X axis wobble along Y during X motion (and vice versa)
Hey guys, I just completed my V-Core 3.1 and noticed a motion system issue. here are the summary of the problem:
Issue observation:
1) X axis wobble along Y during X axis movement (same for Y axis): https://drive.google.com/file/d/1KBno_NA2EgNNHVjQ_mW-qBEWDhvndrYf/view?usp=share_link
2) Printed parts exhibit “wavy” surfaces along X and Y axis (they are supposed to be straight)
3) X wobble is more pronounced when toolhead is nearer to either ends of the x gantry
4) Wave surface is very consistent and appears on the same area of re-printed parts.
Troubleshooting steps that I've done so far:
1) Double checked upper and lower belts are exactly the same length, mounted as symmetrically as possible
2) Double checked Y linear rails to ensure no binding is observed over entire range of motion
3) Re-tensioned belts, using Voron 2.4 method to measure frequency to ensure equal tension.
4) Verified that the motor pulleys are positioned such that the belt runs in the middle of the toothed section with visible gap on the top and bottom
5) Checked pulley and idler rotation smoothness, noticed upper belt has significant more friction: caused by deformed top motor bracket printed part due to excessive motor screw tightening: reprinted and solved, but unfortunately the issue still persists.
6) Check for motor pulley off-center wobble: removed both motors and run a print for direct observation, https://drive.google.com/file/d/1zzXlGQ1N5aefQ2jg43cFf9m_aX-4KLHk/view?usp=share_link. I don't see any excessive wobble, so i cant be sure about this.
7) Turned off input shaper and pressure advance: same issue
8) Gradually reduce belt tension to bare minimum : same issue
9) Replaced V3.1 x axis front facing idlers with 3.0 toothed idlers: same issue
10) Swapped XY motor with Z axis motors (to verify that its not a motor issue): same issue,
At this point, Im pretty much out of idea and wondering if it is a firmware related issue? any help is very much appreciated!
Google Docs
IMG_0041.MOV
30 Replies
national-goldOP•3y ago
Issue observation -
1: I have no idea what the video is supposed to be showing, maybe take another and explain what you are seeing as you film?
2: have you verified that the X axis is at 90⁰ to the Y axis? Sometimes people over tighten one side and end up racking the X axis.
3: have you checked to see if your wiring from the tool head to the electronics panel isn't causing strain and pulling on your tool head?
Troubleshooting -
3: don't use frequency apps to adjust belt tension. You want it just tight enough that it doesn't slap the frame. Then use your adxl to generate graphs and get the X and Y lines to match a d have well defined peak.
national-goldOP•3y ago
Sorry for not clarifying, the first video where the toolhead moves is supposed to show the X axis wobble along Y. If you look at the X joiner printed part or the MGN12 block and refer to the green line that i added, you should be able to see the movement along Y axis,
The second video where the motors are just rotating is trying to capture the little wobble of the pulleys, Im not sure if such degree of wobble is sufficient to cause the issue.
Yes Ive checked the wiring and there is plenty of slack. Noted on the belt tension adjustment, i'll give it a try and see how it goes, thanks again!
I still think your x axis is racked. So looking from the side, instead of it being a like a sideways H (x axis being the horizontal line, y being the two vertical lines that form a H), the X axis is slanted and it looks more like a Z (obviously not that slanted or you would notice)
national-goldOP•3y ago
I've already double checked the gantry racking, and measured it using a bevel protractor and its pretty much spot on 90 deg. printed part also measures pretty close to 90deg (see attached image) so i don't think its a racking issue.
If you move the X axis gantry all the way to the front of the machine is there a gap on either side (i.e. does one side hit something before the other?)
national-goldOP•3y ago
nope, both printed parts hit the front bracket plates simultaneously
Ok, then I think we can rule that out
national-goldOP•3y ago
additonal information, looks like someone else had this issue and it seems that his conclusion is in the pulley tolerances: https://drmrehorst.blogspot.com/2020/12/x-axis-wobble-in-ummd.html
However, if thats the case, I believe most other users should be facing the same issue since the pulleys are all supplied by Ratrig (I assume that they are from the same manufacturer)
X Axis Wobble in UMMD
Testing for X axis wobble in UMMD.
