buck simulation result and actual testing result??

[batistuta]

hi,all
now, i am busy on testing my first power manage chip, during the testing, i puzzled by the test result.

for example, my current mode buck converter, when i simulate it under spectre, it can load 0mA-- 500mA and the output voltage is not change, almostly stable at 3.3v, but when i test the buck on the PCB board, my buck can not load larger than 300mA, it means when iout>300mA, my buck's output voltage not stable, changes a lot...
i don't know why?? why simulation result and testing result not match?
can you give me your ideas??
thanks.

 

[DenisMark]

Do you aware about switching noise/substrate coupling during design? Do you apply any techniques to decrease both?

The most probable reason is that your analog blocks (e.g. banfgap) are affected to noise of switches. If you have testpoint on bandgap output you can try to introduce some capacitance by probe. It may increase stability if the reason in switching noise.