Hello all, After scouring the forums for the past few weeks and tearing down my machine a few times I'm at a bit of a loss. I have a workbee cnc style machine running on the openbuilds blackbox controller with openbuilds nema 23 high torque stepper motors on all axis. The machine also has Lead screws and linear rails on all axis instead of wheels/belts/ etc. The issue I'm having is that even if I square the gantry to the end plates of each y axis, power the machine on, confirm gantry is still square to end plates, anytime i make a y axis movement the machine seems to work itself out of square. I've confirmed over and over again that the machine frame is square, I have replace the linear bearings on the side that is "lagging", I've tried switching the y axis and y2 axis cables between each side, switched stepper motors between each side, and no matter what I do the problem reoccurs. When i tore the gantry apart to replace the Y axis linear rail bearings I ran some y axis movements before rebuilding the x axis gantry and confirmed that the issue still presents itself even when the y axis and slave axis are moving independent of each other and not connected by the x axis. Typically over a 36" y axis movement I will end up with at least an 1/8" difference between the end plate and gantry plate on the y axis and slaved y axis. Any ideas for next steps trouble shooting wise / what the culprit could be?
If you then go back to your start position do the Y gantries line up again? - one possibility is that your two Y leadscrews do not match in pitch. Try swapping the leadscrews and see if the problem switches to the other side. Alex.
Hey Alex, Thanks for the response! They y gantries do in fact line up again when returning to the start position. The pitch of the y leadscrews not matching is what I'm thinking at this point has to be the culprit. For some reason i was convinced i had already swapped them and that it wasn't a pitch issue but am now starting to think that isn't the case. Going to swap leadscrews now and will report back. Zach
Just finished swapping the lead screws out and sure enough, the problem switched sides with them. Will be ordering a couple tonight from the parts store. Zach
Zach - I'm curious if this fixed your problem? I have been having the same issue. Just in case others run into this problem, I bought one of the KNOCKOFF 1500mm x 1500mm Workbee hardware kits since I wanted the larger format machine. I didn't notice the issue until I started working on a project to cut a mold for a fiberglass part that was roughly 38" square. That's when I realized the Y1 and Y2 were out by 4mm over about a 1200mm distance. Probably not a big deal for most projects, but for that project it did since the fiberglass parts were supposed to mate up. I was fairly confident that it wasn't losing steps since the gantry was coming back to square when I moved it back to the front of the machine. I made sure there was nothing mechanical going on with the machine, so I was pretty sure that wasn't the issue. I had checked my calibration distances against X and the Y1 side of the machine and their calibration values were pretty close, so I thought swapping components between the X and Y2 might work. I started with swapping the X stepper to Y2, but that had no effect. I then swapped the lead screw from X to Y2 and that helped, but didn't eliminate the issue entirely (it went from 4mm out over 1200mm to 2mm out over 1200mm). Being new to CNC I didn't realize some of the subtleties of GRBL and a slaved Y2 axis. If it supported separate Y2 calibration, problem solved (I hope they are able to add that at some point). The only other option I could see was to switch to a different controller, or find more accurate lead screws. Open Builds lead screws are ±0.1mm per turn (which is pretty good after looking around), but over a roughly 1200mm working area that could still be a pretty big difference between Y1 and Y2. I did a lot of looking and found Helix Linear Technologies (Tr 8x8 RH SS Trapezoidal Screw - Trapezoidal Screws | Acme Screw, Lead Screws, linear actuators, hybrid stepper actuators | Helix Linear Technologies) out of Ohio. When I reached out to them they said their trapezoidal screws are ±0.0003 in/in which looks really good. It also sounded like they might be able to check and sort to ensure a specific accuracy along the entire length, but I'm not sure what that would cost. Of course, their 1800mm lead screw is $131 vice the Open Builds 1500mm lead screw at $50, but I guess you pay for accuracy. I still haven't decided what to do (or what's good enough) but I'd be curious on what you decided, or if anyone has any better solutions to the problem. Greg
Consider replacing the leadscrews with Original components (openbuildspartstore.com) - inaccurately manufactured leadscrews are cheaper but impossible to correct in software
I second @Peter Van Der Walt 's suggestion to replace the leadscrews with better quality ones - some of the 32 bit controllers can calibrate steps/mm for Y1&Y2 separately, but that would be more expensive and in the long run a lot more work. Alex.
Plus that doesn't solve badly rolled inconsistencies along the length (correct in couple centimetres, then quite off for a bit, then under for a bit, then right... Etc. Average is a value you calibrate, but still will not be good
I'm wondering about your statements wrt the Open Builds lead screws. If they're ±0.1mm per turn and the two screws are different within that tolerance range, there's still a potential for the same problem, isn't there? The only thing I could think of to solve the problem was either very accurate screws (to minimize the error), or matched screws (to allow calibrating it out with a single calibration value). So far I haven't seen anyone that sells matched screws, so I was leaning toward the more accurate screws. Am I thinking about this incorrectly?
