The joint where the 6" wide plate bolts on to the back of the gantry would have to be perfectly square for the gantry to also be square. The distance is so great that the angle on those joints doesn't have to be out by much at all to get 1.5mm out. Also I'd be amazed if the frame is actually perfectly square, so using it as a reference could be misleading. The other thing is the position of the X bearing mount plates and hole positions - they could be slightly out too. All these tolerances stack up, so it's hardly surprising to me that there's a 1.5mm error ... don't worry about it.

If in the end the machine still wont cut square there's ways round it. If it's not much at all then you can position the X-home switches so that mach3 bends the gantry very slightly to square it when it homes. Clearly that's a bad plan if it's out by a lot as it would put a lot of strain on the structure and motors. You can also compensate in mach3 by using 'formulas' which enable you to make one axis a function of another axis. So you just put X=mY in the X formula box where m is the measured gradient / error. The difficult bit is measuring the error to start with as there's no straightforward reference.

Quote Originally Posted by Web Goblin View Post
How do you manage to synchronize the bosch drills:heehee:
Maybe he's using the steppers as encoders to provide feedback...