Large format 3d printer/cnc, and linear rail HELP NEEDED

arduino CNC_Machine grbl Large_Format linear_bearing linear_rail Marlin MPCNC ramps_14 reprap

Please Login to Comment

I am in the planning stages of a potential large format 3D Printer build, and want some input. I am considering what the best budget option for the linear slides/ rails is. Currently, here are my options, and the pros/cons:

Aluminum t slot with v bearings: more expensive, possibly higher accuracy
Aluminum square tubing with 3d printed bearing blocks that clamp 608 skateboard bearings to all sides
EMT conduit with 608 bearings like the MPCNC
12mm metal rods with linear bearings
Supported linear rail with bearings

What are your thoughts?

I plan to build the machine in a gantry style that takes design aspects from cnc machines that have a stationary bed. I hope to eventually equip it with a E3D Volcano or Supervolcano hotend, and a diy Pellet extruder for recycling old prints, and printing with cheap pellets. As for a heated bed, I will likely be using multiple smaller heated beds tied together and powered through some relays. I am also considering having dual control boards with a multiplexer/ demultiplexer circuit to switch the stepper and endstop wiring between control boards. This would enable me to run a 32 bit control board for 3d printing, and have the option to switch control over to an arduino uno with a GRBL shield for cnc milling and laser cutting. I have never seen anyone do a dual control board setup before. This is the best way I know of to achieve the dual function, other than building a custom PCB, or switching firmware for different uses. Please let me know if you are aware of a better way of doing this. For now, I am working with a budget of $500 USD to get a working printer with the volcano hodtnd, and adding on features as I go.

Any ideas are appreciated as always

If you want to have a 750mm cube as the build volume, you will have a lot of issues with wobble on the gantry unless it has wide supports, which make the overall printer much larger. What were your reasons for choosing a moving gantry rather than a Core-XY, HBot or Ender-5 style?

The Ender-5 style has more belts but it is still a very simple x/y/z style movement and with duplicate belts on the Y with a common drive shaft, would keep things nice and tight for what you are doing. The simple X/Y/Z movement would allow you to swap controllers in the way that you want and frame style could easily be made more rigid if you plan to try to add milling or similar later. It's also an approach that allows for the cheaper "rubber wheels" approach now and swap out to linear rails later if required.

I had originally planned for the gantry style because a glass and aluminium heated bed would be heavy to move, but I think that with enough gearing it could be used in just the y axis. I think my new plan with have the bed moving in Y, the extruder in X, and the extruder/x axis rail moving in Z. My other end goal (I know my goals seem excessive), is to make the upright frame part (x axis and z axis) tilt 45 degrees to print on an infinite belt setup. I know that this likely has zero practical purpose other than being cool, but If I have the opportunity to have a 750mm x 750mm x infinite printer, why not?
To put that size into perspective, I could print a full size kayak for under $60USD in ABS pellets. Obviously it would be stronger if it was made in a different method, but for prototyping, something like this could be very viable. I can't imagine that there are too many people offering 3d printing services on one of the many print ordering websites that could print something that large.

Or, in a different sourcing method, if you could somehow get 500 2 liter soda bottles, you could get approximately 21Kg of PET plastic for a very low cost, and save the environment at the same time. I don't know where I would get this many bottles, but if you could buy them at the current return rate in the US, (5 cents per bottle), it would only be 25 dollars for the 20kg. These prices never cease to amaze me, because I am spending that much for only 1kg of PLA filament.

How large is "large"? There have been various comparisons of linear rail vs rod and the rods tend to compare very well. Have a look on YouTube - there are some good examples where people have swapped out rods for rails on established designs and done before/after print comparisons.

Where rails come into their own is when the weight of the printer would cause the rods to sag as the rods have to support the weight as well as act as the guide. Rails are supported so you have far fewer issues.

From what you are describing, it sounds like you will have a potentially very heavy moving mass on the head end. This will cause you lots of problems with speed and accuracy. Are you sure this is the route you want to take, as the general approach is to minimise the mass of the head end as much as possible. If the head end gets very heavy, you may want to consider a threaded rod rather than belt approach. That comes with its own problems (which people will tell you about in no uncertain terms ;-) ) but they will handle a heavy head end much better. Expect threaded rods rather than belts to be much noisier and slower...

If you know all the above then my apologies... but please share your progress as it will be an interesting build!

