Help! Controlling a huge 3D printer

Hey Guys. First off, thanks for any input you can provide.

I've got a fully-functioning MendelMax that's working great.... so great, in fact, that my boss wants me to try to convert an old gantry that we have into a very large 3D printer. My issue is trying to figure out how I'm going to control the thing. The gantry uses Ethernet controlled linear rails for X, Y, and Z. It seems to me that my only options are to try to convert a RAMPS board to control the Ethernet motors or to somehow convert the Ardiuno firmware into something that I can run on the computer and then control the motors via the standard Ethernet commands. Controlling the temperature and extruder motors are additional complications.

Has anyone else had a success with a project like this? Any suggestions on a best approach?

Thanks,
Kevin

Hi Kevin,

I've converted my Isel-CNC-mill into a RepStrap by replacing the original interface board by an adapter to an Arduino Mega + RAMPS 1.2 (see attached image).

Here I've simply wired the pins of the RAMPS, that normally controls the Pololu-drivers (CLOCK and DIR for XYZ) and the reference/end-switches to the corresponding pins of the backplane ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Thanks Victor. That's interesting, but it won't work for my set-up because the motors on the linear rails have the Ethernet ports directly on them. There is no controller board... everything is controlled from a computer. You've given me a thought though. I'm going to take a closer look at the motors and see if I can bypass the Ethernet port. Maybe I can figure out a way to use the pins from the RAMPS board to directly control the motors (in a fashion similar to what you did).

How's your old CNC mill working out as a 3D printer?

... it worked really good! - but this was mainly a test for exchangeable toolheads and software, as I'm using it mostly for milling and lasercutting ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

how big is your CNC VDX? and how big is ur gantry gtg252b ?

gtg252b - So how big is 'huge'? If it's CNC router table size (8' by 4' approx), you might be better off replacing the motors/electronics with something you can get under control with normal reprap hardware/software. Larger than that, I'd imagine the motors won't even be stepper motors, they'll be servo motors with encoders, with custom hardware. Without some knowledge of the comms protocol you're going to struggle. If you can get the thing moving under machine control, that's a start. You may find more help in the forums at [www.cnczone.com] to work out how to get it under control. Something like LinuxCNC (formerly EMC2) is likely to be the control software you'll be working with, which has a steep learning curve but is extensively used in machine automation.
You should be able to run a 4th axis as the extruder, and slicing programs will be able to create the gcode. But you'll want to have a decent size nozzle, or builds will take forever! Also, there will be plenty of plastic shrinkage to contend with. But something this size should be achievable: [vimeo.com]

after reading your paragraph and looking at the vimeo link at the end ... i think i am about to watch something that will blow my mind and it did ... that kind of build is something i am trying to get at ... but i am building it in a bedroom so its not going to be as big as what i think it will be LOL ... fingers crossed i am hoping to finish my machine and start testing very soon

droftarts Wrote:
-------------------------------------------------------
> gtg252b - So how big is 'huge'? If it's CNC router
> table size (8' by 4' approx), you might be better
> off replacing the motors/electronics with
> something you can get under control with normal
> reprap hardware/software. Larger than that, I'd
> imagine the motors won't even be stepper motors,
> they'll be servo motors with encoders, with custom
> hardware. Without some knowledge of the comms
> protocol you're going to struggle. If you can get
> the thing moving under machine control, that's a
> start. You may find more help in the forums at
> [www.cnczone.com] to work
> out how to get it under control. Something like
> LinuxCNC (formerly EMC2) is likely to be the
> control software you'll be working with, which has
> a steep learning curve but is extensively used in
> machine automation.
> You should be able to run a 4th axis as the
> extruder, and slicing programs will be able to
> create the gcode. But you'll want to have a decent
> size nozzle, or builds will take forever! Also,
> there will be plenty of plastic shrinkage to
> contend with. But something this size should be
> achievable: [vimeo.com]

redreprap Wrote:
-------------------------------------------------------
> how big is your CNC VDX?

... it has a working range of 550x500x90 mm (XxYxZ) with a footprint of maybe 1mx1m ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

The gantry that I'm trying to retrofit has a total volume of ~2'x2'x10", so I guess it's not really huge... definitely not when compared to the thing in that video!

I would look into Mach3, your boss can buy it if its got the functionality you need.

