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.

TXS0104E-Q1: The voltages are abnormal

Part Number: TXS0104E-Q1
Other Parts Discussed in Thread: SN65DSI84-Q1, TXS0104E

Hello TI teams,

    The TXS0104 is used on our board, the schematic is show as below, the VCCA is powered up before VCCB, sometimes the voltage of TP_SCL/I2C_SCL and LVDS_EN/EN are abnormal, could you help me to solve this problem. Thanks 


  • And what are those voltages? Can you show an oscilloscope trace?

  • Hello,

    In addition to Clemens' answer, what is the state of the OE pin during this time? Are you able to tie OE low via a weak pulldown resistor until both power supplies are powered up to see if anomaly is gone? 

    Not directly related to the issue, but please note that the current external pullups will contend with the internal 10kohm pullups, resulting in higher VOL levels. We recommend to keep external pullup values to be >50kohms if pullups are absolutely necessary by the system. 


  • Hello Clemens,

        The power up sequence is showed as below. I change R75 from 10Ω to 10KΩ and add a 1uF capacitor on OE pin.


       When OE is enabled, the SCL and SDA signal is abnormal, especially the SCL.


       First, when I short OE to ground, the SCL and SDA signals are normal, but when OE is pulled to high, the SCL and SDA signals are abnormal. (The EN signal in the picture below is LVDS_EN of the schematic)


       Please help me to solve this problem, thanks~

  • Hello Jack,

       The OE pin is controlled by hardware, there is no more I/O pin can be used to control OE signal.

       In the system side, the SCL and SDA are pulled up to 3V3 through 2KΩ resistor. In the i.MX8X internal, pullup resistor is enabled, the typical value is 50KΩ. Those pins are configured to INOUT, and the drive capacity is 2mA. Does the 2KΩ resistor need to be removed?


       In 1V8 side, the SDA and SCL are pulled up to 1V8 through 4.7KΩ. The matching chip is SN65DSI84-Q1. Does the 4.7KΩ need to be removed?


  • The TXS is a passive switch; its edge accelerators are active only for a short time when a rising edge is detected. At all other times, the lines are pulled up only with resistors. If you see a voltage going down, the only cause can be an external device actively pulling it down.

    To find the device that sinks current, determine the current direction by measuring the voltage drop over series resistors like R71/R73 (or NC4/NC5).

  • Hello,

    Please allow the OE pin to reach an input high state + enable time (ten) before sending signals through the device. The current capacitor on the OE pin can be reduced (perhaps 0.1uF) to reduce the RC time constant. 

    The TXS0104E has internal 10k pullups; the current external pullups (R21 and R28 on system side, and the 4.7kohms on DS184 side) needs to be removed. Also see Effects of pullup and pulldown resistors on TXS and TXB devices for more information for impacts on VOH/ VOL levels.