Dryspin Z with no linear hardware

[Ryan Carlyle]

Igus Dryspin screws have a thread shape that allows side-loading the nuts, unlike ISO trapezoidal and Acme threads. And Dryspin screws are really, really straight, according to people I know who have used them. For example, Shauki B made a screw delta using one linear rod and one Dryspin screw per column, supported top and bottom, and it worked just fine. No wobble.

So... How about a Z stage that ONLY uses two dryspin screws? No rods/rails. Each would have two nuts spaced apart for moment stiffness, with a slight spring preload between them to take up thread wear. 

[mroek]

Interesting idea! Getting enough moment stiffness might be a challenge, unless using fairly beefy screws. It depends on many factors of course, and for a small bed size I guess even the smallest diameter screws could be used.

[Steven Butterfield]

Any idea how they get the lead screws so much straighter? Is it the manufacturing process, better tolerances/ quality control? Or is it that the trapizoidal screws are just that cheap?

[Patrick Barnes]

I have a related question... 

Say you had 3 linear steppers (screws) supporting a print platform up in the air. (with a somewhat-compliant linkage)

Assuming the platform is flat and stiff and the steppers are individually controllable; You get real bed leveling and no large cantilever forces.

But of course, the screws are not straight, and the platform is not adequately constrained against XY movement or Z-axis rotation.

Is there something that would remove those three degrees of freedom without over-constraint?

On Mon, 13 Jan 2020 at 08:38, Steven Butterfield <steven.butt...@gmail.com> wrote:

Any idea how they get the lead screws so much straighter? Is it the manufacturing process, better tolerances/ quality control? Or is it that the trapizoidal screws are just that cheap?

--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/26ba6b43-e783-4dad-8a9f-895e7d03a97c%40googlegroups.com.

[David Crocker]

The usual solution is 2 linear rails, one on each side of the bed. This is what the RailCore printer uses. In theory it is over-constrained because the distance between the rails is fixed and they don't allow the carriages to rotate much, but in practice it works as long as the leadscrews remain sufficiently in sync to keep the bed close to level. And yes, you can use the Z probe to do true bed levelling, which is supported by RepRapFirmware and Repetier firmware.

[Ryan Carlyle]

I think it's quality control. Dryspin screws are rolled, Tr8 screws are lathe-cut, both may have some residual metal stresses that cause warping in slender shapes. But there may be some significant process control or heat-treatment step or something that's different, I wouldn't know. 

With linear rails and ball screws (which are ground) the problem with cheap Chinese products seems to be that they just don't care if the tooling is getting worn out or the surface finish is poor or the bearing balls are crap. Cheapo Chinese screw vendors sell us shitty product because we accept shitty product. 

[Ryan Carlyle]

I'm thinking Dr12x05. That's about as stiff as 10mm linear rod, and easily changed out for 1205 ball screws if there ends up being an issue with plastic nut friction or whatever. Could go bigger diameter than that, say 20mm, but the bigger Dryspin screws tend to be really long thread lead (fast screws) so you lose resolution and need more motor torque. 

[Ryan Carlyle]

I have seen somebody (on G+) lift a bed with 4 screws and no rods/rails. They say it worked fine, no wobble. If your screws are straight, and you have 3-4 screws so there's no bending loads, it should work. You'd want to constrain both ends of the screw in that case I think. 

I personally wouldn't want to use 3 screws with no rods/rails for true bed leveling, just in case the screws get out of sync you'll be twisting the nuts HARD on the screws. 

On Monday, January 13, 2020 at 10:27:24 AM UTC-6, Patrick Barnes wrote:

I have a related question... 

Say you had 3 linear steppers (screws) supporting a print platform up in the air. (with a somewhat-compliant linkage)

Assuming the platform is flat and stiff and the steppers are individually controllable; You get real bed leveling and no large cantilever forces.

But of course, the screws are not straight, and the platform is not adequately constrained against XY movement or Z-axis rotation.

Is there something that would remove those three degrees of freedom without over-constraint?

On Mon, 13 Jan 2020 at 08:38, Steven Butterfield <steven.butterfield2008@gmail.com> wrote:

Any idea how they get the lead screws so much straighter? Is it the manufacturing process, better tolerances/ quality control? Or is it that the trapizoidal screws are just that cheap?

--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.

To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+unsubscribe@googlegroups.com.

[Whosawhatsis]

I think Shauki was doing that at some point, not sure.

Of course, the Makerbot CupcakeCNC did that, but they were lifting the extruder over an X/Y platform. I remember seeing a mod where somebody put a larger build plate on that platform and built a new extruder X/Y stage over it.

To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/7ff7ee5c-e800-485c-a2cc-7d4a47064591%40googlegroups.com.

[Whosawhatsis]

The first machine I saw using the 3-screw independent system for auto-tramming used ball joints, but it didn't add much range of motion because they were still a fixed distance from one another. If you REALLY wanted the platform to be tiltable, you could mount a ball on each Z carriage and put the printbed on a kinematic coupling that spans them. Kinda seems like overkill when the whole point is to automate getting rid of the tilt, though. You'd also need each screw to be combined with a linear slide that produces full 6-axis constraint, but at least the kinematic coupling would prevent them from creating an over-constrained system.

--

You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/f3d4f624-1e9d-4f51-b21d-60708aab5027%40googlegroups.com.

[Ryan Carlyle]

That's exactly one of the design suggestions you will be seeing when you review chapter 4  :-D

Another option I like, but ends up being sufficiently complex that it probably isn't worth doing, is putting the Z stage on 1-2-or-3 belt-synchronized screws, then putting the build plate on a three-leg leveling jig with one fixed ball joint and two very small screw drives. Like this: https://www.amazon.com/dp/B07QYB4C7P

So you'd get to build your big 500mm travel Z stage or whatever without worrying about excessive tilt breaking anything, even if you use back-drivable screws. But you can still three-screw autolevel. Probe at the ball joint running the main Z stage, then use the two micro screws for fine leveling. 

To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+unsubscribe@googlegroups.com.

[Dushyant Ahuja]

This is what I've been thinking about doing - though not sure if support is available in the current firmwares.

I have a belt-driven printer with 800mm Z and am too cheap to use ball screws of that size :-)

Will probably get to it once my current project backlog gets down.