I just got my Workbee CNC together and am starting my first projects / cuts with it. I am having an issue with the bottom of the cuts. Here is my spoil board. The lines you see were cut with a 1" mill and they are just that. Lines... you cannot feel them when you run your hand over them. So the spoiler board is Parallel to the X-axis. Here's my results on MDF. 1/8" End mill for the first cut and a 1/4" end mill for the clearing. Feed rate was 80 and plunge was 30. Any thoughts? Kevin
Looks as though you didn't get the Z zero quite right when you changed bits, and you may have some play in your Z and X gantries. What was your depth of cut when you surfaced your spoilboard? Starting at the bit wiggle things looking for anything moving that shouldn't be. Places to concentrate on are lock collars, bearings being seated correctly, eccentrics adjustment, but anything loose could be an issue. Alex.
Thanks Alex. The spoiler board cut depth is probably 2 - 3 mill... I used my touch probe between bit changes. I should have mentioned that before. But I'll check for wiggle and see what I can fine.
The lines on the spoilboard is indicative of a Tramming issue. Router isnt perfectly perpendicular to the bed
2~3 mm doc is a bit much for surfacing the spoilboard - I never do more than 0.5 mm/pass - you'll be throwing the spoilboard away before you get full use out of it . As well as a possible tramming issue that doc can cause deflection of the tool with similar results. Alex.
I guess I should clarify. I made my file to a cut depth of 1mm. I ran that file 3 times to get the entire spoil board surfaced. I'll do some looking into my tramming issues and see if that helps. Thanks Kevin
It took 3mm to get the high spots out? wow that's some major issues with levelness or you need more support (sag). Visible lines are unavoidable, if you can't feel them that's good. As for the clearing issues, I would take a close look at your z screw collars and your flex coupler. Make sure you have no play in the z- screw up and down, if you do adjust your collars, and make sure your coupler set screws are set on the motor flats and grind a small flat onto your screw for the other grub screw. Cheers Gary
avoid all tool changes while you are chasing any issues like tram/play/backlash. why introduce that obvious variable?
It was in the anchoring. I don't think I have the spoil board flat to the beam in one corner. I've tightened the bolt as much as I dare but I can see a small gap... Figured after I milled it, it wouldn't matter. There is no movement in the spoil board either so I think I'm safe. I know famous last words. I'll have to get out to the garage and feel the z-screw to see if there is any play. Another thing I thought of was my stepper motors were engaged when I check the z-axis before. So it would have had everything drawn up and tight. I'll check it again with the power off to see if there is any play. So stay tuned.
So yesterday I went back out to the garage and was just pushing on the z-axis motor to see if there was any play at all in the z-axis. I kept thinking there has to be something loose. When I would push against the z-axis in the Y direction there was a small amount of flex... I thought that should be okay it's minimal. But then I got to thinking minimal is my issue. So I went back out and did some more investigation. The screws on the end of the gantry that attaches the v beam to the Y-plate was half a turn loose. So I went through both sides and checked all the screws to be sure they were tight. Done... I then created a 4 inch circle toolpath with a 1/16 DOC. I used a 1/4" end mill on a piece of MDF. My Spindle Speed was 16,000ish and my Feed Speed was 100. My plunge rate was 30 and I changed my step over from 80% to 60%. Here is the result. Is that a fair enough test or does anyone have a better suggestion on how to see if my tightening those screws has solved my problem. That just seems way to easy... I'm not that lucky. I guess another good test would be to just cut my project again and see what it does or cut my circle again with a deeper DOC. Thoughts gentlemen? Kevin
A good test for backlash is, using the same centre point, two squares at 45° to each other and a circle of the same diameter as the sides of the squares. The circle should touch both squares exactly in the middle of each side. Alex.
The standard "Circle Diamond Square" tests, officially titled as "NAS 979; Composite cutting test 4.3.3.5" is a good one to use to verify everything Mark which way it came off the machine: using calipers you can check X, Y and Z dimensions (what you drew in CAD vs what you measure in the cut part), backlash in both straight and diagonal moves, roundness of the circle, etc
Yes, just post it with measurements indicated and NB direction it was on the spoilboard (so we know what was left, right, forward and back, or x-, x+, y-, y+)
Okay here we go. CRV file is attached. So is the PDF with my results. Let me know if you have any questions.
Also notice the z is all goofy too. On the bottom plain. I can provide those measurements to if you want Peter.
Please do include them, I haven't had time to dig in just yet, hoping @Alex Chambers and others may also chime in shortly Have you done the X, Y and Z calibration?
I have not. I had hoped to work on it this weekend so I can do it then. Or is it important to do that prior to the Circle Diamond Square test?
I understand it's the holidays guys so don't take time away from family to work on my stuff. Like I said I hope to work on it this weekend so if you can get to me Friday that would be awesome. If there is anything you need from me I will try to get it done Friday as well.
Peter, With the block oriented with the Y pointing up like in the picture the thicknesses are as follows. The thickness should be 0.250 for the square and 0.500 for the diamond Top Left of Square - 0.292 Top Middle of Diamond - 0.510 Top Right of Square - 0.290 Left Middle of Diamond - 0.502 Right Middle of Diamond - 0.502 Bottom Left of Square - 0.300 Bottom Middle of Diamond - 0.505 Bottom Right of Square - 0.296 If that doesn't make since let me know and I can create a graphic for it. Thanks guys!!!
Always, and feel free to iterate over it to get it as good as you can of course Do it a couple times - until you are sure calibration can't fix it any better, then either make peace with the physics of material deflection (How to calculate V-Slot® deflection) or if its worse than that, check what is "Not built as good as it can be" - as its a DIY kit - accuracy is also affected in how square you put it all together, how well adjusted everything is, etc For Z moves, remember the base of your stock is uneven, ideally deck off the spoilboard, and then use the machine to deck off the stock using the surfacing wizard too. Then measure pocket depths from the top (zero) surface - after calibration it all first though
Perfect Peter thanks. I'll get it dialed in on Friday and then cut a new Circle Diamond Square and get you the new measurements. Have a great Thanksgiving.
Kevin, it's not clear to me whether you have done this but... Remove the spindle (or router). Loosen the flexible coupling on the motor shaft so that the leadscrew shaft is totally free to move - be absolutely certain that the coupling is free! Put both hands under the bottom plate of the Z axis and push up firmly to make sure there is no upward (and down) movement in the Z axis leadscrew. If there is, take up the slack in one of the lock collars - I would adjust the bottom one. Then re-attach the flex coupling. The flexible coupling should never take the weight of the Z axis.
Okay I checked that Christian and the z lead screw is solid. No play whatsoever. Next to calibration.
Spent the afternoon putzing with the calibration and I can't get this thing dialed in to save my life. So frustrated right now.
Is the problem consistency? Does it return to the same height every time? If not you have a motor connection issue or your anti-backlash nut is too tight, wheel binding, or the z nut isn't in line with the z screw bearings. Also, If using black box turn your current up. Cheers Gary
