Re: Shark HD5 gantry\router not square to board
Posted: Sun Sep 29, 2019 9:32 am
There are several issues to deal with in getting these machines to work correctly. One that has not yet been mentioned yet is speed of cut and depth of cut. Basically, it relates to how aggressively you are pushing this machine. The standard values from Vectric software are for more solid machines and can be too aggressive in speed and depth for a Shark.
The easy test is slow down all your cuts and see if some of the defects go away. Something that will run fine at a speed of 20 may look horrible at a speed of 80. It does not help when manufactures quote rapid travel speeds like 200 when the machine can only cut reliably at 40. etc.. So first thing to do is slow down.. and see if things get better.
I am not sure what kind of lead screw nuts that Next Wave is currently using on the machines. These are the nuts that the lead screw turns to make the axes move. In past days they were solid plastic with no anti-backlash springs. If they are using Anti-backlash nuts on the lead screws this can be an issue as well. In order to be "anti-backlash", there are usually two sections of the nut that are pressed together by a spring. That forces the nut to be snug on the lead screw at light loads. However, when you are at higher cutting forces, the spring in the lead screw can be overcome by the cutting force, and it gets pushed away from being tight on the lead screw. This is hard to diagnose as it only happens when the machine is cutting with too much force. So, Anti-backlash only works at cutting forces lighter than the spring force on the nuts. At high cutting forces, the anti-backlash nuts can be worse than the plain solid ones for holding tolerances. So again, slow down and take light cuts. If those cuts are fine, it is a force related issue with machine flexing, or overcoming the anti-backlash nut tension.
Aligning the X and Y axes so that they are 90 degrees apart may need to be done by adjusting the coupling points of the bar that passes under the bed and connects the gantry sides together. If you cut a test square, it should then give you a square with true 90 degree corners.
"Depth of cut across the bed" issues arise form the table not being the same distance to the gantry at all 4 corners. Many folks have made their tables adjustable to get the table equidistant from the gantry, and then lock it down when adjusted. Planing a spoil board also does the same thing. What you are trying to get is a table surface that is equidistant from the router bit at all places.
If you have a flat work piece, you should not have to use virtual zero techniques, If the workpiece is not flat, and you are carving shallow fine cuts with V bits, it pays to flatten the surface the workpiece first, as you would plane a spoil-board
The router should be perpendicular to the leveled table. The router mount may need to be shimmed to make that happen. Symptoms of the router not being square to the machine are the ridges you may see when planing the spoil-board.
There is a lot of good info in older postings on this site, and the search tool can be very valuable.
The easy test is slow down all your cuts and see if some of the defects go away. Something that will run fine at a speed of 20 may look horrible at a speed of 80. It does not help when manufactures quote rapid travel speeds like 200 when the machine can only cut reliably at 40. etc.. So first thing to do is slow down.. and see if things get better.
I am not sure what kind of lead screw nuts that Next Wave is currently using on the machines. These are the nuts that the lead screw turns to make the axes move. In past days they were solid plastic with no anti-backlash springs. If they are using Anti-backlash nuts on the lead screws this can be an issue as well. In order to be "anti-backlash", there are usually two sections of the nut that are pressed together by a spring. That forces the nut to be snug on the lead screw at light loads. However, when you are at higher cutting forces, the spring in the lead screw can be overcome by the cutting force, and it gets pushed away from being tight on the lead screw. This is hard to diagnose as it only happens when the machine is cutting with too much force. So, Anti-backlash only works at cutting forces lighter than the spring force on the nuts. At high cutting forces, the anti-backlash nuts can be worse than the plain solid ones for holding tolerances. So again, slow down and take light cuts. If those cuts are fine, it is a force related issue with machine flexing, or overcoming the anti-backlash nut tension.
Aligning the X and Y axes so that they are 90 degrees apart may need to be done by adjusting the coupling points of the bar that passes under the bed and connects the gantry sides together. If you cut a test square, it should then give you a square with true 90 degree corners.
"Depth of cut across the bed" issues arise form the table not being the same distance to the gantry at all 4 corners. Many folks have made their tables adjustable to get the table equidistant from the gantry, and then lock it down when adjusted. Planing a spoil board also does the same thing. What you are trying to get is a table surface that is equidistant from the router bit at all places.
If you have a flat work piece, you should not have to use virtual zero techniques, If the workpiece is not flat, and you are carving shallow fine cuts with V bits, it pays to flatten the surface the workpiece first, as you would plane a spoil-board
The router should be perpendicular to the leveled table. The router mount may need to be shimmed to make that happen. Symptoms of the router not being square to the machine are the ridges you may see when planing the spoil-board.
There is a lot of good info in older postings on this site, and the search tool can be very valuable.