TPS552882-Q1: TPS552882-Q1

Hi,

I'm using this part as a buck boost where my input voltage is 10V to 30V and my output is 18V, 6A.  My schematics are below:

My input is 12V from a benchtop power supply and is very clean.  This is eventually going to drive additional power supplies and eventually a comExpress module.  When I used the comexpress module, the 18V dips down to 6.2V when the comexpress starts to boot and draw current and then my system resets and goes into this continuous hiccough mode. 

To simplify debug, I took off the comExpress module and simply put a constant current electronic load on the supply.  With very little load on the supply, I scope the boost SW2 signal and it is switching at the appropriate frequency.  SW1, the buck side is not switching.  Everything seems normal.

When I begin to increase the load current to about 1.6A,  then device begins to fall out of regulation as shown:

The yellow trace is the 18V output which beings to dip and the white trace is the input current and goes to almost 18A!

Below is the Boost SW2 node.

And below is SW1 which is the buck node.

If I slowly reduce the electronic load, everything begins to work well again.  Thinking that maybe the ILIM resistor was populated with an incorrect value, I removed it from the board and measured it. It was 20K.  I tried a 10K but the behavior is the same.  

I have also tried the follow:

* removed R9 which basically pulls up the UVLO to the 12V input. 

* I was able to sneak in a 24.9K from MODE pin to ground and that did nothing.

All the values were gotten from Power designed.  The inductor is rated for 25A and has a DCR of 5.11mOhm so this should be more than enough for this system.

The layout followed the datasheet recommendations so I'm pretty confident that it is good.

Why is this device behaving so poorly?

Thanks!

John