Preferred Z Axis Style?

What is your favorite type of Z arrangement for a Core XY machine? Belts/Screws? Cantilevers?

Just need some opinions for my printer build.

I'm building a SmartRap XL300 which uses belts. They work, but if you cut stepper power the full weight of the bed will crash down because the belts offer very little resistance. I will probably work screws in for the Z on mine at some point. I also have to see how the 4 linear rod setup on the XL300 works over time. At least with screws the bed will either stay in place or move down slowly because the screw will offer some resistance. You really don't need Z moving fast like X and Y usually need to do.

I am using three screws connected with a belt (powered by one motor). I've chosen 2mm pitch/lead screws, so I could be using full (or half) steps to move Z and still get reasonable layer heights. I don't think microstepping would be able to provide accurate heights alone... Z speeds are quite low with this setup, but it's not like you need them to be fast...

Edited 1 time(s). Last edit at 08/31/2016 03:30PM by Dalius98.

Cantilevers is the easy solution for small bed, but the bed structure need to be rigid.
I using a cantilevers 200x200, supported by 2 x 12mm rods and one 8mm lead screw with 2mm pitch, using this setup for almost 1 year and it works with no issue.
I know upgrading to a 310x290 with 2 x 16mm rods and 8mm lead screw with 1.5mm pitch, but this was not tested and can not guaranty it will work.
Any bigger bed and i think cantilevers is not more a option.

Quote
Dalius98
I am using three screws connected with a belt (powered by one motor). I've chosen 2mm pitch/lead screws, so I could be using full (or half) steps to move Z and still get reasonable layer heights. I don't think microstepping would be able to provide accurate heights alone... Z speeds are quite low with this setup, but it's not like you need them to be fast...

Are you using any linear motion components like smooth rods or hiwins? Or is it just leadscrews?

duplicate removed

Edited 1 time(s). Last edit at 09/02/2016 10:44AM by MatthewHall.

For most applications, screws are the way forward. I've never been keen on the belts method as it can't hold its own weight, and belts tend to wear faster. Belts are much better for high speed applications, but the Z Axis will never have to move that fast. A leadscrew is the way to go. I'm not even sure why people go for belts, as surely a length of GT2 belt and pulleys costs just as much as a screw + nut?

As for the cantilever, this is probably one of the most complained about issues of the ultimaker style machines with many 'fixes' floating around. I'd personally go for 2 or 3 leadscrews and guides instead of a cantilever design. I'm not saying cantilever is bad, many parts of our machine are 'unsupported' and work just fine, but this must be adequately designed to work. Cantilever does have a clean look though, and is simple to implement.

Edited 1 time(s). Last edit at 09/02/2016 05:23PM by Origamib.

Quote
Origamib
For most applications, screws are the way forward. I've never been keen on the belts method as it can't hold its own weight, and belts tend to wear faster. Belts are much better for high speed applications, but the Z Axis will never have to move that fast. A leadscrew is the way to go. I'm not even sure why people go for belts, as surely a length of GT2 belt and pulleys costs just as much as a screw + nut?

As for the cantilever, this is probably one of the most complained about issues of the ultimaker style machines with many 'fixes' floating around. I'd personally go for 2 or 3 leadscrews and guides. I'm not saying cantilever is bad, many parts of our machine are 'unsupported' and work just fine, but this must be adequately designed to work. Cantilever does have a clean look though, and is simple to implement.

Cantilever design is not about a clean look or convenience, it really should be about proper engineering principles: not over-constraining the mechanism, while constraining it sufficiently. Most 3d printers fail at this, however.

Quote
MatthewHall

Are you using any linear motion components like smooth rods or hiwins? Or is it just leadscrews?

Oh, I do use smooth rods too

Thanks dalius, how much did the whole z assembly cost?

Quote
MatthewHall
Thanks dalius, how much did the whole z assembly cost?

Well, it will depend on the length...
I've used three 500mm/12mm rods (~8€ ea); three 500mm TR10x2 leadscrews (had to buy 2m length screw and cut as needed ~13€); 3x LMF12LUU bearings for 4$ ea... I'm also using Delrin nuts to try and minimize the wear on the leadscrews... those cost me 20$ for a pack of three... Then there are belt, pulleys and bearings for about 13$. Total is around 85$ - quite a lot now that I think of it, and I don't have anything to compare it to (first 3D printer that is not even finished yet). So, sorry I can't really comment on the performance yet.

