Mosfet Failure

Discussion in 'The Projects Forum' started by magantiamala, May 8, 2012.

  1. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    Dear Sir/Madam,
    I have started working on new project and i am totally new to it..My company has assigned me the first new task i seriously need everyones help as early as possible..because i have very less time:(

    Development of a test setup to characterize the robustness of power MOSFETs during short time overload in linear operation

    The task is to define a test setup for different types of power MOSFETs (e.g. Infineon CoolMOS, TrenchFET ...) for measuring the robustness of the transistors in linear mode, for which there is no or only limited specification in the datasheet.

    Expected results:
    - Robustness (survival time) dependent on current and drain-source voltage
    - Failure mechanism of the transistor
    - Variation of robustness of a certain type of MOSFET
    - Comparison of transistor technologies (e.g. Superjunction vs. Trench)
    - Comparison of implementations (e.g. Superjunction-FETs from Infineon vs. ST)
    - Paralleling of MOSFETs, how good is the current sharing between paralleled devices
    - Test setup efficiently useable for tests of larger quantities of parts, with automation (e.g. control of function generators, switches, automated test result evaluation)
    - Test setup universally useable for evaluation of new devices when they become available on the market

    This is my total project..

    Let us consider a normal circuit with resiatance, voltage ,mosfet and anything we like we can design but it when current or volateg increases the mosfet dies i need some ideas how and what values should be taken to connect the circuit and i want to know few ideas???

    First my question is that i want to know few ideas what are the causes (which point )the mosfet dies so that we can be below that point???
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,103
    3,038
    This is an enormous, complex and specialized project. If you're looking here for help, I suspect that you are not comfortable taking on such a challenge. Why is the company asking you to do something you seem unprepared for?

    I'd spend time with the vendors, and try to learn as much as possible from them. There may be industry standard test procedures and certainly they will all have their own recommended tests to share with customers, so that their specifications can be verified. If you're lucky, the tech service folks may help you establish a testing lab if they believe it might lead to sales. Is your company a customer for mosfets?

    Heat and temperature are going to be key variables to measure and control. Very challenging.
     
  3. #12

    Expert

    Nov 30, 2010
    16,278
    6,789
    Many datasheets for individual MOSFETs contain test fixture schematics but, making a test fixture for a lot of different MOSFETs will require serious engineering skills.

    As for your last line, specific failure modes, gate blowout (sensitivity to static electricity on the gate), over-voltage on the drain, overheating, and "safe area of operation" are a few.
     
  4. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    #12:Thank you for the reply, i want to know is at point the mosfet is going to be dead soo that i can be below that point.
     
  5. #12

    Expert

    Nov 30, 2010
    16,278
    6,789
    I think you just sprained my brain.
    It's impossible to answer that except to say, "Stay within the specifications provided by the manufacturer".
    The only possible way to test a transistor and be sure not to blow it up is to test for whether it meets the manufacturers guaranteed parameters.
     
  6. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    Can anyone please explain me the behaviour of mosfet and provide me few youtube links to see practiaclly how the mosfet behaves?

    what are the limiting factors?

    and why does it break?
     
  7. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    Limiting factors are listed in data sheet, it'll break when you do what exceeds the data sheet.
     
  8. wayneh

    Expert

    Sep 9, 2010
    12,103
    3,038
    In normal use, I'd wager that >90% of MOSFETs that fail, do so by overheating. They're also destroyed by static, so there are other failure modes.

    But again, the vendors will likely have a better perspective and advice on this.
     
  9. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    Can anyone explain me the current voltage characterstics and what are the differences between superjunction and trench mosfet.
     
  10. rohitd

    New Member

    Jan 29, 2011
    14
    0
    find the book on electronic devices and circuits by Milman. it is a good book. also see the books on this website. they are really good. i appreciate the writer.
     
  11. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    rohitd:Thankuuu:)
     
  12. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    I am new in testing department, i have never worked on testing..can anyone help me kindly the steps how to test the failure mechanism of transistor??
     
  13. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    You need some test circuits, and then drive them with more and more power.

    Get a thermometer as well and measure!

    Then you get to a point 96C OK for 5 hours, 105C and the MOSFET makes "pop". Actual temperature might be different from measured temperature (higher).

    Yes too much voltage + heat together, there is really a point when they will burn out.
     
  14. wayneh

    Expert

    Sep 9, 2010
    12,103
    3,038
    I still can't believe they have given a newb such a complex challenge. I think you should seek more direction from your supervisors. What do they really want to know?

    The experimental conditions you establish depend on what question you want answer. One question might be, "How hot can this part get before it fails within 1 minute?" Another question could be, "What is the MTBF when operated at 50°C ambient and rated continuous current?" These are both valid questions but require completely different tests. Maybe you want to know how much static they can survive - another completely different test.

    Whatever tests you decide on, all the variables that matter must be controlled and recorded. Voltages, ambient temperatures, device temp, current flows, air flow (cooling) rates, heat sink size, and so on. Then, you'll need to test several pieces to establish a mean and standard deviation. It would be useless to run a test and forget to record ambient temperature, for instance, since this is likely a key factor.

    Have you followed my suggestion to ask the manufacturers for help? They will all have standard tests they perform for quality control and may be willing to share some of those test protocols with you. They may also have very valuable insights for you on how to test whatever it is you want to learn about.
     
    RamaD likes this.
  15. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    @wayneh:Thanks for your kind suggestion..Ya i have asked my supervisior regarding the data sheet so that i can know abt what point the mosfet fails, so that i can be below that point..he told me that he has only limited data in the data sheet and they wont belive much in that:(

    I am a student and its my thesis, so its like i have to learn and implement...its kinda tough job i know but i should finish my thesis to get my degree...
     
  16. #12

    Expert

    Nov 30, 2010
    16,278
    6,789
    Pick ONE parameter to test and build a way to test that ONE parameter.
    When you get that right, pick ONE other parameter and build a way to test that.
    When you get that right, pick ONE other parameter and build a way to test that.
    Repeat as necessary.
     
  17. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    Can anyone help me in desiging the test setup for lekage detection of mosfet.
     
  18. #12

    Expert

    Nov 30, 2010
    16,278
    6,789
    Choose one mosfet. A leakage test might be in its datasheet.
    If not, we will start from there.
     
  19. wayneh

    Expert

    Sep 9, 2010
    12,103
    3,038
    To see farther, stand on the shoulders of giants.

    Read everything you find here and here. Lots of good stuff.
     
  20. magantiamala

    Thread Starter New Member

    May 8, 2012
    25
    0
    Can anyone please provide me the information on what kind of effects are appearing externally before mosfet damage..

    It would be really kind if anyone helps me..I need the external effects causing before mosfet failing.

    thanks in advance.
     
Loading...