Thanks for the suggestions! I will likely be using threaded rod for the z axis, and dual steel corded 12mm wide GT2 belts for X and Y. If the steppers cannot move the axis with the belts, I have a bunch of large GT2 pulley for making belt gear ratios that I can use to increase torque. as for linear rails, I am going to ask another comment above on my original message.

Comment has been deleted

Take a look at my 2 machines:

I opted for aluminum extrusion frames with MGN12 Linear Rails/Guides.
Strong, stable, accurate.
Not all that expensive, as I got cheapo stuff from Amazon.
It all works great.
I'l do the same for my next build, whenever that is. :-)

Them's my .02¢
Your mileage may vary.

Maker Mods

Yes I had considered that rail system, and they are what I used in my last printer build. The difficult part is that it is surprisingly hard to get longer lengths, without paying a lot. Thanks for the input!

Anyone who has to ask this kind of questions on such a forum has NEVER achieved ANY kind of properly working machine !
You obviously have no clue in the required fields of expertise. The allocated budget is ridicule and further prove my point.

This printer would be my 8th printer build. All of my previous builds are successful printers with no issues. I fully know what I am getting into, and have priced out the starter build of a printer frame, with a volcano hotend as well under my $500 budget.

Don't use two boards, isolating them would be a nightmare and far too esoteric for a school boy. Repetier supports printing and CNC. Don't use RAMPS, you'll end up pissing away that money and then buying a real board. The cost for a CNC spindle is around $100 for even a cheap POS Amazon one and that's not including the speed controller or any machining related equipment such as gauges or bits.

I'd settle for a large form printer and upgrades over time.

You clearly did not read my post above. I am starting by building the frame, and setting it up for basic printing with a volcano hotend. That portion is what the budget is for. I am using Ramps 1.4 for testing purposes only because it is what I happen to already have. Once I have the basis of the printer, I will then be switching over to a 32 bit controller, and adding in the GRBL controlled Arduino Uno and stepper shield. Then I will add in the pellet extruder, spindle, and laser. I already have a spindle from a scrapped Prolight cnc mill that a local university auctioned off. As for your rude comment about me being a "school boy" that doesn't know how to isolate the two boards and switch the stepper connections, this is something that I have already done on a previous reprap that I wanted to add a laser cutter to.

I am quite frankly irritated by the fact that you and the other commenter MKSA above feel the need to criticize my intentions. Your are obviously missing the point of the RepRap community, which was formed to push the limits on the machines that can be made on a budget.

"I am a high school student interested in building"
It's rude to call you a school boy when you proclaim yourself to be a school boy???

"I thought the whole point of the RepRap community was to push the limits on what machines can be made, without spending a ton of money."
You missed the point by a wide margin. Reprap is about rapidly self replicating machines, hence the 'rep' part of the term.

Your solution to swap connectors between boards is... frankly very dumb when it can be done on one board and done in software.

I'll say, you have high in the sky dreams, little knowledge and much hubris about you. Take a step back for a second and realize that you're not inventing the wheel here; you're renting a wheel and calling it a new invention.

I believe that the part you are missing is that I was not referring to what a RepRap is, but rather what the community now tends to focus on. Once again, if you read my comments, I am using the dual boards because it is what I already have on hand. I will not be physically switching the connectors, but rather using a Marlin extension program that uses a circuit to switch which board is feeding inputs to the steppers. Once again, this is a system that I have already made previously, so it makes sense to use it now. Also, I was only offended by the school boy comment because you seem to assume that just because I am young, I don't know what I am doing. I know that what I am doing is nothing new, and that there may be better ways to do it. I am simply doing it this way because this is the equipment that I already have. Other than that, I appreciate your suggestion of Repetier, which I may switch to later on when I move past the testing stages and switch to a 32 bit control board.

If you've already got a system that doesn't involve switching the pins, awesome. I also thought of traditional relays or SSRs. You can pick up 12v automotive relays on the cheap from junk yards, just make sure you can test them there and don't forget the flyback diodes. A 9v once across the right terminals once shouldn't pop them right off. There are also known footprints and CAD models for them. A good choice all around except for the constant current draw one way.

I didn't mean to make it sound like an insult and thought I'd point out some things to watch out for. I had something better typed out but I refreshed and lost it. Repetier has an online config tool with diagrams, it's pretty sweet. Have you looked into the hyper cube?

