The Marlin firmware, by default, ONLY pays attention to the "endstop" switches when executing a homing maneuver. At all other times, the "endstop" switches are IGNORED. And there's only one "endstop" switch per axis, so even if the switches were not being ignored you could still exceed the physical limits of travel in the other direction and get the motor grinding like you say.
It is possible, if you really want to try it, to compile the firmware with the "endstop" switches enabled all the time - but it's a whole lot harder to configure a version of Marlin from the source code than to just download a pre-compiled hex file, and that still doesn't fix the problem on the other end of each axis.
So, back to basics: yes, there are settings you need to update; that's really pretty common and should be expected. If you are willing to work with us just a little more, issue an M503 command and copy and paste the entire list of parameters so we can see if something jumps out at us.
See my first post (before Mooselake's); I think I did answer your quest for info on the "soft limits;" I'm sorry if it wasn't clear enough that I was referring to the same thing you were referring to. Although I can see now that I should have pointed you to the first post in the chain about setting the software limits of travel:
viewtopic.php?f=15&t=10876 (All generations of Printrbots have used fundamentally very closely-related hardware and software so in most cases guidelines for the "Plus" apply equally to the "Simple" or the "Play.")