Why the moving platforms?

I started building a 'big' CNC router out of T-slot extrusion after finding a router design that I quite like: [www.cncrouterparts.com]

It's a fairly standard gantry router, but uses inexpensive cold rolled steel strip and skate bearings for the linear ways. I started designing a scaled down version that would be good for a RepStrap and light milling, but then had a look at some of the other T-slot based designs discussed here. I see they all use a moving table instead of a gantry, and appear rather limited in envelope as a result. Is there a major reason to avoid a gantry system that I'm missing? Or is everybody simply basing their axes on Mendel?

---

[haveblue.org]

Hmmm ... I think I have noticed a bit of cargo cult copying of Mendel about:

http://reprap.org/wiki/Metal_Mendel
http://reprap.org/wiki/Wooden_Mendel
http://reprap.org/wiki/Eiffel
http://reprap.org/wiki/Isaac
http://reprap.org/wiki/LaserCut_Mendel
http://reprap.org/wiki/Foton_3Dprinter

Aside from "Mendel as holy grail", people are just copying what they know works.

And moving-platform is simpler, mechanically, cabling, etc. You do chop your machine's print envelope in half ...

I think the reason we don't have folk doing up a proper gantry-type t-slot bot is because we don't actually have a proper gantry-type t-slot bot.

Happily, one of our lead gantry-type t-slot bot developers may do one up, after a bit of very gentle nudging from you and me:
http://reprap.org/wiki/Rather_Blue_T-Slot_Bot

I look forward to seeing what folk get up to.

---

Watanabe is doing up http://reprap.org/wiki/WatBot. Watanabe, how is that machine going?

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

Big machines tend to be moving head, small machines moving table and medium machines a combination of both.

The extra build area you get from moving head drops off as the machine gets smaller due to the fact the head can not move the full width of the machine (due to its own width plus the width of the gantry verticals) so there is a point where it makes more sense to move the table instead.

The size of Mendel is about the cross over point where there is not much difference. For example the x-axis is moving head and the y-axis moving table but the machine it is actually wider than it deep for the same travel. The width of the head plus the two z-axes is more than the travel of the table.

---

[www.hydraraptor.blogspot.com]

My reasons for sticking with a moving Y in XCBot is it seems logical to me to 'present' the print, when finished, out the front of the printer..
To enable cooling it, and prying it off the platform.

Al...

---

[araspitfire.blogspot.com]

My current and first Repstrap has a moving Gantry - fixed bed (I will document it one day !) but I have never used the full 250 x 500 this provides and it's very slow unfortunately.

It has been very good to experiment with as I can just bolt down whatever material I like to the bed regardless of its weight or shape, this was good when testing bed materials and experimenting with a quick-and-dirty heated bed made from an Iron.

But I don't want to scrap it and I need a faster machine so I'm changing it at the moment to have a moving bed, lighter weight and be hopefully much faster.

On a related note -
If the Extruder motors+filament feeders were suspended/mounted above the moving X/Z we could speed up X and get the nozzles much closer together for multiple material designs. - Bowden cable feed - Very much like Huxley but have more than one on Mendel.

Just mount all the heavy Extruder parts on the top of Mendel and cable feed to the hot ends.

Rich.

---

[richrap.blogspot.com]

Yes that is what Rhys did here: [blog.reprap.org]

---

[www.hydraraptor.blogspot.com]

Thanks, That is really neat. I think I'm going to try doing that on my new machine.

Does anyone know where to get 3 or 4M of PTFE tube for the Bowden cable? - I can only find very expensive reels on Ebay and GRRF has not had any in stock for months.

Is anyone an expert on Bowden extruders in this forum? I could do with some pointers.

I guess the tube needs to be quite stiff and have a 1/4" ID or will 4mm ID and 6mm OD tube be ok?

Has anyone tried 1.7mm Filament in a Bowden extruder? - Any benefit to using the smaller diameter in this method?

Sorry for hijacking, I should really have started another thread.

---

[richrap.blogspot.com]

mcmaster carr sells ptfe tubing

---

repraplogphase.blogspot.com

Thanks.

Anyone know of a UK or European source?

---

[richrap.blogspot.com]

There are countless configurations of CNC machine out there and they each have there advantages and disadvantages. I think that machine size to build area is one of the less important design considerations. If your machine was 300mm longer in one direction, it probably wouldn't put you off if it was a better machine.

The Reprap is based on relatively inaccurate parts, so stacking the Z axis on the Y axis on the X axis accumulates any errors or backlash in the mechanism so wouldn't be the best way to go. The Reprap stacks only two axis on each other. (The X axis on the Z). The Y axis is fixed to the base. This is a more accurate method. The Mantis 9 is another CNC made from inaccurate parts and this uses a similar design philosophy.

Its a matter of opinion, but probably the best configuration seen so far for FDM is the Unimaker / Fab-in-a-box setup, which has a light fast head. The advent of the Bowden extruder seems to have given this the edge at the moment for single head machines.

As for CNC milling XY table is probably the best for normal sized work. For larger work the only way to go is a gantry setup because you cannot move very large workpieces.

I mentioned in another thread today that Mendel can be easily modified to make the y axis much, much longer. I think it would be a rare situation where you would need a large x and large y. and if it was needed, i would worry about accuracy of the bed height over those large distances. However, having one axis that is longer allows for much greater design freedom.

i think it would require modified y rails that sit above the frame elements, a new frog/y-cariage and maybe a Nema 23, as well as reworking the belt (perhaps gluing the belt onto the bottom of the bed and using it as a track?)

