Z artifact diagnosis help

[jrlowe]

My 3d printer is a prusa i3x which I have had for a little over a year. Over the course of that time, I have made many upgrades, improving the print quality. The last thing I can not work out, which has stumped me for months now is z artifacts, which I have dug online for hours but all the answers are things I have tried. Here are upgrades I have made:
-got rid of low quality threaded rods and go 8M 2mm pitch ACME Leadscrews which run very smooth and straight
-brass antibacklash nuts for the rods
-aluminum flexible coupling
-decoupled the x carraige from the z axis (custom made decoupler) (This was designed to prove that the problem was not z wobble, for, the print quality it the same)
-Constrained the top of the leadscrews with 608 bearings (with and without yields the same problem)
-reprinted elliptical (bad) extruder gear which came with the kit
-calibrated extruder (calibrated very accurately)
-Tried with cooling fan on/off, heated bed on/off
-printing the same object normal speed, and very slowly (- 10 mm/s) and the quality was the same (which I believe rules out vibrations and backlash)

As you can see below, I vertically lit a Z resonance test print, revealing ugly misaligned layers.
In this print I:
-printed very slowly
-heated bed off (to prevent expansion & contraction of glass)
-Makerbot PLA @205 degrees
-.25 mm layers (rounded based on reprap calculator)
-Made 2 towers at different speeds, (normal and very slow (this one) and they turned out the same)
-had cooling fan on

Z wobble would indicate that the big bands are evenly spaced, matching the thread pitch, but I did not notice any repeated banding, or anything matching my leadscrews.

One possible culprit could be my hobbed bolt as seen here:

which is slightly not concentric, but there I don't believe such artifacts could be caused by this.

Printing at lower layer heights like .1 make the bulges smaller and less noticeable, but they are still there.

To rule out more possibilities, I calibrated my z axis steps based on my leadscrew, stepper drivers and motors, which is not rounded at all. I also used a calculator on reprap to find the optimal layer height for my z axis setup that will result in no rounding errors over time, which I used in the print.

If anybody knows what is going on here that would awesome! I've been pulling my hair out trying to get this thing figured out.

[3dkarma]

Raise the Z axis about 50 cm and check the horizontal gap between the lead screw and vertical smooth rod by the Z flexible coupler, and by the lead screw nut. If they're not the same you may get some banding issues. If they're not the same for both left and right, you may get some banding issues that'll look fairly odd.

Are you getting any noise from your linear bearings? If they're on their way out they may be exascerbating the problem. Consider replacing with igus bushings.

If you've been upgrading your printer, seriously consider replacing the hobbed bolt. You can at least eliminate this as a source of the problem.

[jrlowe]

Raise the Z axis about 50 cm and check the horizontal gap between the lead screw and vertical smooth rod by the Z flexible coupler, and by the lead screw nut. If they're not the same you may get some banding issues. If they're not the same for both left and right, you may get some banding issues that'll look fairly odd.

Are you getting any noise from your linear bearings? If they're on their way out they may be exascerbating the problem. Consider replacing with igus bushings.

If you've been upgrading your printer, seriously consider replacing the hobbed bolt. You can at least eliminate this as a source of the problem.

Alright so I just recieved my new hobbed bolt, which I got from Air Wolf for $12 including shipping and it is very high quality and very even all around. I put it in my extruder and am reprinting the tower as I type this. Unfortunately, I can already tell that there is not a noticeable amount of improvement, if any. At least I ruled out a source for the diagnosis.

For my linear bearings, they seem to be fine, and they make zero noise at all and still run smooth, and I have greased my rods to make it even smoother.

I read that stiff couplers were the way to go, especially with decoupled x carriages, so Im wondering if I should try nophead's coupler from thingiverse. This eliminates and unwanted play in the flexible couplers because my antibacklash nuts and decoupled parts should be able to handle it fine.

[jrlowe]

Here is another print showing the artifacts. On the left is printed at .25 layer height and the right is a little over .1mm. As you can tell, the layers are much more even on top of each other, which is odd, because if something is off, or z wobble, there will be more layers for error. (Ignore the blob on the left, the computer freezes on circles sometimes). Hopefully this can help anyone identify the problem.

[morgue427]

Is the nozzle loose had a problem with mine. Left and right side of cube looked great, front and back was jagged ended up the nozzle had play in the extruder a big washer fixed it for me.