It looks like your middle assembly is loose, somehow. The uneven milled surface on the bottom of that coaster seems to back that up and the gcode is circular. A milled surface should be pretty amazing. Being able to see such distinct lines says either you are using too high of a step over, your tool is not perpendicular to the work surface, or something is loose.
Extend your spindle down to about an inch from your spoil board. Put your finger on the end of the bit and wiggle it. If it wiggle noticeably you are lacking proper rigidity. It looks like you are only loose on one axis somehow but the video is so zoomed in I can’t tell what axis exactly but it will be the one that is perpendicular to the flat spot.
There was a little play on one axis so I tighten the tension bolt and the play is gone. But then it seems as if it is not square and fighting itself. I did check the z and there is a little play when the axis is don about 5 inch but anything above 3 and the there is no play. So that is causing my problem now as I have the work piece raised.
What would be the best way to correct this?
I was thinking that if I loosen all tension bolts with the assembly on the machine should be able to use the motors to help keep the assembly square while I retighten the tension bolts. Also should all tension bolts including the 5", get snugged up after teverything is square?
Agree shorter is better, I am going to shorten it. But that isn’t causing my problem at the moment only when I want to cut at full z reach. I’m running these at at less than 2 inch and there is no flea on the bit there. As for the middle, I did follow the instructions when I rebuilt 2 days ago and that seems to have caused my problem I’m having now. Squared it with the z assembly in place. Looks like I have a middle assembly rebuild in my future. Thanks for all the help. I’m gone for 9 days now, so will be done as soon as I’m home.
Ok, so rebuild the middle assembly following the instructions very closely, I must have checked and rechecked it 20 times or more. Reassembled it o t he frame and ran the coaster gcode again. Ended up with the extact same results. Ran it extremely slow at less than 1 inch z height, there was no deflection in the bit when pressed on. I’m thinking that maybe the overall size of the machine is causing my problems. What size of a machine do you think I could run my router with, without troubles?
It could be anything. Improperly set drivers, maybe bad gcode, maybe its too big, router runnout, z length? You kind of changed too much to narrow anything down.
What steps do you think I could take to completely eliminate something or anything. Kind of not sure what else i could do?
-Have set the drivers based on my motors, also tried a different set of drivers, and in other stepping modes.
-I uploaded the gcode and you assured it was fine and circular.
-I have cut squares and diamonds and measured diagonals and sides and all are within spec of what I should expect, to eliminate router runout.
-I ran the z height and less than a inch to eliminate deflection as much as possible.
Any thoughts would be greatly appreciated, and thanks for all the patience and help you have provided me with.
Just for giggles I’m cutting the coaster on my router right now. So far from the 45 seconds it’s been cutting I can tell your stepover is way too big. Looks like it’s almost 80%? Though I’m guessing on the bit diameter, so I’m using a 1/4" 2 flute end mill.
Measurements came out 88.7 and 88.5 at 90deg from each measurement.
sidewalls are 9.5 and 9.7 one way and 90deg I get 9.5 and 9.7 again.
I was getting some stuttering from this gcode, not sure if it was the code, or my router is just cold, it’s about 50 in my barn tonight. I don’t think that’s too far off for the shitty plywood I’m using. Also not the sharpest end mill. This is one I kind of burned up running too slow of a feed rate.
Pic
Tape a sharpie to your router and draw a circle. I’m wondering if it’s just the mill is too dull and it’s pulling the router off axis.
Hey Barry thanks for testing the code it certainly looks circular which beats mine. I know the stopover is a bit much, never noticed the setting on estlcam until after. Will try the sharpie out like suggested, what size of a machine are you using.
Ok. That’s the same size as my build. Just wondering what your setup is, I am using conduit with mid-supports and all parts are 525. I’m using ramps with tb6600 using 1/16 stepping.
I changed my motors from parallel to series and ran them at 1.22a at 13.8v. They call for 2.5 amp at 3.1v. So what should I run them at in series and parallel? I did notice that in series they have way better holding torque.
Also, I’ve tried drawing the 2 coaster circles. Got the same results, circles aren’t circles and start and stop position is shifted. Losing steps? Best way to diagnose missed steps?
So, also I’m running my ramps with all jumpers removed therefore putting them in full step but changing that to 1/16 on my tb6600 drivers. Could this be causing a problem, possibly with the firmware and my driver setup? Using the new version by vicious1, rc7. I only have a4988 drivers, is it possible to run my motors off these for testing purposes?, in series or parallel?
Sorry for all the questions, any help or thoughts would be great!
If you are running in full step mode, you are probably cutting the circles off. think of drawing a circle with only straight lines top bottom left and right will be flat. You need to run in at least 8th stepping. You have effectively lowered the resolution of your machine.
If you have a ramps board, why not try the regular drivers so we can eliminate a potential issue. Those drivers are unknown to me and we could be wasting a lot of time with them. Why don’t we get it working as recommended then you can change things?
I can’t give you an amp setting I have no Idea what your drivers can handle.
Max in series would be 2.5A, max in parallel would be 5A.
Ok, got a loan of a buddies drv8825 drivers and tried the coasters again, with the same results. So that ruled out my electronics. So then, it hit me, backlash. I was getting the flats on the sides when motion reversed. So I slowed the code down and then I could see that when there was a change in direction the middle assembly was lagging the outside rollers. So backlash, I’ve since hooked up my old electronics to use linuxcnc to add backlash compensation and I am now getting circles. This is just a temporary fix, I’m going to have to reduce the tension in the middle assembly or rebuild with stainless on a machine this size.
We use belts just for that reason, there is basically no backlash with belts. At least not enough to worry about.
So either your pulleys are loose and slipping in the d slot or you belts are to loose or too tight. You haven’t put up a full a full machine picture to have a look.
No it’s not a belt issue. I can see the conduit flex at the middle assembly in a change of direction. My center conduit is 5’ long in x and y. The middle assembly tension is greater than flex strength of the conduit. So I have to adjust the middle assembly tension or use a stronger material, stainless instead for conduit.
Ok, update. I rebuild the middle assembly again paying great attention to the bearing and tension. Made sure every bearing just made contact with conduit. It wasn’t easy while trying to then square up the center. Got square to about 1/8 over 5 feet. With a little more tinkering i should be able to get it a little better. Once reinstalled, I’m now getting circles that look like circles. On a build this size I could really see where stainless would other much more benefits to rigidity and smoothness over distance.
All in all, it is a great design and a great community that forum offers.
