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.

SN74LVC2G17: Failed part. Possibility of Latch-up or floating input ?

Part Number: SN74LVC2G17

Greetings !

We have problem in our design. Some 5-10% of devices have broken logic buffer SN74LVC2G17DSFR.

I have several possible variants what's going on. Device has UART and power supplied through 2.5m cable, marked as hot plug on picture with ESD protection on input.

Can you please advice if some of these variants can lead to device fail ?

1) Our logic and microcontroller (uC) on board have power supply that goes wired with UART in one cable.
   We have moment during power on (several seconds or more) where Vcc power for logic IC is not available yet but output pin 1Y(AJ_UART_TX) already have around 3.3 V on it.
   So could it be violation of design which can lead to device destruction  by output voltage higher than Vcc+0.5 ?
2) Could situation described in p. 1 lead to latch-up of device ?
3) There's possibility of device connection without UART connected - only power. Can it lead to situation when floating or even fixed on some voltage like ~1.5 V input 2A (AJ_UART_RX) will turn on both HIGH and LOW side output MOSFETS and destroy device ? (actually on failed part 2Y output have 9 Ohm resistance to GND)

Also do this IC has some numeration patterns to recognize if it's not fake ?

Thanks Waldemar

  • Hi Waldemar,

    Some 5-10% of devices have broken logic buffer SN74LVC2G17DSFR.

    How are these devices breaking? Can you describe the issue?

    1) Our logic and microcontroller (uC) on board have power supply that goes wired with UART in one cable.
       We have moment during power on (several seconds or more) where Vcc power for logic IC is not available yet but output pin 1Y(AJ_UART_TX) already have around 3.3 V on it.
       So could it be violation of design which can lead to device destruction  by output voltage higher than Vcc+0.5 ?

    Yes, this is possible.

    This device includes Ioff, which is basically back-flow prevention to avoid this issue. In order for the output to be ensured to be in the high-impedance state, the VCC pin must be at 0V relative to GND.

    It looks like in your case VCC is disconnected - ie it is floating at an unknown voltage. It is possible for this to cause the device to have the outputs biased on and could result in a short to VCC or GND.

    2) Could situation described in p. 1 lead to latch-up of device ?

    This would be very unlikely. The device is tested for latch-up to 100mA. It is more likely that the current caused physical (heating) damage to the output structure.

    3) There's possibility of device connection without UART connected - only power. Can it lead to situation when floating or even fixed on some voltage like ~1.5 V input 2A (AJ_UART_RX) will turn on both HIGH and LOW side output MOSFETS and destroy device ? (actually on failed part 2Y output have 9 Ohm resistance to GND)

    No, there is no input state that can cause the outputs to turn on both the positive and negative drivers. It is possible to produce oscillations if the input is floating, and it will draw some extra power at the supply pin -- however Schmitt-trigger buffers are designed to handle this current and won't be damaged just because of a floating input.

    My recommendation would be to add a large pull-up or pull-down resistor at the input to prevent the input from floating (100kohm would work).

    Also do this IC has some numeration patterns to recognize if it's not fake ?

    The datasheet includes the topside markings on the Package Option Addendum (near the end of the datasheet, before the mechanical drawings).

    I would recommend to only purchase parts from reliable vendors to ensure you get genuine TI parts. If you think that you have received counterfeit parts, please let us know and we can get you in touch with our counterfeit parts team. Typically they will want to see the shipping label and images of the devices at a minimum.

  • Emrys, thank you for such detailed response !

  • About breaking,

    When power unit plug connected with device firstly and then apply power to power unit.

    Failed device have such yellow [1Y output ] vs blue [Vcc] waveform when power applied:

    On the other hand working devices have waveform like this:


    Removed fail part has abnormal >36 mA current consumption without any load while inputs still remain HiZ.

  • The SN74LVC2G17 operates on a nanosecond scale, so I suspect the issue is hidden in this initial connection, shown at a second-scale in the scope shot:

    My suspicion is that the supply has a voltage spike (barely visible above) when the initial connection is made that could be large / long enough in some cases to activate the output, and thus provide a back-power path for the supply. 

  • When we say, "Pin 1Y already has about 3.3V on it", it's important not to leave out the fact that this bias is provided by a 10K pull-up resistor to 3.3v rail on the far side of the hot plugged cable.  The max current from this bias is ~ 0.3mA.  The datasheet says it's OK to violate VCC+0.5v, as long as we don't violate +/- 50mA.

  • Thanks point t up Robert - it's actually written on schematic picture  with question "to 3V3 uC on other side, 10K pull up".
    It also worth to point up that 1Y pin is located on A2, B2 USB type C pins instead of SuperSpeed differential pair #1, TX, positive. Is there probability that USB 3.0 SS TX line can be powered at time of connection ?