Scale up help needed

I'm looking to build a fairly large printer, I can source a 600mmx600mm heated bed, the problem I'm running into is trying to figure out which printer would be the easiest to scale up to that size. Right now I'm in favor of the Prusa i3 but I'm worried about sag across those distances. On top of that I'm having a little bit of difficulty figuring out how big the frame would need to be to accommodate a bed that size (size up math is throwing me) and coming up with a supply list, just rods the rest of the supplies are easy and the same for the base kit.
To be honest this is my first solo build, live by the go big or go home motto, and want this to be a success so please any help you people can give would be greatly appreciated.

Hi sl86,
Most 3D printer designs are scalable within a narrow margin (for example, the Prusa i3), and I am not aware of any design that scales up to a heated bed of the size that you have indicated.
In any case, with very few exceptions scaling up is not easy, it is actually a complex mechanical design problem and generally requires a redesign, in part or as a whole, of the 3D printer.
I can only suggest that you assemble a standard sized Prusa i3 and gain the experience to later try to create your own design.
One more thing: present FDM technology is just not adequate for printing large parts, for various reasons. You'll get a better understanding of these issues once you have built your first printer and printed a few kg of filament.

H-belt printers are quite easy to be scaled up.

---

Blogs:
Meine 3D Druck Abenteuer
[3dptb.blogspot.de]
FLSUN Delta Drucker für Deutschland
[flsun-deutschland.blogspot.com]
Books on 3D patents:
[goo.gl] (english)
[www.amazon.de] (deutsch)

I agree with AndrewBCN. Downscale is ok for any design. But upscale is generaly a bad idea. Physicaly, no design allow to upscale its bed size about 400% neither in the future. The prusa i3 is initialy designed for a 200x200mm bed. On the X and Y axis the 8mm rods are limit for that distance, because of their own weight. To me they flex already too much, especialy on Y. You can raise the Z with less problems as long as you can have rails and screws straight enough and a stiff frame. There's a 12" Prusa I3 working well, so the build area might be 200x200x305mm (XYZ). You won't be able to go further, or you will need to studdy a new iteration, by calculating each component to fit the new size. Appreciate, components size and price grow exponentialy with buid area expansion. I guess you can easily double the price of the i3 by increasing the print area of 30% ! OR, you will increase the size and compose with a lower quality.

A lot of people think that a 200x200x200mm build area is too small. Actualy, it's bigger than what they can think. And it overtake most of everybody's needs. A bif print area sounds good, but nobody imagine the print time, the price, the mecanical skills needed to succed into a big printer build and daily use. Do you realy need a larger print area ? Printing time grows with the part's overall size, especialy on the Z axis. Even with a larger nozzle and a lower print quality. Appreciate a good small machine can print much faster with a good repeatability that allows you to print bigger parts in smaller chunks, then assemble them. For less money, you can run multiple small machines at the same time for a faster print and more flexibility for your production needs. Big printers are needed only if 80% of your prints needs at last 80% of your build area, and if money and/or time and/or quality realy does not matter. Think about it. If you realy needs something large, is the 3D printing the tool you need ? What about a big CNC mill ? It will work much faster and will allow the use of stronger materials.

My advice is to write a specifications chart, about your printing needs. Plan on one yeat production (if you can, or a monthly production) and list the parts, their size, the printing time, the raw material expenses, and so on. If only 50% of your production needs a big machine, you're need a smaller printer. If it's to print numbers of the same part, the solution is not a big printer but several linked small ones. Dont think big until there's no other ways.

Edited 1 time(s). Last edit at 03/29/2015 05:11PM by Zavashier.

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

Thank you, the people I built with before always just used generic designs and that's why I'm no longer working with them. Someone posed this challenge to me and I couldn't help but accept. H-Belt does look like the right option, I was worried about bed stability at this size but this design could deal with that. Again thank you for your wonderful help

Quote
maboo
H-belt printers are quite easy to be scaled up.

maboo,

Just checked your blog, you have not personally designed any printer of any size and the largest printer you seem to have built that can print good quality parts is... a standard size Prusa i3!

So I am not quite sure on what you base your statement above. Probably you are just repeating something you read somewhere else... so would you care to post your sources, please?

Quote
maboo
H-belt printers are quite easy to be scaled up.

Sorry maboo, it's just wrong. Even a horizontal belt transmission is not infinitely scalable. The stretch of the belt is a percentage of its own lenght. If not supported, the belt may bend under its own weight, even a GT2. And we're talking only about the belt itself, not about pulleys required precision and alignment. For any machine's component, size and specs matters definitely.

Edited 1 time(s). Last edit at 03/30/2015 04:31AM by Zavashier.

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

Hello I am just about to start on a large print build.