So i would say that the moving platform is a benefit in that it can be modified to meet the needs of the designer, to some degree.

martinprice2004 Wrote:
...
>
> As for CNC milling XY table is probably the best
> for normal sized work. For larger work the only
> way to go is a gantry setup because you cannot
> move very large workpieces.

Totally untrue.
The Largest CNC milling unit I've ever seen, had a moving bed. The gantry could go up some 20-30 feet, the bed was about 70 Feet long and 20 feet wide, and could easily hold parts weighing 100,000 Lbs. (50 tons)

It would be hard to have the toolhead have the Y travel needed for parts that size, so instead it's done by a travelling platform built into the foundation, with a gantry overhead for X and Z movements, just like a Mendel.

I have been working at a machine shop for nearly two years now. Though its not a long time, I sure have gain a tremendous amount of knowledge and skill. From not knowing a thing about machining to just recently (being the first at my shop) to write programs for 5 axis machine.

The X on Y table design may reduce working area on 'normal' size machines that you would see at your local shop.
The reason is for this (at least for metal machining) is that you want an extremly rigid z axis because of the heavy milling.
A moving gantry design, is possibly prime choice for the do-it-yourselfer because of the increased work footprint allowing for a smaller overall machine size. However, you are sacrificing accuracy and rigidity.
For example: If your ball screw is a class C7 which is 50 micrometers (.0019") per 300mm (just over 11")tolerance. And if the screw is mounted under your working table to a cross member to the gantry (in most cases). 'And' then you have maybe a tiny bit of slop in your gantry bearings. The distance from the screw to the top of the gantry (center of the mounting location of the z axis) is quite large. The amplitude on that is going to be in an educated guess about .010" out roughly just to give you an idea.

Now it may not seem like a whole lot but it depends on your applications.

Personally I chose (currently in the designing stage) I am going with rail guides which are a more pricy but more robust and accurate. Solid gantry with the X moving on the gantry and the table in the Y axis.

This gives me less working area on the Y but equal on the X compared to moving gantry type. But I do gain more rigidity on the gantry while eliminating any torsion or binding under load if I had a moving gantry.

Hope this gives you a 'feel' of why the different designs.

xclusive585 Wrote:
-------------------------------------------------------
>
> Totally untrue.
> The Largest CNC milling unit I've ever seen, had a
> moving bed. The gantry could go up some 20-30
> feet, the bed was about 70 Feet long and 20 feet
> wide, and could easily hold parts weighing 100,000
> Lbs. (50 tons)
>

I was curious as to this machine you have seen. I googled "Worlds largest milling machine" and it came up with the following link.

Worlds largest milling machine

This appears to be only 60ft x 12 ft and has a fixed bed and double gantry setup, not a moving bed machine which your machine has. I personally am not convinced that this is the largest ever made of the fixed bed type however I can't imagine a moving bed milling machine being much larger.

Your machine appears somewhat larger than the worlds largest (A moving bed machine is normally twice as long as the bed, so your machine would be about 170ft long and 40ft wide.

Could I ask where this machine is and what components that are 70ft long it machines.

If this does exist, could you contact the Guiness Book of World Records.

Edited 2 time(s). Last edit at 02/22/2013 07:53AM by martinprice2004.

I think having seperate axis' is stupid, when you have both axis' together then its always accelerating the same part around, and that makes the max speed of each axis nearly the same(depending on the configuration), and eliminates variables from . look at the ultimaker, its designed for speed, it has no motors moving them selves around, and is symetrical. the makerbot on the otherhand has a moving platform which makes sense because the hot end actually weighs more than the build plate with whatever is being printed. if you take into account for milling metals or heavier materials it only makes sense to have a moving gantry because it always weighs the same and therefore has one less variable to account for.

Separate axes makes sense because an extruder with onboard NEMA17 motor weighs about the same as the build platform, so you can split the moving mass in about half and half between the axis. If you are willing to make the compromises necessary to move the motor off the extruder, then it of course makes sense to arrange the printer like Ultimaker, Tantillus or Rostock, with print bed that doesn't move in XY. But with offboard motor you can never achieve quite the same accuracy of extruder output than with an onboard motor.

ttsalo has is correct,
You need to keep the mass of the moving parts as light as possible

I have a solid metal frame CNC gantry router and that design is fine for that purpose
I tested an extruder on my CNC before building a dedicated printer
The test print I did took 1.45 minutes on the CNC router

I just built a 3D printer with separate X and Y axis and very light moving parts
The same test print took 35 minutes on the 3d printer

It moves approx. 3+ times faster than my CNC router can
The mass is just to much with a moving gantry type design

For the same printed object would you rather have a 30 minute print job or a 2 hour print job?

---

My projects are listed here - [www.el34world.com]

I've been working with a fixed-bed printer for the last while; first itteration is called "Tiny" as it's a 700mm x 400mm x 250mm print area. I have a CoreXY cable set up so moving the steppers around hasn't been an issue at all. As for print capability and quality, I get fantastic results. My maximum speed is 580mm/s unloaded or a little over 500mm/s with a single extruder. I have printed up to 300mm/s on Tiny.

I am gearing up to start selling my own 3d printer called the TS300 and TS1000 which are larger-scale printers and, honestly, I love the fixed bed. It allows for simplified wiring as the bed heater wires never move. I have three sets of wires, similar to most other machines, one that runs to my bed and frame, one that runs to my gantry and one that runs to the hotend.

In my opinion, the main reason why printers move the bed is because it works, others have proven it.

You can find out info about my own stuff at tesseractindustrial.com or ottersoft.ca (personal blog where I've been documenting stuff).

Cool looking printer Otter!

---

My projects are listed here - [www.el34world.com]