Layer Consistancy

@akhlut

I use a similar z-axis wobble elimination method on an Ord Bot Hadron. I was actually going to try changing my z steps to try to reproduce your result tomorrow.

I agree that your z axis design should virtually eliminate z wobble as long as the z threaded rods are not making contact with the x carriage at any point aside from the z nuts. If they can make contact, then you can still get z wobble. Thus, with this set-up, it's best to use a z couplers that are rigid. Those flexible couples can result in the rods making contact with the x end.

Just to add to the mix, here's how I set my steps per mm in my Prusa2 that uses 5/16" rods:

#define DEFAULT_AXIS_STEPS_PER_UNIT {80, 80, 3200/25.4*18,98.3*25/24.15*50/35.18}

The z steps per mm is being calculated by the compiler and definitely has more than 2 decimal places. With this, I get perfectly lined up walls as long as z wobble is not in effect: [www.flickr.com]

I went ahead and did a quick test at different z steps per mm. Left: 2267.72, Middle: 2267.71654, Right: 3200/25.4*18. No difference.



Edited 1 time(s). Last edit at 03/03/2013 08:18PM by brnrd.

I have put a few small dabs of CA glue between the Z bushings and X mounts... I could wiggle the whole assemble before, now its sturdy. There are also Z constraints on it. Printing something now and Ill probably know tomorrow morning if it worked.


-Nick

@ brnrd,

Seems reasonable.

Going to print out new X-ends when I get a chance.

---

- akhlut

Just remember - Iterate, Iterate, Iterate!

[myhomelessmind.blogspot.com]

Wow, printed off the new parts with the glue on the X ends and the layer alignment is pretty much perfect!

There are a few sub-millimeter "bumps"... theyre are like blobs, but nearly invisible... any ideas?

Will post pictures if I get a chance.


-Nick

could be from oozing, one way that helps is to increase perimeters. if using slicer the latest version is better at it.

akhlut Wrote:
-------------------------------------------------------
> OK.
>
> I changed the Z-steps in firmware from
> 2035.5809493712193 to 2035.6
>
> 2 prints
>
> Same Gcode
>
> No mechanical changes, 1/4" ACME screws
>
> Same plastic
>
> I think the image below speaks for itself.
>
> [i.imgur.com]

If you had asked me what changed between those two prints, I would have guessed that you got your temperatures under control. Any chance you tuned your extruder's PID between those two prints, or made any other changes? That doesn't look like the results of a change of 0.002%.

Dale,

no changes to temp. PID control of both hotend and heated bed. Also, heated bed has nylon standoffs to help decouple it from the leveling screws beneath. Typical procedure is to heat bed to 110 for 5min to allow it to stabilize in Z, then start printing.

Also, I want to thank you for your X-Z decouplers! I've gone and redesigned them to fit my machine, but they're a lifesaver!

---

- akhlut

Just remember - Iterate, Iterate, Iterate!

[myhomelessmind.blogspot.com]

I thought those looked familiar. I see your machine assembles from 90° in the other direction.

A change that small made that much difference. Interesting.

Yeah, the Z axis is flipped 180° - smooth rods close to the frame, drive rods on the outside. Maybe this in and of itself is an issue - maybe it helps induce z-wobble. Never considered that. But flipping back to a standard X-end would mean a redesign of the machine. Personally I'm thinking the X-ends themselves need a complete overhaul at the minimum...we'll go from there I suppose.

But yeah, I was kinda shocked at the difference. Maybe I had chosen the exact wrong layer height. Who knows. But I've since moved to a true full-step layer height and prints are looking even better.

---

- akhlut

Just remember - Iterate, Iterate, Iterate!

[myhomelessmind.blogspot.com]

I think the issue is with iround and how the firmware handles floats. it is looking for numbers that are greater than zero based on a float. a number can be near zero and still be above.