I will probably spend about 300 dollars on linear guide rails and bearings for starters, sbr16 rails have a T shape that supports the rail over its entire length and fixes it to the support structure. That should reduce sag.
I will be mounting the rig inside a kitchen cabinet to give thermal and humidity stability. And the project will be called....

Majjiska läden

magic box. You can guess where the cupboard came from. The build area will as a result be quite
large, probably 400 deep by 600 wide by 400 high. The carriage will resemble that of an ultimaker
but stiffer rails and a double x and y carriage rail. The carriage should have room for a double bowden hot end
with separate extruders of course and a carriage mounted geared extruder for flexible filament.

To drive this I will be using a six stepper driver Rumba card. X, Y, 5:1 geared stepper Z, two bowden and one gearhead.

wish me luck as this will be my first build.

Majjiska Läden.

I just want to clarify that I don't expect to print 600 x 400 x 400 models. But I do wish to create some prototypes that would be best in one piece and about 350 square and flat. No printers that I can afford can do that that I am aware of.
I wish to print nylon, so I have to keep the moisture out of the filament in a box or cupboard.
Making the enclosure that I build in a kitchen cupboard is an obvious progression for me.

Many people complain about flex in their smaller printers, that is why I am spending heaps on the rails and lead screws, geared steppers etc. It should be cnc grade once I am done. I have access to laithes and mills also for the brackets etc.

I am open for suggestions though. I plan to order the electronics. Motors and rails this week.

Adam.

Quote
adam_tigger04
...
wish me luck as this will be my first build.

This being your first build, I would suggest you save some money and build a Prusa i3 for starters.

With that, you will gain experience which, in this case, is much more relevant than luck.

I'm with Andrew and Zavashier on this one - It should be made a sticky; "For anyone contemplating an extra large Reprap build..."

I don't want to knock anyone down for having dreams or ambitions but

What you would gain in knowledge from assembling and using a regular kit printer (i3 etc), I believe it would easy pay for itself against the cost of constructing anything non standard or large. The understanding of the print process will reshape your ideas of what you really want. The cost will really be 'peanuts' in comparison to something the size you suggest and you could always sell it on.

Scaling up isn't straightforward, as stated. The relationship is cubic, not squared. Think; Elephants legs versus ants legs.The bigger the printer, the bigger the support structure. The bigger the support structure - the 'even- bigger' the support structure to hold it's own weight.

If you really want a 600 cubed print area, make a cardboard box at least 800 cubed. Now find a place in your house that it can live. Permanantly. Not just when its wowwing you with huge (extremely long) prints, but when the novelty wears off a bit. Imagine it weighs at least 50kg, probably more. Will you move it? Can you build it strong enough to be able to move it without distorting it every time?

There have been similar posts about big printer builds before but most of the posters tend to disappear and never be heard from again after a short time.

Good luck. Let us know how you get on.

Alan

Quote
alan richard
There have been similar posts about big printer builds before but most of the posters tend to disappear and never be heard from again after a short time.

Yep; I also noticed

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

I agree to a certain extent but at the same time I would encourage others to experiment, do not quit your day job though ;-) and do not spend more than you can afford.

Edited 1 time(s). Last edit at 03/30/2015 02:00PM by lunarkingdom.

I think that the best option at the moment would be a h-bot or CNC style extruded aluminum gantry.

Talked to my boss, the person I'm design this for, and he said he would prefer something more along the lines of a CNC, says h-bot is a little more complicated than he wants. This is just going to be a display piece at our shop so speed isn't really a problem.

As to why people leave the forums I can't see that happening, I mean all the positive feedback and people actually reading what I've posted. So supportive.

You're getting paid for this?
How cool.

Have a look at how 3DP do it.

That arrangement may work out better for you. It is the right sort of scale for what you have in mind.

Openbuilds have just launched a new range called C Beam. That and their v slot extrusion would make for a nice sturdy assembly of that scale.

Keep us posted. Sounds like a great project.

A

I understand the concerns you have with a large build and the problems it creates like where do you put it and how do you make it stable. My oversized build is actually an answer to those problems and I don't wish to move it once built. I don't wish to make it a feature of the house. I am hiding it in a cupboard under my workbench. When researching what printer to buy or build for the last nine months I kept hitting the same troubles that some people regularly have with most printers available for a good price- the frame shakes, needs bracing, z rails lockup, z motors stall, wires break off the hot bed, they get cogging from the belt teeth, wish they had room for another extruder, fumes from pla, nylon absorbs moisture so you need to keep it in a cupboard. The cooling effects cause curling..... need to put it in a cupboard....

it did not take long before building the prusa and bracing it and hiding it in the cupboard under my bench evolved into.....
BUILD THE PRINTER IN THE CUPBOARD. This is actually a very logical step.