How do you think using just three leadscrews and no smooth rods would work?

Quote
MatthewHall
How do you think using just three leadscrews and no smooth rods would work?

Badly
Even if they are really straight, you won't be able to match the precision provided by linear rails/rods... Keep in mind I am no engineer, but I'm fairly certain that this would end badly for multiple reasons.
If you want to reduce costs and/or have a small buildplate, cantilevered design might be the most suitable for you.

What you are suggesting is to use screws as guides aswell... It could be possible but you would need to design the screws so that there is no play in the nuts whatsoever. Standard 8mm diameter acme screws and brass nuts will not be up for the job. You are probably looking at using large diameter ball screws.

However it is implemented in the end, I bet it will be more expensive than just using guides + screws

I use belts with a dual smooth rod z axis. No complaints. It works just as well as screws but its easier to setup. Plus the mechanism looks great.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Demond, what is the design you used? Is it on thingiverse or something? I've been interested in the style of belts in z axis as is done on the smart rap core

Click on my things I made link above andreasl dual z axis belt alu version modified by me. Works like a charm.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

By the way of Z axis, one of the issues with two motors for two lead screws is the sync. One motor coupled to 2-3 leadscrews could sound weak (I could be wrong)
Could it help to link the 2 motors + leadscrews with a belt? This way the motors steps might have better chance to stay in sync. Just thinking loud..

Yeah I think a lot of people who have gone down the route of dual or triple leadscrews have preferred one belt around all of them with possibly a geared motor setup, or have used a second belt with a small to large drive pulley to achieve enough torque to turn the screws and improve resolution.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I have been looking at the sync issue. In a forthcoming release of RepRapFirmware for the Duet WiFi it will be solved, on the assumption that the motors will move very little when not powered. This should be true of most designs using leadscrews, but perhaps not for ballscrews. That release will also provide for the Z motors to be driven independently do bed levelling.

However, a single motor driving 3 !eadscrews via a belt is still a simpler and cheap solution, unless the bed is very heavy.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Sounds good, it's been a classic issue. Will definitely consider when building a serious machine next year.

edit: @DjDemonD : Yes having a gear to have some more torque is probably good if driving several shafts.

Edited 2 time(s). Last edit at 09/29/2016 07:11PM by orea.

Quote
filipeCampos
Cantilevers is the easy solution for small bed, but the bed structure need to be rigid.
I using a cantilevers 200x200, supported by 2 x 12mm rods and one 8mm lead screw with 2mm pitch, using this setup for almost 1 year and it works with no issue.
I know upgrading to a 310x290 with 2 x 16mm rods and 8mm lead screw with 1.5mm pitch, but this was not tested and can not guaranty it will work.
Any bigger bed and i think cantilevers is not more a option.

How about a 400x400 bed with aluminium tooling plate 8mm thick cantilevered ? works without issues on two 16mm smooth rods. The plate that holds the 4 linear bearings and the support brackets is 15mm thick. You just need beefy components and support brackets. It works actually for any size. The secret is to scale up the whole assembly, not only the bed ... Does it worth it ? Well, it was extremely easy to build and assemble, only one stepper, and costs were extremely low. Would I upgrade to two or three lead-screws ? I don't have any reason to do it at the moment, I get high quality prints with no Z artifacts. Let me know if I should post some photos.

Can you post a photo of the Z axis mechanism and closeup of a couple prints made at the far end of the bed? Cantilevering is simple mechanically which is appealing, but I have yet to see one that doesn't bounce.