If you move the X axis by hand does everything move smoothly?
national-goldOP•3y ago
yeah it does, except for the motor resistance, things are pretty smooth, when i push the X axis along Y, the printhead remain stationary along X and vice versa
Well, I suppose you try swapping one of the Z motors out for the one that is having resistance and see if that fixes it?
national-goldOP•3y ago
sorry when i meant motor resistance, i was referring to the typical inductive force that you feel when trying to manually rotate a stepper motor, so its perfectly normal. I've also already swapped out the motors in one of my troubleshooting steps, so i think i can rule out defective motors as well
When I move the x axis by hand along y the extruder wobbles back and forth along x. But it hasn't impacted my printing quality or straightness. This is new behavior that happened after I did the 3.0 to 3.1 upgrade.
conscious-sapphire•3y ago
I haven't watched your video yet, but based on the description you provide I can say that I probably have a similar problem but less pronounced. I can only spot it while testing and it doesn't affect the prints.
I've double checked all the parts of the printer and haven't found any issues. At the moment I can only suspect either the belt or the printed parts of the gantry.
I have a spare belt but I was too lazy to replace it so far because my problem doesn't really affect the prints.
UPD: Oh, I checked your video. My "wobble" amplitude is maybe 0.5 mm, not that much as yours. It is more like that tiny wobble I can see on your video before the obvious drastic move at the end.
I've also got some belt wear away from the steppers which is obviously uncontrollable with the current design (3.1) as there is no height adjustment at the idlers. I am starting to think the bearings supplied with the upgrade kit are not what was dimensionally expected by the designers or they are not working well with the belts. The flange bearings don't appear to do a great job at keeping the belts in track, I noticed that during the initial install. The prints I made for the upgrade were accurate and square. The hardware fit tightly in the printed holes which is good. It's also possible the print design(s) for the upgrade may have error(s). I never had this sort of belt wear with the 3.0 design and there was no x wobble. I am experienced in 3d printing and also I am a mechanical engineer. I don't think I made a building mistake although it's possible. If I made an error, I didn't make it the first time I built the 3.0. I only removed what was necessary to do the upgrade. It's looking like I am going to have to break down the printer and rebuild. I only make these posts here because I believe it is possibly (probably) related to the OP's issue.
Just guessing: could the wobble be caused by the belts climbing over the bearing flanges and then falling back? If so, then there should be some belt wear pretty quickly. I had an issue with belt wear with 3.0, when I used my self-sourced flanged bearings (12mm diameter), and had to switch back to the original toothed ratrig pulleys. I now upgraded to 3.1 and enclosure 2, which seems to come with 13mm diameter pulleys, and have not noticed wear issues
national-goldOP•3y ago
Yep same for me, I am a mechanical engineer and I’ve build quite a number of very precise fixtures and equipments. After triple checking my assembly, I’m pretty sure I’ve got everything as close to ideal as possible. I might try to downgrade it to 3.0 and see how it goes. Thanks
I do notice some dark belt stains on the flange section of the flanged bearings, more pronounced on the motor mount side. Perhaps the stacked bearing center line axis is tilted, but it is quite difficult to measure and verify since they are all hidden within the printed parts. I’m also suspecting that i might have a slightly faulty SKR Pro V1.2 board. I’ll swap it out with a octopus and see if that fixes it
It's a shame because the bearings are a big improvement over the 3.0 design which weren't robust enough and were noisier. It may be a new design approach is required. Or, go back to the belt idlers but use higher rated bearings in them.
conscious-sapphire•3y ago
Do the FMMM mod and you will be able to see your belt math in the motor cages. It is a good mod in general and will help you to diagnose your printer.
