TXB0108RGYR

If VCCA is 1.8V, VCCB 3.3V, can nOE be controlled by a 3.3V I/O pin, or will this destroy the chip?

I'm seeing some strange behaviour on the TXB0108RGYR.  Setup:
 
1.  GND (pin 11) and PAD are connected to GND.
 
2.  VCCA is 1.8V.
 
3.  VCCB is 3.3V.
 
4.  A 3.3V output is connected to B1 (pin 20 - CH1).  A 1.8V input is connected to A1 (pin 1 - CH2).

As you can see, the 1.8V output is not clean.  The enable is at 3.3V for the entire pulse.

For anyone interested:  the TXB does not actively hold outputs at 0 or 1.  Only the edges are accelerated through one-shots.  The input above has a pullup, and it contends with the TXB's pulldown at logic level 0.  The TSX0108E has a Npass FET that connects the input to the output at logic level 0, and provides a much cleaner output (low-output is around 60mV).

Hello Inderjit,

I'm sorry I was not able to reply to your earlier post sooner, but it looks like you were able to come to the recommended solution.

For applications that require pull-up resistors - the TXS0108E is the best choice.

It is not recommended that you use TXB translators with pull-up or pull-down resistors, unless they are at least 50kOhm.

The TXB device does have a weak drive (4kOhm) that will hold the output at 0 or 1. However, if you have a pull-up, (in this case it looks like ~15kOhm?), it will contend with the internal TXB buffer, and exhibit the behavior you show in your scope capture.

I've included a link below to an applications note that details using TXB translators.

http://focus.ti.com/analog/docs/litabsmultiplefilelist.tsp?literatureNumber=scea043&docCategoryId=1&familyId=707