Print Quality

On all of my prints, they appear to have an even band running up the z direction in the print. I have done tons of research on this and I have not fixed the problem. I made sure the steppers are not rounding to certain layers. Also, the threads match the bands perfectly, an m5 course thread. I have learned that this is from z wobble. I unconstrained the threaded rods by allowing some movement side to side in the tube couplings but this had no affect. This print is .2 layer height with .15 infill. At this point I have tried so many things with no success so I'm hoping someone on here has fixed a similar issue to where the banding is not visible.

What printer are you using? Knowing a bit more about how your Z axis is configured will help troubleshoot your issue.

swirv

I am using the prusa i3x. Something to note Is that due to the poor frame design, the vertical axis slightly wobbles along the y, but I don't think this is the issue because the bands go all around the print. In the picture I have a view of the z axis threads and the coupling to the motor. Just to toss out any information...the threaded rods do have the ability to move around I the coupling horizontally to prevent restriction. I also marked with tape on the top of the threads to make sure that neither of the rods slip in the couplings, and they don't. Thanks for the reply by the way!

The banding doesn'nt necessarily come from the X carrige being pushed around by the Z-leadscrews.

No matter how loose you couple the motors to the leadscrews, if they are bent, they will enter the nuts at an angle, and this angle will change as the leadscrew rotates, this again means that your point of contact between the leadscrew and the nut changes with rotation of the leadscrew, and the end result will be non-linear movement in the Z direction.

When the Z direction doesn't move linear, some layers will be more compressed, and therfore wider than others, and this will be seen as banding.

The effect is made worse by the fact that far from all nuts have the thread cut perfectly perpendicular to the faces that the carrige rest on.

Alright, so even if I have the z axis guide rods with linear bearings on the x carriage then there is still a little bit of shifting room for the lead screws to move it around I'm assuming? Would this change in the z due to different angles of the screw in the nut even if I can tell both lead screws are spinning the same amount or would the change be too small to notice...This reason does make perfect sense but I thought that because I marked the screws with tape at the top to visualize their rotation and they were always the same that this wasn't an issue, but we must be talking like fraction of millimeters, and that is definately not noticeable with tape. To solve this solution should I just buy some more lead screws and hope they are straight or also try different pitches and size?. I assume that thicker lead screws are more likely to be straight but if they are bent, they are stronger and force the whole x carriage assembly around with every slight bump.

We are talking about small differences in layer-to-layer height, but imagine 3 layers in a row at 0.8mm and then 3 layers at 0.12mm when the expected layer height was 0.10mm, that is a difference of a whole layer.
On a complete rotation of the leadscrew, the height will match again, but when the screw rotates further it starts over, and the layers of the first half of rotation might be too small, and the next too heigh.

If you has leadscrews with the exact same pitch as your layer height, the problem will be completely elliminated, but printing with 1mm - 1.5mm layers involves other challenges..

Finding straight leadscrews can be dificult, M8 threaded rod isn't exatly produced with linear motion in mind. I would start trying to straighten them by hand, and if this fails, go the the hardware store and find one that is straight.

Alternatively find some trapezoidal leadscrews, they are designed for linear motion, and are generally way better quality than hardware store threaded rod.

That makes sense. What are some or these challenges printing at .1 mm? I've printed a few things this resolution and the prints turned out really well except for the banding, which is a little less noticeable, but still present. Finding a lead screw with the pitch of the layer height I want to print would result in having to get threads almost 5x finer than now. I will try using a trapezoidal screw and see how that works out!

To avoid wobbling effects you could also try to decouple your Z and X - this way no matter how much your leadscrew wobbles, it woun't affect your head position at all. CHeck this one out:
[www.thingiverse.com]
Good luck!

What does it it mean to decouple the z and x axis'?

It means that the x axis is not tied to the z axis. The nut that is moving on the leadscrew is on a different part from the x-ends. The x-axis lies on top of this part but is not fixed to it. The added bonus is that if by any chance your hot end hits the bed it won't be forcing down, as the x axis is not being actively pulled down.

The poster is already using m5 rod on his i3x with M8 smooth rods. That design in itself should take care of the z wobble from everything I've read as long as the threaded rods have some flexibility with the vinyl tubing. I have an i3 alpha with a similar setup and my m5 rods are not very straight at all, but I have no issues like what's showing here. The thingiverse decoupler looks like it's for machines that use M8 threaded which are known to have more issues with z wobble.

Edited 1 time(s). Last edit at 07/18/2014 10:23PM by brucehvn.

Do you know what may be causing it then?

Printing a couple of decouplers is a lot cheaper than buying trapezoidal leadscrew and not a huge investment in time. The wobbling could come from what Ralf suggested above, which would also be corrected using this method. If it doesn't improve then start looking elsewhere.

I found some trapezoidal leadscrews here: [openbuildspartstore.com] Along with some anti backlash nuts here: [www.robotdigg.com] Has anyone tested any of these? Would these solve most of my banding issues?

It can be a problem with your layerhight and your z-axis pitch.
Take a look at this.Taxonomy of Z axis artifacts in extrusion-based 3d printing

As a person with Z wobble issues I feel for you In the article linked to in the slic3r manual:
http://goo.gl/ci9Gz
They mention other issues that can cause z banding. One of them was tension on the filament.

Quote

A similar and probably more common error is resistance from the spool pulling against the extruder, which will tend to lift it. This is most prevalent on machines that move the extruder on the Z axis, especially if it is on a cantilevered platform like on the Thing-o-matic. In these cases, the filament will pull until there is enough tension to pull it loose, and the extruder will fall back to its normal position.

You may want to check that out too.