I have the following problem with my Anet A8.

3d-faq
3d-qa
3d-help
gplus
(Jan19998) #1

I have the following problem with my Anet A8. There are gaps around but mostly on the right of holes. They are between the perimeter and the top solid infill. Another thing i have found out is that this problem only occurs at a height of up to 5mm. I think it is a mechanical problem but i have no idea what it could be.

Maybe someone can help me?

(ThantiK) #2

Usually this is caused by pulley set screws slipping, hot end tilting in the mount, or similar mechanical issue.

(Kevin Danger Powers) #3

I had an issue where my Z axis screws were slightly bent and it caused the print head to wobble back and forth as it moved to higher layers.

Another possible issue is that you’re under extruding (though it doesn’t really look like it).

A possible fix is to use a concentric profile for your top layers instead of a zig zag. Is would cause the printer to just put more rings around the hole. It’s worth a shot.

(Richard Karlson) #4

You might try increasing extruder heat or turning off the fan so the melted filament flows together.

(ThantiK) #5

@Richard_Karlson that’s exactly what he doesn’t want to do. You don’t solve problems with 3D printers by covering them up. This isn’t just a visual thing, it’s also going to make his parts mechanically inaccurate.

Find and fix the problems at the source. Find the mechanical issue, fix the mechanical issue. Otherwise the only thing that will be worth printing is vases and other decorative items.

(Jan19998) #6

Concentric looks even worse.
missing/deleted image from Google+

(ThantiK) #7

@Jan19998 – What kind of couplers are your Z axis using? A picture of the printer would help perhaps. Sometimes manufacturers don’t realize the problems with some of their component selection.

(CELSS) #8

On ID circles, filament has a tendency to pull to the inside. Might try printing infill first before perimeters. Then (at least in Cura), you can increase the overlap, so the circle has something to stick to.

(Jan19998) #9

Normal Anet A8 couplers momentarily stiffen with tape.

Upgraded rods, linear bearings and belts on x and y axis. A few printed upgrades.

The overlap is at 25% and one time it came out good without the gaps.

This exact problem came after upgrading the things on x and y axis. Especially the belts.missing/deleted image from Google+

(ThantiK) #10

Just had a quick idea. Raise your Z axis up maybe 75mm; check the linear rods in the Z axis. Make sure there aren’t grooves worn in them. Could be allowing the X stage to rock.

Essentially, diagnostic is like this: Grab things, twist things, push and pull things…tighten up anything that moves differently than it should.

If you ‘upgraded’ things on X and Y, check what components it affected. Belts should be tight, pulleys should have their set screws double checked and tightened down.

(Jan19998) #11

On the z axis rods there are very little grooves in them.

(ThantiK) #12

@Jan19998 Depending on how deep the grooves are, they could be your issue. I’m guessing the grooves you’re seeing disappear about 5-10mm up?

That’s my best guess so far. Only so many avenues to test remotely like this. :frowning:

(Jan19998) #13

grooves fade out up to about 4 to 5 cm. why are they happening?

(Jan19998) #14

I have reversed the z axis rods and printed again but nothing changed. missing/deleted image from Google+

(ThantiK) #15

@Jan19998 The grooves are likely because the rods given with your machine are probably not chrome plated precision rods. The hardness of the chrome balls in the linear bearings are going to wear grooves into them. Since you’re homing to the build plate, you roll on that 4-5mm of surface the most, and it wears in more often.

You’ve checked your belt tensions on X/Y, as well as the pulley setscrews for those axis?

(Jan19998) #16

Both things OK. The only thing that I found is that the upgraded Glas fiber belts are more flexible than the original steel core belts.

(CELSS) #17

This is a more extreme example with a much larger hole. Worn threads are not the issue (this came off a delta), although bed adhesion clearly played a role here, in addition to the circular path filament shrinkage. After a few layers the problem went away. missing/deleted image from Google+

(Jan19998) #18

Infill before walls in on.

(Jan19998) #19

Changed a few settings like speeds, line width, overlap and filament diameter. Now it looks like this.

missing/deleted image from Google+

(ThantiK) #20

@Jan19998 You’re just masking a mechanical problem somewhere. While this does look better, there’s something loose. It might be good enough for you. Your call.