wow. you did shim it up well. Start with resetting your controller after you move the sled to the center of the board. Once the sled is centered, the reset command is (you can type in the black terminal window and press enter):
$RST=*
This will wipe all your stored memory on the controller.
After you reset, do not try to move z axis or the sled or it will just go typically up or down, but it won’t stop…
You then use the “reset chains” and zero the z axis. it doesn’t really matter at this point if the sled is perfectly aligned or the z axis is at zero, you just need to tell it that it is at zero so it doesn’t get confused. after you do that, you should be able to precisely move it and set the home / center position. You might consider not using the sled edge calibration. If you use the precision hole calibration and set the hole to 10 " in (or 254 mm) so the sled doesn’t have to 4x4 over the skirt edge to get to the spot, then it should be easier to run the calibration though you do have to wait for the holes to drill. 10" is far enough in that the edge of the sled should be 1" inside the skirt. Also if you have a spare end mill, you can put it in the hole that is cut and use it to measure from.
So Makermade sent me a new calibration guide which is fairly different than the process I have been using via 2Tankards. It’s simpler and the process for evening out the chains is better. It got me a much better result leading up to the Edge calibration- accuracy at 1.5mm.
However, when I do Edge I have the same result. The numbers are just insane : -376 in the top left corner, +304 in the lower right. And then on the top right it drove the sled off the frame again, albeit in a different spot than the previous attempts.
I guess what I am hearing is to skip the Edge step and go to precision, but the machine is just so, so far off. It can find center just fine, accuracy is 1.5mm. But it clearly has no idea where the edges are. Still safe to skip to Precision?
Thanks,
jason
I’m sorry, I’m at a loss. It seems related to measurements being wrong. Which you have triple checked. Maybe post screenshots of the measurements so someone with Makerverse expertise can evaluate just to get that second set off eyes.
Based on the measurements you sent above your running a standard 10’ frame approx 18" above workpiece.
it doesn’t seem like measurements to me because i am able to get the centering pretty good. so it seems like the frame is ok and the machine is able to move around correctly, i have nothing weird she trying to jog it.
it seems to me like somehow it is mangling it’s own math when it tries to do automated calibration movements
It sounds like your controller may be the issue. Not sure if it is memory corruption or a bad connection, but you need to contact makermade and see what they can do for you. This is not a new problem for them.
sounds about right to me. is it possible to wipe whatever is on the Arduino and update the firmware? or is that not the cause? i guess that wouldn’t touch corrupted memory
followinf guidance from makermade i deleted my workspace and completely started over from fresh, the idea I guess was that some major error had been made at some point and subsequent calibrations were built on that. doesn’t really make sense to me in light of the “wipe settings” button. but I tried it anyway. it did absolutely nothing to resolve my problem, and once again during edge calibration it drove the sled off the frame. no matter what I do, that happens. It isn’t the measurements I’ve made and I’m not making an error putting them in. it just doesn’t work right.
I am having a similar problem, I am sure the frame is square and plumb but when I get to the edge calibration and click the top middle move command using a 100mm offset the sled goes to -4mm instead of ≈100mm from edge. Then moving to top right I get 66,160, bottom right 70,156, bottom middle 100, bottom left 160,56, top left 154,77. This results in a calibration of ≈95mm which is caused (IMO) by the sled going way off scale on the top middle.
If I apply that crazy calibration then run edge again, the sled goes to the same exact locations. It seems the applied calibration has no effect on subsequent tries when it should have down something different.
I thought maybe the values were not getting set in the EEPROM but I can set and change values manually in MV or with a terminal.
I am trying to go thru the code to find out why it is slewing so far to top middle.
BTW, this was after executing a $RST=* to reset everything.