Winch Z stages

[Ryan Carlyle]

Why doesn't anybody ever lift Z stages with a little winch? I mean, obviously there are line-wrapping challenges and such, but it seems like the sort of thing zany hobbyists would experiment with more often. 

[Patrick Barnes]

But then you couldn't print upside down! :-O

Perhaps it's a problem with Z-stage stiction? (static friction)

On Mon, 18 Nov 2019 at 14:20, Ryan Carlyle <temp...@gmail.com> wrote:

Why doesn't anybody ever lift Z stages with a little winch? I mean, obviously there are line-wrapping challenges and such, but it seems like the sort of thing zany hobbyists would experiment with more often. 

--
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/08ae3069-6461-4dd5-b4fa-9f4429986b4f%40googlegroups.com.

[Whosawhatsis]

You mean rolling up a string of some kind to lift it? Once the string starts wrapping over itself, the effective diameter of the pulley will increase, and it will lift more each rotation. There will also be a smaller inconsistency effect as it walks back and forth across the spool. Better to just use a return to the bottom side (like a belt drive), which also means you don't have to rely on gravity.

On November 17, 2019 at 19:20:53, Ryan Carlyle (temp...@gmail.com) wrote:

Why doesn't anybody ever lift Z stages with a little winch? I mean, obviously there are line-wrapping challenges and such, but it seems like the sort of thing zany hobbyists would experiment with more often. 

[Ryan Carlyle]

A spool at the bottom, running to a grooved idler at the top, should get you plenty of travel on a single spool wrap before the fleeting angles get bad and you're meaningfully distorting the print or starting a second wrap. People successfully run fishing line spools for Deltas and CoreXY, both of those need more travel than a Z stage. 

I totally get why there are downsides to spool winch type designs... what I don't understand is why I nobody seems to do it. People do much harder things with winches, like the Flying Skydelta cable-driven parallel robot mechanism, or the Hangprinter. Seems so simple for a basic Z stage. 

Only one I can think of is the RepRap Wally. And that kind of NEEDED a winch drive since the Z stage was mounted on a linkage, not linear hardware. 

On Sunday, November 17, 2019 at 11:32:33 PM UTC-6, Whosawhatsis wrote:

You mean rolling up a string of some kind to lift it? Once the string starts wrapping over itself, the effective diameter of the pulley will increase, and it will lift more each rotation. There will also be a smaller inconsistency effect as it walks back and forth across the spool. Better to just use a return to the bottom side (like a belt drive), which also means you don't have to rely on gravity.

On November 17, 2019 at 19:20:53, Ryan Carlyle (temp...@gmail.com) wrote:

Why doesn't anybody ever lift Z stages with a little winch? I mean, obviously there are line-wrapping challenges and such, but it seems like the sort of thing zany hobbyists would experiment with more often. 

--
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]

Oh, another problem with using a pulley without a return to pull the axis down is that if the axis gets lifted (or fails to drop when lowering) for any reason, the line will go slack and jump off the pulley, and likely get tangled in something. All of this can be avoided by closing the loop with a return on the bottom side.

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/69a33b32-87be-4ec5-b5af-0c418f31fbd9%40googlegroups.com.

[Russ Taber]

Not a winch per se, but a cable drive that does the z-lifting with help from counterweights. Had it at MRRF2019.

Lifted  my Core XY cable driven frame well enough for the project intended--this 6-ft sculpture.

Initially thought of it for a I3 style lift.

But sketch-thinking, evolved to the system above.

[Ryan Carlyle]

Russ, I'm pretty sure I remember seeing your printer at MRRF but don't recall noticing it had that kind of Z lift. Maybe I did know that? My memory lately is pretty shot from chronic sleep deprivation tbh. 

I was just writing in my book vol2 that a traction winch like that was unlikely to work well due to slippage, but I see you have proven me wrong. I found spectra-on-steel to be too slippery to make it work, but in hindsight spectra is just about the worst possible cable selection for a friction drive. 

• What kind of cable is that? 
• Pulley material? 
• Print/machine/buy the pulleys? 
• How consistent do you think the layer height is? The sculpture prints have horizontal artifacts (which does not detract from their attractiveness), is that a cable drive issue or extruder or unknown?
• Do you recommend the approach? 
• Can I use your photos/renders to illustrate the drive type?

[Russ Taber]

Hey Ryan, a bit of history. Ran across the cable drive idea some time ago. Was spectra line over machined pulleys. 

http://3dprinterhell.blogspot.com/2013/06/a-better-filament-drive.html

Also, Fusion3 printers used a core xy cable system with accuracy so it seemed viable for my ideas—cheap z-lift that isn’t overly complicated. CAD modeled a few ideas.

At MRRF2018 chatted with Sanjay from E3D. Asked what he knew about cable drives. Said he had not studied them much, but asked for my email address. Immediately emailed me a link to a Stanford professor’s lecture on YouTube.

Reuben Brewer’s cable transmissions: https://youtu.be/jKZIvseA1Nk?t=389  -- skipping the 6 minutes of extraneous classroom talk.

From that, I realized it was highly viable—especially using more wraps and moving the pulleys closer to increase surface contact. Wrap something enough times and minimal effort (tape) is all that is needed to hold the end—no matter how slippery.

Tom’s MRRF2019 interview showing my TowerBot: https://youtu.be/-FvGRm7I1Dc?t=16m24s

Used generic Spectra 1.2mm 8-strand, 300lb. test. The 80-90lb. I used for the counterweights stretched too much over the large distance—up and down each vertical edge and framing the bed—roughly 20 meters for my printer. 

Considered steel cables but for the strength/size I needed would have to use larger v-groove pulleys for cable arcing tolerance.

Initially printed the pulleys with PLA, but the cables eventually cut into them. The walls were thin encasing the bearings so went to the next size up tying to keep the diameter it a ratio with the stepping rate. Went with PETG with far better results. Did 5 or 6 wraps. One more might not hurt. Nylon would probably be best, but was running out of time. Or cost no object, machined metal ones.

Vertical accuracy, to the best of my sleep-deprived memory and lost scribbles, was a shocking fraction of 1% for repeatability with fast z lifting 5 times (a mm or two over 1800mm movements. Far better than my initial core xy tests. But those were skewed due to the breakdown of the PLA pulleys. 1mm nozzle, 0.8mm layers, and 1.5mm line widths allow for a bit of variance.

Horizontal artifacts are mostly due to the core xy system. The bushings on the carbon fiber may bind a bit. A bit of play in direction change—cheap v-groove pulleys with screws as axles. Long Bowden feed for slower pressure response. The cable pulleys tend to walk up and down on the screw used as axles.

I didn’t account for the weight of the cable bunch at height for the counterweights so the first tall print struggled for the last 5-10% until corrected with 14oz. bags of marbles.

A larger NEMA for the z-lift, either a long 17 or a 23 would surely have less struggles. 

Yes, you can use my images for your book with accreditation. When you’re closer, I can provide you with higher rendered images.

Recommending it if you are going for cheap, some portability, and a bit more in size flexibility. Helical threaded rods would be far superior in many other ways, but much more costly.

I considered a z-drop design (could be quite interesting with more of a winch system) but with my machine so tall, would have been a challenge to work on the core xy parts. and the considered heat bed would have cable challenges. 

[Ryan Carlyle]

Thanks Russ!