issue. They are not shorted to ground. There are also no issues with the LOAD since I was able to use it on another board with a good LM5157 circuit. 

Thanks,

Deniel

  • Hi Deniel,

    Thanks for reaching out. We're right now unable to view the attached schematic, it would be great if you could upload it in a different way to this thread.
    Furthermore, it would be very helpful if you could in addition attach the filled out LM5157 Quickstart Calculator for SEPIC Regulator Design, as it would help us during the schematic review and troubleshooting.

    Thank you and best regards,
    Bryan

  • Hello Bryan,

    Thanks for the response. I re-attached the schematic in my reply thread.  I have also attached the PDF version of the schematic

    Regarding the LM5157 design. I used the TI webbench to design it. I also have a previous email thread with you where I also attached a filled up version of the LM5157 with out design parameters.

    Here is the link to that thread. Let me know if you have additional questions. 

    e2e.ti.com/.../lm5157-request-for-evm-gerber-files

    LM5157_SEPIC_circuit.pdf

    Thanks,

    Deniel

  • Hi Deniel,

    Thanks for attaching the files.
    The schematic itself looks okay to me. Especially the fact that the design was running fine for several days before points towards the possibility of a one-time event where the device took damage.
    If I understand correctly, after supplying BIAS and ENABLE, VCC and SS stay at 0V and the power supply draws a high amount of current.

    There might be an internal short within the device due to damage.

    The next step would be to check how the damage could have happened.
    It is always possible that an ESD strike hit the device.
    Another cause could be an overvoltage at some of the pins during operation. LM5157 has an abs max at the SW pin of 50V.
    In SEPIC topology, the SW voltage calculates to VIN + VOUT = 15V (max Vin) + 12V = 32V
    There should still be a good margin to the abs max rating, but voltage spikes could still be able to reach that threshold.

    Do you have spare ICs of LM5157? It might be helpful to resolder a new IC onto the board and measure SW and VCC during normal operation.

    Thanks and best regards,
    Niklas

  • Hello Niklas,

    Thanks for your response. A couple of clarifications:

    1. BIAS is connected to VLOAD node which is connected to a 3 cell battery (8.25-12.6V) or a 15V adapter. We have a charging circuit that does a power pathing function. This means BIAS always had a voltage permanently connected in our testing of our main board where the LM5157 SEPIC subcircuit is located.

    2. Only the VCC pin to ground is shorted. But it does not cause a high current draw from the BIAS pin, when powering the LM5157 circuit  by itself via EN pin going above the EN threshold voltage via the FAN_ENABLE node. The power supply current only increases by 100mA. 

    3. If there were any shorts that would cause a sudden overcurrent, our whole unit would turn OFF from the Battery Short Circuit  Protection feature or the 15V Power Adapter Short Circuit Protection Feature. This was also tested during battery operation, where the battery voltage was just under 12V, which would have a calculated SW node voltage of 24V

    4. The whole PCBA was inside a plastic enclosure since the beginning of our testing. There is no direct way to contact the LM5157 circuit. 

    5. I do not have a spare LM5157 by itself but I have other PCBA that still have a functional LM5157. From probing the oscilloscope and DMM. The VCC measurement in normal operation is around 4.7V. When probing the SW pin to ground, It shows a square wave with the peak voltage of 24-26V when the battery is fully (12.6V) When powering the LM5157 circuit with 15V power adapter. VCC is still around 4.7V and the SW pin has a peak voltage of around 28-30V.

    With these new information, do you have other thoughts on what may have caused a short on the VCC pin of the LM5157? 

    Thanks,

    Deniel

  • Hi Deniel,

    Thanks for the detailed feedback.
    Please allow me some additional time to look further into this.
    I will get back to you beginning of next week.

    Best regards,
    Niklas

  • Hi Deniel,

    If the device was fully enclosed during all this time, and ESD strike cannot be the root cause.

    There is also no external connection at the VCC pin that can lead to overvoltage from an external side.
    It also does not sound like any abs max limits are violated in this application.

    The next steps I would take would actually involve starting a failure return process, the the broken unit gets analyzed to find out what cause the damage of the IC.
    Are you familiar with the failure return processes at TI?
    A lot of information can be found here:
    https://www.ti.com/support-quality/quality-reliability.html

    Best regards,
    Niklas