---

Random Precision

Kevin, the 3D printing club at the University of Washington (WOOF) converted an old plasma cutter gantry to a 3D printer over the summer. We didn't have to worry about converting electronics or anything, as we just used the original CNC movement software, so unfortunately I can't help you there. Our blog has all the work we have done on our printer, but you may have to wade through some posts to find it all. Our blog site is:

fabbersuw.blogspot.com

However, you will need to keep some things in mind before building this. A large printer will not work as well as a small printer, because at that size warping becomes a huge issue. You also need to decide what resolution you want of your print. We printed a boat that had a layer thickness of about .5 inches, and even that low resolution took about 19 hours to print. Printing at that size at the same resolution as your Mendel will result in prints taking days or even weeks to finish.

Huge printers are definitely cool, and I'm not trying to discourage you from building one, but it's important that you (and your boss) know what you are getting into. It's not going to just work, and even if it does, it will be incredibly slow comparatively. And, chances are, you will not get even close to the resolution your Mendel has.

Thanks for the input. I'm definitely concerned about warping... more so now than I was to begin with. I could probably build a large heated plate using some aluminum clad resistors mounted on an aluminum plate... the weight of the build platform wouldn't matter since it's not moving. Do you think having the build platform be heated would prevent curling at this size, or would the forces just be too large?

John, thanks for pointing me there. That might work. I see that Slic3r even has Mach3 as a 'G-code flavor' option. I still need to dig through the Mach3 documentation to make sure it will work with our motors.

Cool boat Pomeroy.

Check out this print. That's some major warping issues. When you get into the big prints then warping becomes so much of a problem that even if you get the print to stick to the bed then the print can warp so much it tears the kaptom tape (or whatever your using) right off the bed, possibly with bits of glass on it!

Fun challenge though. Try printing on quarter inch thick steel sheet with wet super glue for a base.

---

-Tom

The 'ethernet' connectors may not be ethernet but RS485 serial connectors. These can carry the step direction signal, or even the current for the individual coils. If you can get any pictures or product numbers from the controller, it may help to diagnose what you can use to run the motors.

Quote
gtg252b
Thanks for the input. I'm definitely concerned about warping... more so now than I was to begin with. I could probably build a large heated plate using some aluminum clad resistors mounted on an aluminum plate... the weight of the build platform wouldn't matter since it's not moving. Do you think having the build platform be heated would prevent curling at this size, or would the forces just be too large?

Just because the build platform isn't hooked up to motors doesn't mean that the platform won't move! WOOF extruded into chamfered pegboard to create a mechanical bond. We also put down screws every four inches along the perimeter. The force of the plastic shrinkage was so large that it ripped out the screws and warped the pegboard. The warping is really quite impressive.

Of course, we are extruding HDPE, which is notorious for having a large amount of warpage. But even the guy in the vimeo video above attaches his prints mechanically (I believe he puts a screw at all the corners of a part).

A heated build platform will certainly not hurt, but getting the platform to heat evenly over that large of an area might be difficult. It's something our group has been wanting to do but have not yet had the time.


If you do start moving out of research and into the actual building, please send me a message or post it here. It would be great to be able to collaborate and make large printers more available!

droftarts Wrote:
-------------------------------------------------------
> The 'ethernet' connectors may not be ethernet but
> RS485 serial connectors. These can carry the step
> direction signal, or even the current for the
> individual coils. If you can get any pictures or
> product numbers from the controller, it may help
> to diagnose what you can use to run the motors.

They probably aren't even RS485 serial connectors.
It's fairly common in the lower end CNC world to use RJ45 Jacks just as connectors for low voltage signals between boards. My larger CNC Mill does this, it uses RJ45 jacks between the breakout board and the stepper drivers, as long as they are wired consistently at both ends it's just a connector, they just carry the Step, Direction and Gnd wires. The Advantage is that the connectors are cheap and you can buy various lengths of wire off the shelf.

Here's the actual motors that are on the linear rails: [motion.schneider-electric.com]

It uses RS-422/485. In it's current configuration, each motor has an IP address and the control computer sends the commands over the network to move the gantry... pretty cool stuff.

Digging through the manual, I see that you can run it by ignoring the Ethernet port altogether and sending commands over Step, Enable, and Direction pins. The RAMPS schematic shows that these outputs are already on the RAMPS board, so it looks like I can control the thing directly from an Arduino/RAMPS controller. I think I'm going to have to give this a try. Thanks for the input everyone... it may take me a while to get any useful data points, but I'll be sure to update this thread with my progress.