Hello,
I designed uTCA shelf board with two TPS2359 power distribution controllers on it.
Looks strange, but I fail to communicate with them via i2c: no acnowledge from any of the two TPS2359 connected.
I've tried several hardwired addresses, than wrote an i2c bus scanner - nothing helped.
I use MSP430 i2c hardware programmed to 100kHz SCL clock rate. It works OK with other devices.
What could be a reason for TPS2359 i2c to not to reply/acknowledge?
I looked through all the pins: nothing suspicious.
Could a thermopad be a reason? There is no wording in the datasheet what to do with it. I left it isolated.
Unfortunately it is much more difficult for me to try connect it now (the pad is under soldered QFN36) than to ask for your help here.
Any comments / ideas?
Thank you.
We can't go. We R waitin' 4 Godot.
There are several things you can check. First, the SDA line has a weak pull-up internal to the TPS2359. This is to prevent the pin from oscillating if the I2C is disconnected. The internal pull-up is not strong enough for operation though so an external 2.2kOhm pull up is required from SDA to 3.3V. Also, the TPS2359 can only operate in slave mode so the MSP430 should always be providing SCL. Also, each TPS2359 must be wired to have its own unique I2C address using the A0, A1 and A2 pins. You should also look at the SDA and SCL lines with an oscilloscope to verify I2C timing and that the SDA and SCL are reaching the appropriate logic levels when attempting to communicate.
Typically, the thermal pad is connected to ground to provide heat sinking. Leaving it floating will not cause any problems for the communications. Be sure to check the IC temperature after the device is turned on and providing full power to the outputs though.
oops,
there were SDA and SCL swapped.
too many wires.
shame on me, thanks for the help.
if not this forum, I'd probably desoldered TPS2359 and danced around that thrmopad. :-)