Adventures in Z-ribbing and my first 3D printer part.

[Ken]

I've been following along on the Z-ribbing discussions, and decided to try to build on the suggestion that Ezra had for getting a better mount for the Z lead screw thrust bearings.  I made the design in OpenSCAD and put the scad file and STL file up in Thingiverse. I wanted to keep the steppers where they were, so I made a pocket for the coupler and put the thrust bearing above that. (Mostly because I didn't have anything that I was comfortable using to displace the stepper motors).

I replaced the M3x10 screws that held the steppers to the deck with some M3x20 studs and used nuts to hold the steppers in their original location. Then I mounted the bearing holders on the same studs and fastened them down with more nuts. The bearing goes in the pocket on the top and the lead screw goes down into pocket and into the coupler.  The coupler can be adjusted up and down a fair amount.

I've tried to attach a picture of the mount in place, we'll see how that works on a first try.

I used the replacement endstop holder from thingiverse with couple of alterations to fit around the bearing holder.  (All right, it involved a hacksaw, but I wanted to try it out. I'll go back and fix it up the right way.)

Ken

[Ezra Zygmuntowicz]

Ken-

These are awesome and are exactly what I was moving towards with my start of az scad script/ We tried at least 5 different thrust beqring holders that mounted to the tslot gantry uprigth and that way lies madness as it will; never ever be 100% true and inline with the stepper since tslot is not that precie of a beast even though it is pretty darn precioe. In this case z wobble can be caused by runout at thew bottom of the leadscrews more then anything else. The tops of the screws can wobble around quite a bit without affecting or causing z banding or wobble.

These parts are what I would officially rfecommend if anyone wants thrust bearings on the bottom of their leadscrews. I'm printing a set for my own T1 today and will combine these supports with new linear bearing holders I've mocked up that do not have an "ear" for the leadscrew nut to mount to at all. All they do it hold the blue bearings and bolt to the tslot on the back of the X axis.

I will be combining these bearing holders for the linear blue bearings with new leads crew nut smooth rod followers. These are a forked tongue of sorts that wraps around two sides of the smooth rod but does not rigidly attach the leadscrew nut to the linear smooth rods. They remain completely decoupled which is what we are all shooting for here.

The leadscrews are for motion and the smooth rods and blue bearings are for guidance. Separating these two mounts so they work together still but arent rigidly tied to each other by being mounted to the same printed part is going to be the way to go once I iron out the issues and finish my parts which I will document and release as add ons you can print and install yourself if you so desire.

Here is a preview of the leadscrew snooth rod follower/forklift I'm talking about, these are not final parts bear in mind either.

It will make more sense when I show you all the parts I have in mind working in unison to eliminate z wobble/banding issues for good.

But thanks Ken for these parts, this is the only way I see making a properly aligned thrust bearing support azs mounting to the gsantry uprights has proven futile over many trials and errors.

Cheers-

Ezra

[Ken]

Hi Ezra,

I'm never more flattered than I am when someone likes something I've made.  I hope they work well for you. I was using a lighter structure than most of the regular parts, I hope they're sturdy enough.

Ken

[Ken]

I spent a while last night taking more measurements of the thrust bearings and mounts. The two thrust bearings and mounts have very different behavior.

The lead screws are constrained only in the coupler and thrust bearing. I've got a sliding platform under the X carriage ends that allows the lead screw to move freely in X and Y.

The bearing, coupler and mount that I have on the left side seem to be stable, with no measurable movement in X, Y, or Z. The left lead screw is not parallel to the smooth rod, but there's no X,Y motion at the end of the lead screw, just rotation.

The right side is a different matter. I also collected a few additional measurements on the right side. With the dial indicator on the outside bottom of the flexible coupler, I saw variance from -0.02mm to +0.05mm separated by half a rotation (which didn't seem too troublesome). At the top of the flexible coupling, I saw variance of -0.01mm to +0.05mm separated by half a rotation. That matched the bottom.

The thrust bearing was an entirely different story. Measuring displacement in the X,Y plane, I saw variance of -0.15mm to -0.01mm, which didn't seem good, compared to the left side. Then I measured displacement in the Z direction from the top of the thrust bearing and saw variance of 0.00mm to 0.05mm.

I can't work out the geometry, but I suspect that the lead screw is in a canted hole and some of the eccentricity is showing up as Z movement.

I've ordered some couplers to see how that affects it.

[Jean-François Talbot]

the motors are mounted in slot holes I believe, have you tried to move the motor to help with alignment ?

[Ken]

Hi Jeff,

I didn't see any slots when I switched the bolts over to the studs I used to mount the thrust bearing holder. I'll check further when I get the mounts off again.  If the thrust bearing mount were offset relative to the motor (and the shaft and coupler are straight), I don't think it would show eccentricity, it would just be pushing on one side of the mount. I tightened the mounts last, after the lead screw shaft and the motor  shaft. That should have allowed the mount to settle to the center.

[Ken]

I finally got the other printer parts I've been working on up to Thingiverse.

The floating X carriage mounts are here.

The compact Z endstop carrier that I use with the thrust bearing mounts is here.

And the deck fan duct is here.

I hope they're useful. They work Ok for me.  I still have to replace the lead screw coupler on the one side.

[Rick Zehr]

I like your fan duct - the one by Eugene B seems to be restrictive to me

[Ken]

I hope it works for you. I like messing with openscad, it starts you thinking in different ways.  I was most surprised when it didn't seem to absolutely require support material for the outlet opening..

Ken

[Jean-François Talbot]

So, how is the floating working out in this manner ?

Is it better than to anchor it at the top of the 3 screws going thru the anti-backlash nut, with those screws heat down and nuts up ?

[Ken]

I think it's working Ok. I don't see any X-Y displacement in the X carriage with a dial indicator, even with the bum coupler. The Z motion still gets passed to the carriage, but that has to happen, otherwise lift won't be doing it's job.

Ken

[J S]

I'm missing something... What is the benefit of decoupling? Just less chance of binding or something? Or is this the zbanding issue?

And I printed out your thrust bearing and forgot I needed longer screws! (Just ordered some). I'll try them out later this week.

[Ken]

This is my interpretation of what I've read. If the lead screws and smooth rods are exactly parallel, motion in the X-Y plane from the Z axis lead screws can cause changes in the Z height. Possibly through binding, possibly from other physical interactions. I didn't seen much improvement with the floating x carriage ends and thrust bearing mounts. But I did find that I had one eccentric flexible coupler that I'm still messing with. When I put the thrust bearing mount and the eccentric coupler together I saw the eccentricity transferred from the thrust bearing to the mount. The top of the mount was moving about about .4 mm in the X plane and 0.05 in the Z. (I didn't measure the Y displacement).
I mounted the eccentric coupler and a couple of replacements on a freshly turned stub mandrel in the lathe and they were straight and concentric there, but not on the Z stepper shaft. The stepper shaft was straight (within 0.01mm). So, I'm mostly puzzled.

I tinkered with turning the lead screw in the coupler to minimize the eccentricity and that's given me the best results so far. It's still not satisfactory, but at least it's better.

Ken