<font face="Arial, Verdana">That could be tricky, since min endstop is meant to be at z=0. There is a variable </font>
<font face="Arial, Verdana">#define ENDSTOP_Z_BACK_ON_HOME 0</font>
<font face="Arial, Verdana">where you could try a negative value so that it goes further down. Not sure if this messes up some other computations, so be carefull while testing. I guess z probe height must be 0 if you fix the difference that way.</font>