Should these be loose?

Topics around mechanical design, controller and electronics. Mods & hacks welcome.
haylcron
Posts: 21
Joined: Mon Apr 04, 2016 2:30 pm

Should these be loose?

Postby haylcron » Sun Dec 29, 2019 8:16 pm

I was getting some strange defects in the finish of some prints - especially as the object got taller. Checked some of the bolts while the print continued and noticed these bolts were loose on both sides of the build plate:

Image

Are these intentionally loose? I'd be shocked if they were supposed to be - but I'm also surprised they are all as loose as they are. Just wanted to double check before making an adjustment.
Attachments
IMG_1273.jpeg
IMG_1273.jpeg

Jetguy
Posts: 3015
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Should these be loose?

Postby Jetguy » Mon Dec 30, 2019 12:22 am

They are intentionally loose.
Tightening them is a guaranteed way to make that artifact EVEN WORSE.

https://forum.raise3d.com/viewtopic.php ... ose#p46646

https://forum.raise3d.com/viewtopic.php ... ose#p40865

Short of replacing the entire printer and tuning the PID temperature control parameters of the hotend, there are honestly some things you will NEVER solve in this design. Round unsupported vertical rods spanning over 2 feet, regardless of how strong you think they are don't even begin to reduce or slow down wobble. Leadscrews, between manufacturing, shipping and abuse are pretty much guaranteed to be anything but perfectly straight and concentric- hence the need for a sliding joint with some compliance (AKA loose) between the ball nut and the Z stage itself.

Understand that to replace the ballscrews pretty much is nearly 90% complete teardown of the printer. By design, it is one of the first parts to go in and cannot be taken out without major disassembly.- Hence, replace the printer, use alternative compensation (one of the handful of alternative mountings or methods of the ball nut attachment), or live with it.

haylcron
Posts: 21
Joined: Mon Apr 04, 2016 2:30 pm

Re: Should these be loose?

Postby haylcron » Mon Dec 30, 2019 12:54 am

Thanks for the reply, Jetguy. Since they were all loose, I thought that might have been intentional. The second link you shared is very similar to the issue I'm seeing - only mine has started recently and is much more severe. I understand there are issues with the design, bur I don't think what I'm seeing can just be chalked up to that. I have documented the issue below and would appreciate you (or anyone else for that matter) taking a look and sharing your thoughts.

Album with notes:

https://imgur.com/a/N3rGHjT

Thanks again! Guess I need to go back and loosen those bolts....

Jetguy
Posts: 3015
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Should these be loose?

Postby Jetguy » Mon Dec 30, 2019 2:56 pm

OK, so kind of a major issue in false diagnosis, that is not Z banding caused by a lead screw.

What you have is some form of backlash (AKA lost motion) in the XY system and that presents itself when the pattern of motion changes as the print gets taller. Again that's not Z wobble, that's XY motion backlash.
Time to check your couplers.

https://forum.raise3d.com/viewtopic.php ... ler#p44546

Jetguy
Posts: 3015
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Should these be loose?

Postby Jetguy » Mon Dec 30, 2019 3:23 pm

Again, looking through your posted images, you have some pretty bad problems, and yes, there is in one or two pictures, evidence of what appears to be Z banding based on the lead screw, but it's relatively light and compensation methods and or properly loose mounting of the ball nuts can reduce it. Most likely, this in some form will be present in all prints as long as you own this printer as it is a function of the Z stage design weakness. It will always get worse the taller the object is- again, by design of the Z stage.

However, you also have some rather nasty examples of either a serious backlash failure or combination of inconsistent extrusion that effectively causes that layer to not line up with the previous layer. So yes, you have a problem- no, it's no simple Z banding, it's something much worse. It might be more "fixable" because it must have recently shown up given you showed at least on example of a before and after print meaning if I was a betting person, you are having either a loose motor to shaft coupler or a broken coupler, both cause backlash where the motor turns the correct distance however the coupler fails to transmit the distance resulting in a small shift. Such a problem is intermittent (exactly what your pictures show) and direction dependent (as the slicer follows layer by layer to generate the gcode, I changes the pattern based on details in that layer). Both the motor shaft and the motion shaft of the printer have a "D" shape and then the couplers have setscrews and if that screw loosens over time, the coupler can and does intermittently slip or rock on the shaft leaving you with lost motion and position.

haylcron
Posts: 21
Joined: Mon Apr 04, 2016 2:30 pm

Re: Should these be loose?

Postby haylcron » Mon Dec 30, 2019 11:24 pm

Took the top of the machine off and the couplers are not broken, but were not snug. I tightened them up which took some of the slop out of the carriage's XY movement when locked. Re-running yesterday's gcode to see if that resolved things. Thanks for the input and linking those posts.


Return to “Hardware”

Who is online

Users browsing this forum: Majestic-12 [Bot] and 2 guests