TXB0108: TXB0108PWR- Output Stuck HIGH (1.8V) despite Correct Input and Isolated Pin

Part Number: TXB0108


Hi TI,
We are experiencing a functional failure with a logic level translator IC in our latest production batch. This is the first time this issue has occurred in our device history, previous units using the same design have functioned without error.

The IC is intended to shift data signals from a 3.3V Controller (VCCB) to a 1.8V Modem (VCCA). While the input signals are reaching the IC correctly, the output remains unresponsive.

Testing and Observations:

  • Power Rails: Both VCCB (3.3V) and VCCA(1.8V) were verified using an oscilloscope. Both rails are stable, within the recommended operating conditions, and show no significant noise or fluctuations.

  • Signal Integrity: During hard reset cycles, we monitored the controller-side inputs. The data patterns are correct and meet the VIH/VIL thresholds. However, the modem-side (VCCA) receives no data.

  • Isolation Test: To rule out external PCB contention or modem-side loading issues, we lifted the output data pin from the pad.

    • With the pin physically isolated from the rest of the circuit, the IC still failed to toggle.

    • The isolated pin remains stuck at a constant 1.8V (HIGH).

  • Impedance Check: We performed a continuity and resistance check between the isolated output pin and the VCCA rail. There is no short circuit detected, ruling out a solder bridge or internal short to the supply rail.

    Could you please review the schematic and the behavior described? Specifically, we would like to know if this "stuck-high" behavior on an isolated pin points to an internal latch-up state or a specific failure mode known for this family of translators. 

    Thanks and regards,
    Manish Dhumal


    image.png

  • Hi Manish,

    Please note that VCCA> VCCB level shifting operation is not supported by this device, and the schematic/ description shared conflict this. Can you confirm which is correct? Does the anomaly happen in every channel or signal direction from B->A? This device does not support open drain drivers as well. 

    Regards,

    Jack

  • Hi Jack,

    Thank you for your prompt feedback and for catching that discrepancy.

    Please accept my apologies; I inadvertently swapped the voltage values in my previous message. To clarify and confirm the schematic configuration:

    The actual design configuration is:

    • VCCA(Modem Side): 1.8V

    • VCCB (Controller Side): 3.3V

    • Signal Direction: VCCB (3.3V) --> VCCA (1.8V)

    The failure is localized to one channel and persists even when the output pin is lifted.

    Regards,
    Manish.

  • Hello Manish,

    We are current out of office for the holidays. Please expect some delays in our response. Thank you!

    Regards,

    Josh

  • Can you measure the resistance between the stuck-high pin and VCC? The TXB has weak output drivers (4 kΩ); a much smaller resistance would indicate that the chip is damaged and has a short.

    Is this only a single chip? In general, the most likely cause is overvoltage, either on VCC or on the I/O pin; with only a single channel affected, VCC is unlikely.

  • Hi Clemens,

    I have measured the resistance between the stuck-high pin and VCC as requested. The multimeter is reading OL (Open Loop), which confirms there is no low-resistance path or short-circuit between the I/O pin and the VCC rail.

    This is the only chip in the batch exhibiting this behavior.

    Regarding overvoltage, the supply rail is highly regulated by a TLV73333PDBVT (3.3V LDO). Given that this is an isolated failure and the supply is stable, overvoltage damage is unlikely.

    Regards,
    Manish