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.

TXS0102: Why rise time to VCCB voltage is so long ?

Part Number: TXS0102
Other Parts Discussed in Thread: SN74LVC2T45

Hello

I want to use TXS0102 to drive WS2812 led strips from an esp32 (and possibly reuse the port for 5V input).

ESP32 is 3.3V and WS2812, although it can work at 3.3V, signal should be at 5v.

I attached a TXS to a test bench, with 3.3V on VCCA and 5V to VCCB. Input is a 3.3v square signal from a function generator and output to scope.

Around 10KHz output signal look fine, but at 800KHz output voltage is too weak.

Why the rise time takes around 4µs ?

CH1 is input, CH2 is output

Thanks for your help

Nicolas

  • The edge accelerators are active for only about 30 ns, so they will work only for small capacitances. I'd guess the cable to the LED strip is too long for that.

    For unidirectional signals, better use a direction-controlled translator with proper output buffers, like the SN74LVC2T45.

  • Hi Nicolas,

    The TXS devices have no drive strength. They will not have the drive strength to drive an LED strip. The 5V runs through a 10K pullup resistor so the output will have some 50uA of drive strength, not enough to turn on LED's.

    The reason your output has such a slow rise time is because is your input rise time is too slow. The datasheet requires a 10ns/V or faster input rise time so that the output can follow it just as fast. There are one-shots inside the translator that pullup the device for a set period of time. Since your input didn't finish rising before the one shot turned off, the output finishes it rise in voltage with the 10K pullup. This is why in the datasheet you will see the datarate for open-drain use case is only 1Mhz.

    Thanks!

    -Karan

  • Thanks Clemens and Karan for your feedback

    Ok I see that my input is indeed too slow thus no edge accelerator. I get it now :)

    indeed I might abandon the multidirectional functionnality and go for something like you mentionned.

    thanks for the detailed explanation. My intent was not to actually power the leds from TXS (they have a separate 5V power input), but only the data line that command them and allows to turn them on individually. The TXS output should be enough, I'll test.

    Thanks again

    Nicolas