I tried that opposite movement until it bound and I didn’t want to force it. I’ll try again. Thanks for responding so quickly!
You only reprinted the one side right? If it was both id say push it and even if it breaks who cares… But in your case you only have one replacement. And you don’t need another stinking voron misprinted part LOL
Yes, just one side. I think the abs will be difficult to break.
Looking at a couple other threads, I have another idiot question:
How do you both keep the belts barely tensioned and use the belt to true the x axis - without over tensioning?
LOL that’s where my printer is struggling right now. And when I get done with all these ASA parts I’m going to do what I have to so I can figure out why my x rail doesn’t sit square naturally. I believe I have a skewed Y rail, or both. So before I take it down I will print 2 new ones just to be sure. Hopefully that’s what it is. If not then I have no clue. Talking with Ryan we have gone though and looked at several things and they all are good. Printer is dead square. So there is no reason it shouldn’t sit square on its own. Once I figure that out I will post up the results to let everyone know. And then hopefully with the belts evenly tensioned ill be able to turn this thing up some more and see what it will really do.
When you put your x beam in did it fit really tight? did you have to almost force it in there?
Yes, the x rail fits tight. With the screws that normally hold it removed, it is still in place and doesn’t want to come out. It can be racked about 15 degrees in either direction and it stays in it’s spot.
I think I assembled it on the y rails all as one piece onto the extrusion… Maybe I had to get the extrusion to spread a little… I put a top brace on the front opening. Maybe that just need to be released
I haven’t figured out how to make X rail physically True Perpendicular to the Y rails. But I don’t care…
Am happily using Klipper’s software/config XY skew compensation settings to make up for my poor physical build quality. Did some test calibration prints using a super thin 100mm x 100mm Crown model. Larger model would be better, but the diagonal of ~141.42mm is close to the size limit of what my calipers can measure.
Ideally, software should help trivialize obtaining dimensionally accurate parts despite a crappy initial assembly, and, also help detect degraded/anomalous vibration behavior resulting from loosening grub screw, or, frame/mount screw. Proactively raising actionable maintenance alert(s)…
My skewed X rail maybe in part due to my pulley orientations being wrong. Noticed in Oct but still haven’t fixed, because the printer was printing fine, at least I thought so until I tried printing close to @Lithium366’s accelerations/speeds. Will update if fixing my pulley orientation helps reduce physical X rail skew. First though, before anymore variables/changes, am trying out a new part cooling duct design…
1 Like
But your x isn’t off by like 5 degrees is it?
5 degrees !?!
My X rail is ~0.805 degrees off from where I expected. Not 100% if my expectations are correct though, so included pics to clarify what I’m measuring. Note this is with the Right pulley incorrectly orientated.
|
Left |
Right |
move the rail all the way to the back and measure from there. The point you are measuring from looks like the motors and those move as you tension the belts. If you have one belt just a touch longer than the other it will make a difference. If you go all the way to the back against the idler mounts it should be touching on both sides equally
2 Likes
I wouldn’t measure from the motor unless they are both in the exact same location. With the ability to tension, that can change. I measured from a cross beam I installed on the front and from the back beam. Need to run the cal test to see just how far off it is, but now that it is in pieces, it will be different.
5 degrees was in response to you saying you just let the software fix it…
Got the new left part on. Turns out the left top extrusion was rotated CCW just enough I could tell… Strange. I remember squaring everything. And the rail was being squeezed by the frame. After loosening the front extrusion just a tiny bit, the frame is now just a touch wider up top and with that. The x rail popped right out.
After reinstall, the X rail is within a 16th" WRT the front extrusion. It should work.
Pulley isn’t leaning, but I’m wondering if it wasn’t that twisted extrusion and the pulley bolt squeezing that oat together. I discovered a missing bolt that explains how the y rails twisted. The nut on the back corner was there but no bolt. The new part went on without issue. Lubed up the y rails and x rail, then tried to reprint a corner bracket and I probably changed too many things:
Because the next print turned out less than ideal. The holes are still squished in z and there is some ringing and corner artifacts around these holes. Changes to pressure advance and square corner velocity maybe need to be made. I loosened the belts and made sure it was still square, so perhaps some skew correction is needed. There will be some more adjustments and a few cubes printed hopefully before the petg runs out and it has to switch to another color of pla.
Edit to include some concrete numbers of where things were for posterity and what if anything helped:
- stealthchop 999999 (found that somewhere, but don’t recall. changed to this after it was a 250, but not sure where I got that either… this doesn’t help with noise reduction)
- pressure_advance: 0.6 # likely too high while others default to 0 or 0.0x (I’m seeing online)
- square_corner_velocity: 4.5 # default is 5. only thing to be found on that is “try something. default is 5, most never change it”
- tmc driver current settings for L,R = 0.7-1.0 Z0,Z1,Z2 = 0.5-0.8, Extruder = 0.5-0.7. stopped disabling drivers. looking for macro code to survive reboot without dropping bed. looks like possible, but not sure the syntax yet.
Yep. Been there several times. It can go either way. Change too many things and it goes like that. Or change and it goes great but you don’t know why lol. I do my best to not make more than one or 2 changes at a time so I can try to keep track. I’ve also been making sure to save all the testing under the name and color filament I am using now that I’ve switched to Orca slicer. Seems to be working well so far lol. Ryan got me hooked up today on that horrible noise mine was making. This whole time I thought it was the rails, nope it was stealthchop setting. It was there from the example config I started with set at 99999. Changed it to 0 and its SO MUCH BETTER!!!
- stealthchop 0
- pressure_advance: 0.019
- square_corner_velocity: 5
- tmc driver current settings for L,R =1 Z0,Z1,Z2 = 0.7, Extruder = 0.7
- print height: 0.34
- first layer height: 0.3
- nozzle: 0.6
- print speed 60 mm/s
- walls: 3
- infill: 5%
- extrusion width 0.65 mm
- black PETG 240 °C bed 60 °C
- first layer 230 °C
- fan disabled for first 2 layers
- Printer acceleration max is 10k. Max speed 1k, max z is 400, max z accel is 1k.
Cube 1: 20 min print just to see.
top is a bit rough and cant read the small inset
Dimesions in slicer → 25.4 mm on all sides
measured:
X: 25.49
Y: 25.65
Z: 25.89
This cube I searched up klipper config on this forum and basically “reset to stock” to get a baseline. That is where the 10k accel came from.
The worst side of the cube is the back corner at the top shown here:
I’d really like to get print speeds up around 150, but the first layer is the key here and it is pretty terrible:
I’m going to run a temp tower with these settings and then print the accel test. Not sure what to do about the top layer smoothness, but the sidewall waviness appears to be gone, but a longer run will be required to see if that is just from the scale and short sections.
10k acceleration!! WOW. Did you run the Ellis tuning to make sure it can hold that for 50 iterations?
Cube looks good! Little things here and there but at that layer thickness you wont ever get it real smooth I don’t think
Still kind of throwing darts. If not for other settings for examples, it would still be printing blobs of goo.
Take some time and go though this guide. You’ll get your printer dialed in to what it can do rather than just trying to copy what someone else can do. Seems like all of us are capable of different accelerations, flow rates and all kinds of stuff.
How would the rest of us idiots learn anything that way?
maybe just a whole new forum section then…