• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 62 subscribers
  • Views 357 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.

UCC27614: Failure at very light load.

Fredrik Häggström
Prodigy 30 points
Part Number: UCC27614


Tool/software:

Hi,  
We are developing a new product which uses two UCC27614 to drive each transistor in the MSCSM120DUM027AG. Our gate driver board had the wrong footprint for the UCC27614, so we turned it upside down and patched in the necessary pins. However, we only used one of the pins for VDD and one pin for GND. Could this be the cause of the broken driver?

Are the two GND and VDD pins bonded to the same die pad? 

Setup: 

Drive voltage +20 /-5V

Gate resistance: 3,3 ohm on, 3,3||3,3 ohm off.

Conditions at failure:

Bus voltage: 1V (one volt!)

Transistor peak current: 6A

Pulse width: 100us

Pulse frequency: 100Hz

Kind regards

Fredrik

  • 0
    TI__Genius 14145 points

    Hey Fredrik,

    Thank you for reaching out with your questions regarding the UCC27614.

    Can you please help clarify a few things for me?

    1. So are you trying to drive 1 FET with 2 UCC27614 in parallel? So that would be 4 UCC27614 and 2 FETs in total with 2 FETs being in 1 module.

    2. Can you share a picture of the modification done to mount the UCC27614? If the pins are connected correctly, this could still be causing a lot of inductance and ringing which could be an issue.

    3. Can you share waveforms of VDD, the inputs, and OUT in the condition where it is not being damaged (0V input)? Take these measurements at the pins of the driver for best accuracy.

    Thank you,

    William Moore

  • 0 in reply to William Moore
    TI__Genius 14145 points

    Hey Fredrik,

    Are there any updates or follow ups to my questions above? Or were you able to resolve this issue?

    Thank you,

    William Moore

  • 0 in reply to William Moore
    TI__Genius 14145 points

    Hey Fredrik,

    I am closing this thread under the assumption that you have resolved this issue at this point. If you have, please follow up and let us know what you found. If not, you can reply to this thread with your response to my above questions and an update on the status to reopen the thread to continue the conversation.

    Thank you,

    William Moore

  • 0 in reply to William Moore
    Prodigy 30 points

    Hi, 

    When I tried to answer on this, the whole E2E site was down. Now I'm back.

    I have attached some pictures on the patch. Its not super obvious on the picture, however, the chips are put upside down, some pins are bend down to the pad and GND is applied via the patch. All of the pins have the correct connection except for pin 1 and pin 5, they are floating. 

    All signals looked good running them at low load (1kHz, 25V) for several days, no temperature to talk about. Then one suddenly died. 
    The component that died has now a voltage drop from VDD to GND of around 1.4V. 

    When the chip died, the source to drain voltage to the power MOSFET was only 1-2V. 

    We have tried to provoke the DCDC converter (MGJ2D122005SC) to see if it produces any voltage spikes. However, it is very stable.

  • 0 in reply to Fredrik Häggström
    TI__Genius 14145 points

    Hey Fredrik,

    Thank you for the update.

    Having the gate driver installed with this patch installation could definitely influence reliability and performance as modifications such as bending pins could damage or degrade the IC.

    With pin 1 and pin 5 floating, there is no connection for VDD and ground as seen below. So the driver will not operate as intended.

    Your schematic appears to be okay.

    Is there a particular part that you are trying to replace here? Is this the need for the modifications?

    We may have a driver that can be P2P with what you are replacing.

    Thank you,

    William Moore

  • 0 in reply to William Moore
    TI__Genius 14145 points

    Hey Fredrik,

    Were you able to resolve this issue? If so, please let me know what your findings were. If not, please update with the status of your debug and with answers to my above questions.

    Thank you,

    William Moore

  • 0 in reply to William Moore
    TI__Genius 14145 points

    Hey Fredrik,

    I am closing this thread under the assumption that you have solved this issue at this point. If you have not, you can reply and reopen this thread.

    Thank you,

    William Moore