Looking to build a large volume printer

Hi Guys,

I've been having fun with my GeeTech i3 but I want to print bigger and my friend has put me up to the challenge, he said he wants me to try to build a printer with the following print volume 1,000mm X 1,000mm X 2,000mm he wants to see if he can print a life size version of one his Daz Studio models.

With this challenge in mind these are my thoughts on how to combat the challenge.

I'll use 50mm Aluminium angle for the corners, using plywood for the sides, 16mm Construction Ply this should be straight and square. For X and Y I’m looking at using Threaded rod, for the accuracy in the head being that large a printer, and the Z will be in the build platform.

To get the speed in the head, as I’m looking to get travel speeds of about 500mm/s or faster it will be a bowden set up, but I’m brainstorming with the X and Y to use twin threaded rods on both Axis, so I can use a gearing arrangement to get the speed in the print head and with the bed I’m going to use a rack and pinion set up for the Z. I'm also going to see if I can print the majority of the Gears needed for the drive.

Now the drive arrangement I’m looking at for X and Y to drive the threaded rods, as threaded rods are less prone to jerk like belts are. I am looking at a Worm Gear configuration to keep the torque requirements low but coupled to a step down gearset to speed up the spin of the threaded rod.

With this in mind, I would like suggestions on controller electronics, as I’ll be able to use one stepper each for X and Y but with the build platform, I'll need to use two Z steppers linked with gearing and threaded rod to reach all 4 corners. I'll be putting a BL Touch as standard on the head, but I'll need to look at how to build my own aluminium build platform lined with PEI for printing on.

So i'd like people's thoughts on the challenge I've taken on. And suggestions on recommended electronics to look in to.

Talk Soon
Paul

WOW, it's full moon again?

Let's assume, you were able to build this printer, with a 1x1m heated bed and 500mm/s fast threaded rods. How long would a build of a lifesize Daz Studio model take? I guess, your friend will be old enough by then, to give it to his grandkids...Not to mention the cost for heating the bed and filament used and the risk of failure etc...

Well I'm still going to see if I can build the printer that big, but I want to see how much I can push it, and to try and get the print speed as fast as possible. I want to see when I go up in size if the limitations on speed are reduced accordingly. Cause when the printer gets larger your print speed in reality needs to go up accordingly.

So that’s why I mainly asked for help with the electronics side, as I have an idea on the mechanical side that would do the job.

And in regards to your comment about how long it will take, I’ve calculated approx 1 to 2 weeks to print. with current expected speeds of printers

And i'm designing to be an enclosed cabinet, so its not just a heated bed, but also heated cabinet, due to the increased build volume

Edited 1 time(s). Last edit at 01/02/2017 05:48AM by Celcius1.

look though the forums, every 3rd person asks how to build a over sized printer.. and we point out the obvious flaws in this idea.

Cost: grows exponentially with size.
Print time: grows exponentially with size. (Think months per print)
Warping: grows exponentially with size. (will need the be temperature controlled enclosure)
Filament: you going to need multiple rolls and a way to smoothy change rolls

Type of machine, you shouldn’t do with with an I3 type, imagine 1 meter square bed moving back and forth at speed. death trap.
I would go for a core XY type arrangement, but make the XY gantry move up and down, its a constant weight vs a heavy bed and what ever your printing needing lifted up.

NB you can't just print faster... laws of physics catch up with you, most normal sized printers can't melt the plastic fast enough to match the speed the hardware can already get to.

As for electronics any controller will do, your probably going to have to break out stepper driver and bed heaters anyways as nothing will run what you need out of the box (steppers on Z are going to be need to be hi current.)

I would definitely go for a 32bit controller. I am partial to smoothieware boards, but others here will vocally advocate Duet (because they sell them!)

Edited 2 time(s). Last edit at 01/02/2017 08:09AM by Dust.

Quote
Dust
look though the forums, every 3rd person asks how to build a over sized printer.. and we point out the obvious flaws in this idea.

Cost: grows exponentially with size.
Print time: grows exponentially with size. (Think months per print)
Warping: grows exponentially with size. (will need the be temperature controlled enclosure)
Filament: you going to need multiple rolls and a way to smoothy change rolls

Type of machine, you shouldn’t do with with an I3 type, imagine 1 meter square bed moving back and forth at speed. death trap.
I would go for a core XY type arrangement, but make the XY gantry move up and down, its a constant weight vs a heavy bed and what ever your printing needing lifted up.

NB you can't just print faster... laws of physics catch up with you, most normal sized printers can't melt the plastic fast enough to match the speed the hardware can already get to.

As for electronics any controller will do, your probably going to have to break out stepper driver and bed heaters anyways as nothing will run what you need out of the box (steppers on Z are going to be need to be hi current.)

I would definitely go for a 32bit controller. I am partial to smoothieware boards, but others here will vocally advocate Duet (because they sell them!)

Thanks for the input Dust.

Now your points with the issues raised are valid, but this is a challenge I’m happy to work thru.

Print time: now yes this is true, but I’m working on a new nozzle with a high flow rate specifically for oversized printers. To keep the speed up to a valid print time.
Warping: if you noticed the printer design inclueds a heated enclosure and a heated bed for extra stability.
Filament: I'm designing the machine to be able to have wound on to it about 20kg of Filament, and I’m also working on a design for a filament splicer along the concept of a fibre splicer I’ve used in my fibre optics work.
Machine type: If you noticed I am going with a core type, cause the weight of moving parts needs to be kept to a minimum, and this is why I’m looking at how to get threaded rod moving at high speed, as i found threaded rod is alot less prone to jerking issues.

So to summarise, I am fully aware the issues grow exponentially with size and i need to get the parts and speed to grow accordingly, but as you grow larger the feasibility of tackling these issues becomes easier if you have the right mind set.
I will also note you said to move the head, up and down, but I believe this is not feasible as with a bowden setup, that means having a bowden tube at least 3m or longer, I'd like to avoid this. Which is why I've explored the use of rack and pinion for the Z bed, as the rack setup should be able to handle the weight, and give the layer precision needed.

But once I start working on the various components, I'll provide updates. Cause I'm guessing some builders are curious.

Talk Soon
Paul

But to add, I'm happy with the input, and glad there is people here who are happy to help, the cost I understand is a factor but I have the funds spare, and I like the challenge my friend presented me with.

The easy way of getting threaded rod to move at high speed is to throw it off a cliff and fit a ballscrew...

If you think you can get a threaded rod to outperform a ballscrew, I would be pleased to see it!

Tref.