• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 92 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.

UCC21750-Q1: Sort Circuit Test

Kaushik Prajapati
Intellectual 350 points
Part Number: UCC21750-Q1

Tool/software:

Dear Sirs,

I am using the UCC21750-Q1 in a power supply to drive MOSFETs. I have two different power boards, X and Y. While conducting shoot-through tests, I encountered an issue with one of the boards.
Condition of Test: For both boards, the bottom MOSFET's drain and source are shorted, and I am applying an 8μs pulse to the top MOSFET.
* Board X functions correctly at 800V DC during the short-circuit test. I have attached the waveform for your reference.
Where,
1-CH : Yellow Channel = VDS Voltage across Drain and Source TOP IGBT
2-CH:  Blue Channel = Current while shorting 2X = 128Amp
4-CH:  Green Channel = Gate Signal at MOSFET 
* Board Y, however, exhibits problematic behavior during the short-circuit test, allowing approximately 960A at only 100V DC. The response time is set to 1.8μs. I have also attached the waveform for board Y.
Where,
2-CH: Blue Channel = VDS Voltage across Drain and Source TOP IGBT
1-CH: Yellow Channel = Current while shorting 2X = 960Amp
4-CH: Green Channel = Gate Signal at MOSFET 
I am unable to explain this behavior and would appreciate your guidance. I can provide additional data if necessary.
Thank you for your time and assistance.
Sincerely,
Kaushik
  • 0
    TI__Mastermind 20106 points

    Hi Kaushik,

    I do not understand why Board X CH1, the high-side Vgs, does not droop to 0V for 8us when the IGBT is closed. Maybe the pulses are only 2us?

    Are these two different IGBT part numbers? Different IGBTs will have different turn-on times, and this may explain why the second test reaches a higher current.

    Also, test 1 has a lot of resonance on the gate. Maybe it is not turning on the high-side as strongly as in the second test. What is considered problematic? In a short circuit test, I would expect the current to be very high.

    Best regards,

    Sean

  • 0 in reply to Sean Cashin
    Intellectual 350 points
    Sir,
    You are correct, and I am also puzzled as to why Vds does not drop.
    The pulse width is 8uS, but it trips and reduces the width to 2uS (response time).
    However, as you mentioned, the resonance on the gate, well it serves my purpose by preventing the current rise from exceeding 200A.
    Yes sir, both boards use MOSFETs with the same part number.
    Could this resonance in X board be related to the active miller clamp?
    Thank you,
    Kaushik
  • +1 in reply to Kaushik Prajapati
    TI__Mastermind 20106 points

    Hi Kaushik,

    Resonance like this is often related to feedback from the high-side mosfet acting like a common drain amplifier, as it passes through its saturation region on its way to the Rdson region. You can try to break the feedback loop at this frequency with a ferrite bead on the gate, designed to maximize impedance at the resonant frequency.

    The more permanent solution is to eliminate the LC switch node ringing itself. You can improve the HV decoupling to try to reduce high-side drain inductance, and possible present series ESR at this resonant frequency from the local capacitors. Electrolytic capacitors can do this, for example.

    You can also upgrade to SiC FETs. These have 1/10 the gain of a Si Mosfet in their saturation mode, and will not cause as much positive feedback. 

    Best regards,

    Sean

  • 0 in reply to Sean Cashin
    Intellectual 350 points
    Sir,
    Thank you for your valuable insights. I will implement the experiences you have suggested and see how it goes.
    Best regards,
    Kaushik