Fix z-banding w/ no backlash- "Slop Nut" (Level 2 techie - Consumer ) (Page 1 of 7)

Clever idea.

Now to adapt it to have two nuts for anti backlash.

done

This should keep the nuts tight to the thread in both directions. Instead of using a spring (which I didn't have room for) I offset the nuts. Hopefully this will take all the backlash out.

Make sure the 2 dots on the center nut huggers are on opposite sides when you install them. The nuts are off centered by .2mm in that direction.

You want to be the guinea pig?

I'm really enthused by this.

Unfortunately I'm not enthused by the thought of pulling the Z carriage to pieces again.

The results looks so good that I will test it soon, though.

I have longer, better allen keys this time so hopefully it's less painful.

Printing now, will fit tomorrow. Any tips for fitting?

More Results:

This is the first thing that I printed at .1mm layers. The grey frog was printed I think was printed with the 5/16" rod at .1mm layers. I just needed something as a reference for the pictures. My settings are off for .1mm but the quality is good enough to see that z-banding isn't affecting the frog hardly at all.

Guessing by the vertical stripes, I am getting ready to do a GT2 belt and pulley upgrade. Funny, I didn't notice those too much before this.

Before is on the left in grey. Both models were printed using my malfunctioning .1mm layer heights which needs to be modified still to work with my new hotends. The After is on the right. You can see significant reduction of any bands. The bands that you can see may be an issue with y-axis calibration and belt tension. I had a skipped step in the after print which was in the Y axis so I am hoping to get even better results after a mechanical run through.

please

Ok I've fitted the parts. I assembled them like in your SketchUp section view.

About to print test part.

Here is the experimental 5/16 slop nut. I don't know if the collared base plate will fit as I didn't measure the hole when I had everything off. If it doesn't fit then just use the un-collarded version. The 5/16 nut is too big to have 2 for backlash correction but you should see an improvement without it. An anti-backlash version can be made later but it will need longer hardware.

Print at 50% infill for strength.

Still working on M3 slop nut set.

Attachments removed... see updates below

hmm... i will test it but i fear that without the anti-backlash nut, the result is the same as the original wood plate that i mounted loose on my machine (so the screw can move on the x-y plane)

Can I ask what the hysteresis is like on your setup?
From what I can see to get the slop you have put in the nut must rotate until it jams on the edges of your print. So when you then reverse the bed direction the nut would need to spin in the housing before you get any movement  in the z axis.

Not saying this is bad as if you have all that coded in then you shouldn't see any issues at all.

If the rotation needed to jam the nut is  ~45° then with .8 pitch we are talking about .1 mm slack on the first movement that isn't accounted for. Since it is repeatable, I don't worry about because I can just take that into account when setting the z tab height. Just set it to where I need it and then 1/8th of a turn down.

Other than that, I currently have no compensation in software or hardware. I really should be using the hysteresis fix. So when I get home I will measure out my actual z-axis and change it to something like
M99 X0 Y0 Z0.1 E0 or whatever compensations I end up needing for the z-axis.

Very nice print.

Congrats and thank you on a great contribution to the community.