Page MenuHomeAleph Objects Inc

tool head rotation about Y-Axis
Closed, ResolvedPublic


As the tool head changes directions I am seeing some twisting.
Is this the scenario in which the tool head must be removed in order to tighten the compression on the X-Axis bushings?

Event Timeline

logan created this task.Dec 28 2018, 7:48 AM
logan created this object with edit policy "LulzBot Hardware Products (Project)".
tutley added a subscriber: tutley.Dec 28 2018, 7:59 AM

@logan yea i saw this a long time ago with an early quiver prototype, and i had to tighten the bushings on the x carriage. I think this is due to the 3 bushing design with the self aligner being the isolated bushing

i feel that the self aligner should be paired with another bushing and not be isolated

logan added a comment.Dec 28 2018, 8:10 AM

@tutley There aren't any self aligning bushings on the X-carriage, at least not on Alpha

@logan ah thanks for clarifying that. Sorry i havent had a quiver in a while. Anyways i corrected it by tightening the bushings adjustment screws

logan added a comment.Dec 28 2018, 8:13 AM

Ok I will do that before my next print

logan added a comment.Dec 28 2018, 9:01 AM

@tutley No improvement after tightening. I tightened until I could not twist the x-carriage about the Y axis by hand but the tool head still has a very obvious rotation.

@logan is it the entire toolhead moving?

@tutley Yes it is quite noticeable, I think it may've gotten worse after I tightened the compression also. But without a video for reference or measurements I can't say for certain. Is it possible we're compressing these bushings in the wrong direction?
Looking at the assembly, the bushings are compressed front and back which should cause an oval shape, giving the tool head more room to rock rather than less.

@logan well i think you would have seen the x carriage rock without the toolhead if that were the case

@tutley To clarify, it was still rocking a little by hand but it is much less pronounced without the tool head. I was afraid to tighten any further as drag was already approaching what I would perceive to be too tight.
But if you think about the arrangement of the compression, how would it not oval the bushing elongated vertically causing more rotation?
The fact that ALL parts must be removed from the x carriage in order to adjust (tool head, cap/board cover, interface board, board mount) makes this arrangement far less than ideal.

It is printing one of your vernier scales right now, feel free to come take a look if you like.

@logan the betas wont need all those parts removed to adjust this. it is just removing the toolhead and the screws are right there

For betas could we try compressing the bushings in a direction that would be helpful? By doing so we could likely make the screws accessible from above and below without removing anything.

tutley added a subscriber: west.Dec 28 2018, 12:15 PM

@logan yea that is definitely the whole tool head. I honestly think it is from having 3 bushings instead of 4, or that the 2 are on bottom and 1 up top with the belt mounting close to the top single bushing. Ben Baldwin ran into this when he tried to convert a gladiola to a 3 bushing set up on the x axis. Id like to come take a look at this machine at some point today.

@west do you think we can easily change the bushings to compress from top and bottom instead of front to back?

FWIW i dont think that rocking is any worse than the Taz6 chevron mount gets but i still think it needs corrected

west added a comment.Jan 2 2019, 7:24 AM

The new part for betas seems to work, it also moves the tension screw to the from so the only thing you have to remove is the toolhead T4833. I dont see what changing the tension screws from front and back to top and bottom would do. The belt is mounted closer to the top rather than center is because of how close the rods are

logan added a comment.Jan 2 2019, 7:26 AM

@west Because we're compressing the bushings into an oval shape. Doing so from front and rear elongates the bore vertically, whereas doing so from top and bottom would elongate the bore horizontally. We want to avoid vertical elongation of the inner bore as it will increase the motion seen in the video above.

logan added a comment.Jan 3 2019, 8:46 AM

After loosening the compression bushings, the rotation is still there but less pronounced based off of observation alone, in my opinion. I cant think of an easy/quick way to measure that.

logan triaged this task as High priority.Jan 3 2019, 10:01 AM

This causes a failure of circular dimensional accuracy because lumps are created at the X axis direction changes.

west added a comment.EditedJan 23 2019, 7:35 AM

this moves the toolhead back and forth 5mm at a 15000 mm/min feedrate to see if there is rock in the carriage

logan added a comment.Jan 23 2019, 8:46 AM

To clarify: this is not isolated to Q11 and has been visible on every Quiver I have seen printing. It is obvious.
Q11 is not functional at this time.

logan added a subscriber: Steven.Jan 23 2019, 9:57 AM

Got Q11 running with the latest parts delivered by @west
Its better, but still visibly present.
@tutley @west @Steven feel free to come gawk at it when you have a moment, running a print now

logan added a comment.Jan 24 2019, 9:41 AM

@Steven @tutley @west
I made a mount to attach a digital indicator to the top of the tool head to measure off the top of the X Axis smooth rod to the right of the tool head.
I then ran the rock.gcode first with both X axis compression bushings loose. This was with the new parts and brand new bushings. The numbers reported were -0.04 to 0.18
I then installed the bushings removed yesterday and began running the following gcode

First test with both compression bushings loose: -0.09 to 0.04
Bottom compression bushings tight: -0.10 to 0.06
Top compression bushing tight, bottom loose: -0.03 to -0.07 *best
Both compression bushings tight: -0.06 to 0.05

The indicator can easily be installed to any quiver, it attaches to the inserts used for the cap. It must be left off during probing but can be attached during the purge with one screw.

Here is a pic of the gauge mount

DaniAO added a subscriber: DaniAO.Feb 13 2019, 9:58 AM

Going to see if this shows up in Betas

@DaniAO So you want this gauge mounted to each of the betas to show that this is still happening or? What information is needed to get this verified issue fixed in the design?
I have no doubt that this will be present on every machine, to what extent depends on how the compression bushings are set, which we were not provided with any information regarding what they should be set to.

@logan there hasn't been any updates on since Dec. Have you noticed this on any of the Betas?

@DaniAO Last update was 2/27 (Feb). See unanswered questions above. The beta I received had this issue initially, quite noticeable. I was able to reduce it greatly by ensuring the upper bearing holder is tighter than the bottom, and the bottom tight enough to reduce wobble but not introduce excess drag. I did not get the opportunity to mount the gauge to it so I do not have any data on that.
Otherwise, my exposure to betas has been limited, @matth has that machine now and I have not been assigned to development since 3/8 so my focus has been elsewhere.
@MikeR may have more information as his exposure to betas has been greater.

QB16 on @EricNugent's desk shows this issue very clearly. The rotation is quite bad on that machine.

The machine, QB16, seemed to have a reduced amount after implementing the steps in the earlier post.

@EricNugent can you please confirm if the steps followed to reduce the concern are documented in the Ohai for this part of the assembly?

Steven assigned this task to EricNugent.Mar 22 2019, 2:30 PM

@Steven I didn't see an ohai for this stage of the assembly finished yet but will be sure to include some steps in there to ensure this process is implemented. The unit I performed this on was shipped out for a show so it will take a few more cases of this popping up before we have a sample size large enough to determine if this is a repeatable problem and solution.

kent closed this task as Resolved.Apr 2 2019, 12:08 PM
kent added a subscriber: kent.

Step 7 here outlines the process that we have determined will mitigate this issue.