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.

Why do your TLC3702CP's keep burning up ?

Other Parts Discussed in Thread: TLC3702

Your TLC3702CP's keep burning up. They smoke and get so hot they crack down the center and melt the plastic socket under them. The first time it happened, I thought it was a quirk. But it happened again last week. We operate them at less than 14V. The last one may have degraded shortly before it burned up, but I'm not sure. They are in the middle of a circuit, no leads are exposed to external connections that might have ESD.

Please help.

  • Hi Steve,

    The TLC3702 should be reliable with a supply nominally to +16 V, and even with the absolute maximum +18 V without experience a catastrophic failure. When we see device packages damaged to the extent you describe that is usually an indication that the IC's power supply pin was subjected to one, or more, electrical over-stress (EOS) events. What most often occurs is a higher than expected voltage transient occurs on the VCC line turns on the normally off ESD cell that connects internally between the VCC and VSS pins on. That cell is for out-of-circuit ESD protection during handled and never intended to be turned on when the device is in a circuit and powered.

    The ESD cell can latch on during an EOS and will source high current from VCC, to VSS. There is very little resistance in that and the current can immediately rise to hundreds of milliamps, or amperes, depending on what the power supply can source. Since the area of the ESD cell is very small it has very limited power handling capability. Once it is exceeded the area becomes extremely hot and eventually the silicon melts resulting in a short circuit. The heat generated during the event is great enough to overheat and deform the plastic molding compound the encases the IC die. The result is a melted area in the plastic, to splitting the package. You can see an example in the image shown below.

    It is possible for other circuits internal to the IC to break down during an EOS event so that too could provide another short-circuit mechanism.  

    My suggestion is you monitor the supply voltage being applied at the TLC3702 VCC pin during power on, operation, and power off, using a DSO. Use low capacitance 10x probes. You may need to verify input and output levels as well during those same power sequences. The ESD cells from those pins often terminate back to the internal power supply circuit paths. Look for transient, or operating voltage levels, that exceed the maximum values listed in the TLC3702 datasheet.  If the circuit has different operating modes the pins should be monitored as the circuit is exercised through them as well.

    Regards, Thomas

    PA - Linear Applications Engineering

  • Thank you for your reply. The photo looks like the damaged devices. However, the circuit is driven from a 15 VDC power supply, through a diode, so that's well within specs. One of the two outputs is connected through a 1N914 diode to a 1 uF ceramic capacitor to pull it low. However, it should never discharge the capacitor, only keep it low so it won't charge. The maximum voltage on the 1uF capacitor is 7V. But if it were possible for the 3702 to try to discharge the capacitor, would it have this EOS effect ? All four inputs to the 3702 have at least 10K in series with them.

  • Hi Steve,

    Would it be possible for you to share the schematic? Lables showing where the voltages appear would be helpful. If it is proprietary I could have you send it to my email.

    Thanks, Thomas

    PA - Linear Applications Engineering

  • Yes. What is your e-mail address ?
  • Hi Steve,

    kuehl_tom@ti.com

    Regards, Thomas

    PA - Linear Applications Engineering

  • Confidential schematics, and an explanation of the circuit, have been sent to you.