Banding / Ringing type artifacts?

[Worldspawn00]

@radialincorporated

One pulley has belt teeth running on a smooth surface.

I’m having major x-axis ringing in PETG, and looking at the spacing of the ringing and the belt teeth, this is EXACTLY what I immediately looked at.

The difference in radius caused by the smooth pully going between toothed and non-toothed segments of the belt could certainly cause a belt-patterned resonant movement in the head as it essentially changes speed every time it goes between a tooth and a flat on the belt. And considering that there are 2 belts, vertically offset, these differences can result in a wobble in the Y axis if the 2 rods aren’t completely rigidly attached to each other and to the Y axis outer rods. A sloppy Y axis bearing or connection could be allowing the head to swing back and forth with each passing of a tooth.

IDK why they aren’t using a toothed pully in those spots, but that feels like a huge oversight. It may also be that some portion of the Y axis (rods, rod block, and head) are tighter/looser on some machines which causes the resonance to show up in the print. I’m going to look at what’s possible to tighten in various spots in the assembly and see if any of that helps.

I would also be very interested in testing if swapping the smooth pulleys for toothed idlers would resolve the issue. I wonder if they’re a standard size…

In fact! I’ve found several discussions about other printers and DIY builds exhibiting this same issue, and it was narrowed down to running the toothed belts over smooth pulleys! Motor noise and print quality | Duet3D Forum
https://forum.prusa3d.com/forum/postid/63325/

So, yes, these smooth pulleys almost certainly need to be replaced with toothed ones, the teeth moving on and off the smooth surface is causing pulsing in the axes (I’m not sure why the issue is more apparent in some machines more than others, I believe this goes back to tightness/tolerances somewhere in the X-axis gantry). Some other designs solved the issue by twisting the belts so only smooth parts are running over the pulley, but that’s not practical here due to clearance (I believe).

You CAN get away with a smooth pully, but the diameter must be equal to a 40 tooth pully to have enough teeth contacting the surface at once to not introduce motion irregularities. In the case of a GT2 belt, the minimum diameter would be 26mm, the ones in the bambu look more like ~12mm.

With the gates curvilinear profile (GT, GT2, GT3, etc), it's acceptable to run the teeth side against a smooth idler if it has a diameter equal to or greater than the effective diameter of a 40 tooth pulley of the same tooth pitch.

[drummerjoey567]

So if thats the case, who’s gonna be the guinea pig and swap them? Im almost willing to if they’re a standard size.

[Jrock]

My workaround is I reduced my accelerations

Outer wall from 5000 to 3500

Normal printing acceleration from 10000 to 8800

Reduced ringing but not completely eliminated

If theres a way to completely eliminate these artifacts and print even faster I’m all over it like a fat kid on a lollipop but in the meantime slowing it down a C hair has helped especially when I print silk PLA

Matte white PLA is a whole different story, if you’ve got gaps, blobs, bulges or zits then matte white is the chubby 20 year old recluse living in his moms basement and will show all of those issues front and center

[rovster]

On DIY CoreXY printers (such as Voron, RatRig) are successful GT2 20T tooth idlers on X in use. Those machines run similar accel/speeds as the Bambu’s do.

I have it on my task list to try out. But I didn’t dig deeper yet and miss therefore these information:

• What actual outer-diameter has the idler of the bambu
• and what inner-diameter has the bearing in it (on pictures the pin looks like 5mm for the tension idlers Replace the XY tensioner idler pulley | Bambu Lab Wiki)

If the inner-diameter is 5mm and the outer-diameter of the idler is around 9.9-10mm it can be replaced by standard CoreXY GT2 20T tooth idlers 1 to 1 without any modifiactions.

@Jrock your workaround isn’t a workaround for what he was talking about because it’s not ringing. It is VFA - potentially caused by the idlers or steppers.
VFA can be solved by more speed (not reduction) or mechanical improvements.

[Jrock]

Copy that

Appreciate the education, that explains why I can’t get rid of it completely

But it did reduce the ringing inadvertently lol

At least it looked like ringing

[rovster]

Sure the reduction of accel will help a lot to reduce/eliminate ringing.