Yes I'm thinking that could be part of the problem with the misalignment. Why didn't they make the windows taller and just add material on the top to make up for the loss of strength? Anyway when I re-do it I want to add a bearing on the opposite end like is done in the decoupler mod. Is that the same as FMMM?
conscious-sapphire•3y ago
No, it's a separate mod.
https://discord.com/channels/582187371529764864/878965315596349490/1036332584512262254
1) it almost looks like one of your idlers doesn't have the hole in the center...
2) I have the same and it looks like ringing, but it is not as it is along the whole axis. I am thinking this might be due to the actual teeth of the belt, running against the flanges of the idler bearings. My belt moves slightly up and down (no clue yet why but I am thinking of a misalignment somewher).
I made a video of my x axis wobble and sent it to Ratrig to get their comment.
environmental-rose•3y ago
Maybe I can help you with that. I am also constantly looking into (minor?) belt issues. at some point I had a worn out belt, so after replacing it I wanted the belts to be as visible and accessible as possible. I created the attached minified motor cages in order to see what's going on and to be sure the belts don't slide to the side when I put the motor cages together.
Here is Ratrig's response: "We have accessed your video and this seems to be either some misalignment in the XY gantry, or related to the belt tension. You should double-check the alignment of the motor pulleys, if they aren't correctly aligned, it can cause wobble in the belts, the grinding of the belt might also be caused by this.
If this isn't the issue, then check if the X gantry is moving completely perpendicular to the Y gantry. If neither of these is the issue, try to adjust the tension of your belts, it might be the case that one has more tension than the other, or you have too much tension on your belts.
One last thing that you can check is if one or more of the bearings stacks are over tight, and this can cause the bearing to not spin freely and create friction on the belt, causing this.
Other than this, we do not know what might be causing this, and how to help you solve it, as we have never had this issue in any of our farm machines."
national-goldOP•3y ago
Yeah i pretty much got the same response from Ratrig, but I've double and triple checked those items since they are pretty obvious areas to look into initially, still to no avail
Some updates:
1. I converted my 3.1 to 3.0 (just the X Y gantry area, that includes Idlers, X mounts, and Motor mounts) and I still observe the same issue, maybe a tiny bit reduced in terms of the wobble amplitude, but still very visible. While doing the conversion, I took the opportunity to re-measure the upper frame to ensure that it is still square and verified that all corner angles are well within 0.1deg using my bevel protractor.
2. Measured every measurable dimension on the Core XY system with my calliper and compared those with the CAD model. They are all pretty much accurate to +/- 0.2 deg so I can more or less confirm that the build is fine.
3. Swapped the mainboard from SKR Pro V1.2 to Octopus V1.1, re-installed Rat Os, did all the usual calibration and unfortunately the issue still persists. I'll re-do the wire loom connecting the XY motors to the mainboard to eliminate a possible bad wire crimp.
ambitious-aqua•2y ago
I have recently assembled the 3.1 300 ; I am experiencing the same problem;
Running the commands :
G1 X290 F10000
G1 X1 F10000
G1 X290 F10000
G1 X1 F10000
continuously
It can be seen quite well that along the Y axis the gantry moves noticeably;
I tried making a printout of a cube stretched from X80 to X200 the linear part, the problem to the eye is much less noticeable, compared to the gcode commands sent from console;
Checking the print wall, you can see defects of nonlinearity , the print looks a bit wavy; and repetitive in the same way.
I have done the checks described , but no resolution at the moment, I am quite disappointed 😦
Translated with www.DeepL.com/Translator (free version)
national-goldOP•2y ago
yeah, disappointing indeed especially after pouring so much money and time effort into building it. I gave up and planning to sell the printer to anyone who can deal with the issue (such as printing large sculptures and non dimensionally critical parts). For me, I primarily print functional mechanical models, so form and dimensional accuracy is extremely critical and this will not do. It's unfortunate, but thats the risk of such DIY printers.
optimistic-gold•2y ago
So I take it there is no cure for this issue... Frustrating. I'm struggling with it on my 500mm and need to print full bed parts with straight walls.
It's hard to sell a printer I can't fit out my door...