+ Post New Thread
Results 1 to 2 of 2
  1. #1
    Member level 1
    Points: 423, Level: 4

    Join Date
    Oct 2017
    0 / 0

    Overvoltage in a regulator volatge circuit

    Hi All,

    I have recently seen an electronic circuit which I think has been under high voltage. It is supposed to work at 24V but It can handle up to 50V. The circuit is basically two regulators voltage and a microprocessor.

    The first regulator voltage is MC34063ABD and it is configured as page 11 on its datasheet, it steps down the input voltage to 5V. The second regulator voltage steps down the voltage from 5V to 3.3V to feed the microprocessor.

    I said it was overvoltage because the MC3406ABD is not regulating and it is visible damage, the inductor which is between both regulators voltage is completely burnt and open circuit and the microprocessor is shorted.
    I’m trying to think in the sequence of how the components were failing and I think it is as follows:
    Firstly the MC34063ABD fails because more than 50V applied to it, then it stops switching and all the voltage outputs in SWE pin. The voltage drops in the diode Zener (as page 11 on datasheet).
    The second regulator voltage, the MCP1700T can’t handle that level of voltage and let the voltage reach the microprocessor which fails causing a dead short. As the micro is shorted and there is voltage dropping in the Zener (the output of the first regulator voltage), the micro continues pulling current until the inductor is completely burnt and open circuit.

    Does it makes sense this sequence of events? should I expect different evidences if more than 50V were applied to the circuit?.

    Many thanks.

    •   AltAdvertisment


  2. #2
    Advanced Member level 5
    Points: 38,202, Level: 47

    Join Date
    Mar 2008
    1791 / 1791

    Re: Overvoltage in a regulator volatge circuit

    Overvoltage of the front end regulator is a credible theory
    but in failure analysis you should beware of your impulse
    to believe the first good idea that comes along. Rather,
    collect the possibilities and work the evidence to get the
    probabilities, as priorities for deeper digging (like, there
    may be other ways to fry a switching regulator, but some
    may not be credible given the assembly and access and
    local threat "population" - maybe a "screwdriver short"
    on the output, while a credible way to damage a part,
    is not credible in the hardware as it sat).

    There is always some "wandering around in the basement,
    in the dark, looking for the light switch" element to FA in
    the early going. Don't try to run ahead, take your time
    and diligently explore the failure modes, effects so your
    analysis has a solid chance of including the real villain.

--[[ ]]--