I am rehabbing a CubeX Duo that has a cantilevered bed that rides on 12 mm rails (part of the machine's frame). It seems pretty rigid, but I have't printed with it yet so I'm not sure about how well the cantilever system works. I have seen some very high quality prints (made by others) from these machines. This is the cantilevered bed support from the side and below:





I'm also building a CoreXY machine and have yet to settle on a Z axis configuration for it. The two I have been considering are three screws with two linear guides and a single motor or a more conventional cantilever arrangement with a second screw at the far end of the cantilever to keep it from bouncing. I will use fully support rails- either linear guides or round rails.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
unromeo21

Quote
filipeCampos
Cantilevers is the easy solution for small bed, but the bed structure need to be rigid.
I using a cantilevers 200x200, supported by 2 x 12mm rods and one 8mm lead screw with 2mm pitch, using this setup for almost 1 year and it works with no issue.
I know upgrading to a 310x290 with 2 x 16mm rods and 8mm lead screw with 1.5mm pitch, but this was not tested and can not guaranty it will work.
Any bigger bed and i think cantilevers is not more a option.

How about a 400x400 bed with aluminium tooling plate 8mm thick cantilevered ? works without issues on two 16mm smooth rods. The plate that holds the 4 linear bearings and the support brackets is 15mm thick. You just need beefy components and support brackets. It works actually for any size. The secret is to scale up the whole assembly, not only the bed ... Does it worth it ? Well, it was extremely easy to build and assemble, only one stepper, and costs were extremely low. Would I upgrade to two or three lead-screws ? I don't have any reason to do it at the moment, I get high quality prints with no Z artifacts. Let me know if I should post some photos.

I finished to build my bed, was indicate use high quality 2 x 16mm rodds and one 8mm screw with 1.5 pitch. Was result i get tiny Z artifact (z banding), overall the print quality on the Z is ok, but far to be classified was great.
The z banding is very tiny, but noticeable in a close inspection. The problem i have was nothing to do with the rigidity of the bed or vibrations. The problem is because of the 8mm screw and the Z motor support, is easy to see the bed on each layer is not lowering the same length. There are layer that are more compress that other and i get a tiny ondulation artifact that was the same pattern of the z screw.

I totally agree with you, i still think cantilevered bed work if the components are good quality and scaled up. My error was not scale up the z axis and use a more rigid z motor mount. But i still planning to correct my error. planning to try to secure the Z screw first (remove the weight from the motor mount and motor bearings), and if is still not work them buy a Chinese 1204 ballscrew complete set.

Edited 2 time(s). Last edit at 09/30/2016 11:58AM by filipeCampos.

I wouldn't try a cantilevered design without using fully supported guide rails or linear guides. End supported rails, unless they are huge, are going to flex.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

piezo inchworm

Many different types

ultra positioning for Z-axis

---- one on each corner of bed
or perhaps attach to belt / string for you belt junkies

Expensive for now

[piezo-motor.net]

The "piezo inchworm" look great, though I don't even dare to check the price. Maybe one could look at traditional linear motors, but they might not be steppers?

Also: could linear rails replace smooth rods for the Z axis? I suppose the carrige might have a bit more angle / cantilever play than the larger round bearings though.

Edited 1 time(s). Last edit at 10/05/2016 05:06AM by orea.

The linear rail carriages I have (cheap aliexpress ones) have no play that I can detect but I haven't used them for a z axis, I am using them on a delta.

Don't discount belts they are easy, cheap, less technically demanding to install, long lasting and the resolution is plenty unless you really want to print 50 micron layers.

I was as sceptical about it as anyone until I tried it. The idea now I'd buy 2-3 relatively expensive leadscrews/ballscrews, align them, lubricate them, synchronise them when I have a really good, well supported large z axis already seems crazy.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
the_digital_dentist
Can you post a photo of the Z axis mechanism and closeup of a couple prints made at the far end of the bed? Cantilevering is simple mechanically which is appealing, but I have yet to see one that doesn't bounce.

Ok, here we go... mind the wire mess, I am still waiting for the silicone heater to be delivered from China.. I am using the heater from a pancake maker (1.4 Kw, 35 cm diameter) until then.
You recognize the print.. it's the negative tolerance test, still working on retraction and extrusion rate.

[goo.gl]

Printed at 1/4 of the edge.

[goo.gl]

Back of the Z axis:

[goo.gl]

Cantilever arms and Aluminium tooling plate:

[goo.gl]
[goo.gl]

Edited 1 time(s). Last edit at 10/07/2016 05:45AM by unromeo21.

Very beefy! It looks like a lot of mass for the motor bearings to support, or is the motor in a mount that takes the weight off the motor shaft?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]