I had not thought of the mass of the belt being an issue. I concede that. I will get back to you on that one when I cross that. Had not read of that being a problem. And I have been reading lots of posts etc.

I am planning more than relying on luck. Figure of speach. A large number of problems I believe are solved by using the sbr16 rails on 3 axis. But I know that i will need some guidance on this ambitious project. I will post some drawings and photos soon.

If you could picture a plywood frame ultimaker, fix the x and y rails along the entire length rather than just at the ends and then make the z axis move with two screws with four rails of linear bearings, then that is what I am building.

wish me luck Adam in Brisbane

Alan- thanks I'll look into them

Adam- I am wishing you the best of luck because what your talking about does sound like a cool idea.

Regarding the large build
just for perspective on a large build I just committed to $350 worth of good rails and slide bearings alone. More to come.
This was the most affordable way of getting the precision I was after at the scale I am working at.

Thanks for then encouragement sl86

after the base is installed in location, braced and levelled and fixed, the four vertical rails will be mounted onto some flat bar brackets and shimmed to perfection.

I will then manufacture the hot bed frame from aluminium and a carriage frame holding the stepper motors, pulleys, tension blocks and the x and y rails.

I reckon I can pull this off. I am an industrial electrician/ engineer in the printing industry at the moment. I am currently upskilling to dual trade as a fitter. I often calibrate machinery to fine tolerances......

Just one more comment... think reliability. How long can a printer print without a clog or jam or whatever?

I have a lot of half-finished pieces of trash, most of which took several hours to produce before something went wrong. Some items have taken 4 or 5 attempts to print (admittedly with cheap cheap filament). This on a 100x100x100 printer.

If the probability of clogging etc is linear, then double the size of the printer and it takes 8 times as much filament, then you would get 1/8 of the number of successful prints (and take 8 times as long on each attempt). A 600x600 printer with that reliability would produce one completed object a year (and a very large heap of plastic rubbish).

Update, came up with a rough plan and and a very basic price list, over $1000(U.S.) and pointed out the potential issues with going this big to my boss, he's given me the go ahead for a full design and parts list. Btw, openbuild has been very helpful so far.

Glad you got the go ahead sl.

other issues are effecting this months budget but I am still purchasing the moving parts and controllers. A bit over 500 aus dollars.

I am going to use a Rumba board. It has six stepper outputs. Using drv8825 drivers. They have slightly higher current rating and more micro steps. I am interested in the extra current. If I need a second z axis driver I will still have capacity for two extruders we one driver per motor. I will not series them.

my z axis will use a single stepper with a 5:1 planitary reduction gearbox to drive two lead screws. This will hold my weighty x and y rails.

the other motors are 60mm nema 17s. Still small enough to run on conventional stepper drivers at under 1.8 amps.

Most other things on my build are off the shelf. I have kept It that way so I can service it easily or others can repeat it if they wish. If I was building any larger than mine I would probably use an aluminium extruded frame.

good luck sl. Keep us all posted.

Adam.

Good luck sl86. I will stop hijacking your thread and start my own.

my rails just arrived. Thought you might like to see my expensive solution to many of the upscaling problems that I could conceive, especially sag in the rods supported only at either end. These rails have that T shaped extrusion attached the full length and are really solid. May be a bit noisy. Bearings have very little backlash in them.

good luck.

Adam.

Hello
I had made a scaled up prusa i3 recently, and experience layer offsets in my x axis (the 400mm) axis when the print reaches the height of about 1-2cm. I am sure its not for under or over currenting and there is no problem with my stepper drivers and motors. The belt on x axis vibrates like a guitar string, and I believe thats the cuase of the problem. As I used GT2 belts, is there any other alternative which can handle the 400mm axis?

Quote
adam_tigger04
Good luck sl86. I will stop hijacking your thread and start my own.

my rails just arrived. Thought you might like to see my expensive solution to many of the upscaling problems that I could conceive, especially sag in the rods supported only at either end. These rails have that T shaped extrusion attached the full length and are really solid. May be a bit noisy. Bearings have very little backlash in them.

good luck.

Adam.

I started with zero knowledge of 3D printing and scaled up a machine and it ultimately came out better than I could have ever imagined. Keep doing what you're doing and don't let the nay-sayers get to you. You'll make some mistakes along the way but you'll figure them out and solve the problems one by one and in the end you'll have the printer you wanted. The main thing you need is persistence. You may also need a little money to throw at the problems every now and then. Your first step of lurking in the forums and reading about everyone else's problems was a good one. You can't improve things if you don't understand/anticipate the problems. Keep going!

Fully supported rails is a good start. If slop in the bearings becomes a problem, look into linear guides like this: [www.ebay.com]