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.

LM2901: LM2901 failure when startup

Part Number: LM2901
Other Parts Discussed in Thread: LM2903

Dear TI:

     When used LM2901, found some devices failure.

     The failure occur at PCBA startup, and the failure device PIN2 impedance is 200Ohm to GND, and the OK device PIN2 impedance is infinite to GND.

     The schematic design as below, what reason would causes the LM2901 damage?

     Thanks!

  

  • Hello Eric,

    Start-up issues can be difficult to determine as not all the voltages are defined or stable at startup.

    DANGER: This is an off-line supply and you cannot ground the "ground" line. You must use a line isolation transformer. I have to say that..

    I would get a scope shot of the start-up operation, watching the comparator output, +20V supply and pin 4 (-IN).  Ground the scope at pin 12 of the comparator.

    The inputs and outputs of the LM2903 do NOT contain ESD clamps, so they are sensitive to negative going transients. Any transients less than -0.3V, or greater than 36V on the pins can cause damage.

    One thing I do see is that the S8550 transistors are only rated for VCE of 20V max - you are already at 20V, so any large transients could cause them to breakdown. You may want to consider using higher voltage transistors (>50V). If they brake down, any Q1 transients will go directly into the comparator output.

    I think monitoring several start-up events with a scope will reveal the issues. It's hard to tell form just a schematic.

  • Dear Paul:
    Customer want to know if the transients less than -0.3V or greater than 36V, how long the LM2903 will be damaged ?
    Based on the customer's design and the current phenomenon, is IC more likely to be damaged by overvoltage?
    Thanks!

    Best regards!
    Thanks!
  • Hi Eric,

    as Paul already said: "Any transients less than -0.3V, or greater than 36V on the pins can cause damage." "Any" here means that even only one transient violating these limits can cause damage.

    Have you made the measurements Paul recommended? You should really carry out these measurements! There's something wrong with your cicrcuit and you need to find it out.

    Kai
  • Hi Eric,

    have you already made the measurements?

    Kai
  • Hi Kai:
    Customer test by themselves, and don't reappear again the failure before.