Delta "grinding" past endstops

[Credo]

I have no clue why this is happening, but when I home my delta, it grinds a bit before topping. I've ensure the right endstops are triggering the right towers (ie. X carriage is triggering the X endstop, Y to Y, Z to Z). I've verified M119 - everything's showing as triggered. The only thing I've changed are the stepper motors. I can no longer run any bed leveling, because the vibration from the gears grinding through the belt sets off the FSRs. the connections on the endstops are set to 'Normally Open'

Any help would be greatly appreciated.
thank you in advance.

[Aflac]

I had something similar, though mine turned out to be I had the wrong end stops for the towers. I had changed the example config file I was using and in the new file I didn't notice that the alpha and gamma towers were opposite of the original config file I was using.

[mhackney]

I went through a looooonnnggg period with Smoothie doing odd things like this and it turned out to be related the USB connection. If the CF card auto mounts when you connect smoothie via USB and then attempt to print, Mac OS X (and perhaps other OS) sometimes attempts to access the SD card during the print and that causes odd things to happen - and I experienced exactly the issue you have.

This was one of the big reasons I moved from smoothie to Duet (although I still run smoothie on a mini kossel). There is apparently no (straightforward) way to stop auto mounting CF cards on OS X so I was forced to remember to do this overtime I printed. It got to be a big pain in the rear having to remount in order to transfer gcode files (or remove the CF card to load them) and then remember to unmount, etc. Way too much headache for my production environment.

Try unmounting the CF card and printing a test.