I also run reduced accels due to ringing and simply to have a longer lifetime of the mechanic.

• 8k general
• 10k travel
• 2.5k outer wall
• 4k inner wall

[Jrock]

I know this might sound super uneducated on my behalf but what does the acronym VFA stand for ?

I’m a geek so Variable Fart Acoustics is all I could think of on a whim lol

[msinger]

@Jrock

what does the acronym VFA stand for ?

Vertical Fine Artifacts

[DzzD]

Reading the thread you linked too a bit more, a cupple of people said that changinf pulley made zero difference :

Page 10 | poor print quality

[rovster]

Because on the Prusa’s the VFA is coming from the steppers, a known problem, especially with LDO steppers - also in the CoreXY scene - LDO brought a few months ago a new stepper, which resolved the problem almost completely.

€: What doesn’t mean it couldn’t be the steppers on the Bambus as well…
What also could help, the Bambu have quite a high belt tension for a corexy printer, reduce the belt tension.

[Jrock]

That sounds better than what I came up with, more professional but actually kind of the same thing lol

Thanks Mike

[Cjs10mm]

Do not use TL Smoothers. Please don’t.

Also can we get a thumbs down emoji for the reactions?

[rovster]

So I ran a test to see what the VFA is causing on the Bambu. It’s a tower which increased print speed every few layer by 10mm/s. You can generate that yourself with the newest SoftFever Pull from GitHub.

Well, the Bambu has stepper VFA up to 60mm/s and then we have some by mechanic caused VFA which are less and less pronounced by more speed and are gone with 150mm/s. (which explains, why Bambu doesn’t have any default profile with less than 150mm/s)

But see yourself:

I also had a close look to swap the X axis idlers - there is no way to get the pin out of the idlers without removing the rods → taking the whole axis out of the printer.
I decided for myself that I won’t do that until I need to do a major maintenance. (replacing the carbon rods)

[DzzD]

Great that’s a very interesting test and information, it should be put somewhere in the doc/wiki, here it will disappear into the limbo of the forum.

[Botanic2000]

@dana.nelson

it shows very unique brass x rail bushings with sacraficial graphite beads,

Not unique by any means. I’ve been using same bushings on my old P3Steel printer 7 yrs ago. And they are dirt cheap, two or three USD a pop on AliExpress. I would assume that the price was the main factor, why this type of bushing was chosen by the BL.

[Botanic2000]

I’d be glad to do that, since I have the the same banding problem

The only problem I see on my X1C - pulley axles on Y carriages seem to be pressfit into plastic, so there is no easy way of getting them out to swap the pulleys

[Ortesse]

Setting of infill change other type of artifact but not the artifact show on all picture of this topic. For me It’s the belt and pulley and possibly the carbon rods whose create that.

[Ortesse]

with Softfever fork of Bambu studio you can do settings more precisely:

Modification post :
I have tooth idlers, I ll measure them and if it’s ok I try to replace the smooth idlers. I come back in 4 hours after the end of my print and my manipulation to report my experience with you
My tooth idlers measure:

inner-diameter :5.4
outer-diam:11.9
border : 17.85

I’d know if it is good only after disassemble the Bambu’s smooth idlers. Actually I’m printing, but after it’ll be mechanic time ^^

Modification post number 2 :
I’m trying to understand how to disassemble the idlers without broke something

If you have ideas : I take

[guillaume.leclerc.wo]

I don’t know if it was a fluke but I was able to get much better results on a flimsy Ikea table vs my super solid bolted to the foundation bench. Anyone experienced that ? (Yes I reran calibration before moving the printer location)

Top: Ikea table
Bottom: Super stiff bench

Could it be that most people experiencing the issue have tables with no damping ? Do you guys mind sharing your setups ?

[msinger]

@guillaume.leclerc.wo

I was able to get much better results on a flimsy Ikea table vs my super solid bolted to the foundation bench

It looks like you changed the model as well. Are you also seeing improved results when printing the model on the bottom in your photo above?