TCA4311A: what symptoms occur unread

Please show an oscilloscope trace of a failed read.

Gareth,

As Clemens said, we need a bit more information here. An image of the communication waveform would help, a schematic of some kind would also help.

Regards,

Eric Hackett 

Hi

The situation that customers encounter is that sometimes they can read the information, and sometimes they cannot. A number or FF will appear normally, and XX will appear when it cannot be read.

Thanks

Gareth,

Are you able to get scope captures of SDA/SCL when the failures occur?

-Bobby

Hi Bobby

Provide Waveform as follows

(fail)

(pass)

This might be the problem described in section 8.4.3 of the datasheet. Use weaker pull-up resistors, and/or add capacitance to the bus lines.

Hi Gareth,

As Clemen's pointed out, this seems like the rise time accelerators turning on and experiencing contention, I suspect this is occuring during an ACK. The VoL seen on the line also seem a little high, this could make the rise time accelerators trigger earlier since you're closer to their threshold for when they turn on.

Changing the pull up resistors/loading can help, usually making them weaker to slow down the rise time accelerators turn on.

If you're early in your devlopement, I would suggest moving to TCA9511A since it's a functional replacement to TCA4311A with improvements. The only downside to TCA9511A is the ESD rating is lower than that for TCA4311A.

-Bobby