Yes, I considered the Hypercube, or the hypercube infinite made by Zechyc (see below). I think that to start out, I will use the dual boards and multiplexer/demultiplexer circuit to switch the steppers. Later on, when I have a functioning frame, and motion components, I will switch to a better single controller. The current way that I am doing the stepper switching uses a 2 position toggle switch that outputs a binary 0 or 1. Then I then have a circuit that uses multiplexers to take the 0 or 1 input, and switch the path of the stepper connection. In testing it works pretty good, but I will have to see if that holds true for the printer build. The only side effect of this design is that the circuit has an approximately 3 second propagation delay from the logic circuit used, so the steppers do everything 3 seconds behind what the control board thinks it is doing. In tests, there is no motion distortion, only a delay, which doesn't really impact performance. It is just a little strange that if you use the display on the control board to move an axis, the movement comes 3 seconds after you input the command.


Hypercube Infinite Belt Printer version 2
by Zechyc

Never tried using ball bearings on extrusions.. the tolerances of the extrusions will be important to know BEFORE buying material. I used 12mm rods on my large 450x450mm bed printer. I used 4 cheap MK2 heat bed boards bonded to a large 3mm thick aluminum sheet, with 0.5mm thick PEI film bonded to the top.. backside is insulated with 2 layers of cheap foam board bonded on too. I found a $15 used HP Server power supply on Amazon to power those boards (12v and up to 60 amps).. thing heats up faster than any printer I have. print bed is light enough for a single high torque NEMA17 motor... and I use only steel corded GT2 belts.
One lesson learned with E3D Volcano... my is 12v .. does not have enough power to extrude material fast enough to use a nozzle bigger than 0.8mm with 0.3mm layer height. If ya want to print material really fast, just a 30 or 40W heater cartridge isn't going to cut it.. probably want twice that. I was going to make a heater block with two heater cartridges in it.
Also, IMO.. bed leveling is a must, since the bed is so large... I'm using a BL-Touch sensor.
The idea of using pellets for printing is a great goal ... because material cost using fliament is a killer for many projects.. I'm printing a lot of parts that use an entire 1KG reel.. and some designs that need more than 1KG, and I can't print them .. so a way to hopper up 1-2 KG of pellets would be cool.

Thank you for being the only person in this discussion that is taking my plan seriously. I thought the whole point of the RepRap community was to push the limits on what machines can be made, without spending a ton of money. The rest of the people that have replied to this thread seem to think that I don't know what I am doing just because I am still in High School. As for the extruder, my end goal is either an E3D Supervolcano, or a custom cnc heatblock that uses Supervolcano Nozzles. For the heater cartridge, I have two 50 watt extra large cartridges from a friend's failed filament extruder. As for bed leveling, I will be using eccentric bolts to lift and stabilize the bed, and a BL-Touch for leveling. In addition, I plan to have the pellet extruder as the main extruder for larger prints in the future, and use the other normal extruder to print support material.

Two 50w cartridges should work well with 1.2mm nozzle and probably get you to 0.6mm layers. FYI - I remembered wrong... I'm running a 1mm nozzle at 0.3mm layer max ..with a 40w single cartridge .. that allows printing most large parts with a single perimeter wall, thus reducing material usage.
You could try saving money by guiding on cheap Aluminum extrusions .. if it is straight and/or you can machine the working surfaces straight, if you have access to a milling machine at school or elsewhere (I have mill and lathe .. a must have when ya get older and want lots of toys, LOL).
If you can find descent E3D clone hot ends they work pretty good and are really low cost .. worst case ya swap the nozzles for high quality real E3D nozzles (this is what I do, since I have 6 printers and spares ... but use the hardened coated E3D nozzles on all, since they last forever and the tips don't wear away like brass ones).
I understand what you are saying ... human nature is to ridicule and troll people in forums, .. instead of constructively helping each other. I understand being your age and wanting to save money
If you go the pellet extruder route ... IMO, the challenge there is how to hopper 1-2kg of material, without that heavy weight having to move with the extruder. You might consider a machine design with a stationary extruder, and a heated bed that moves in X, Y , and Z.

I was considering buying aluminum extrusion from https://8020.net/ and using T Slot rollers from OpenBuilds. As for the weight of the pellet filament, I had initially intended to have the extruder move in x, y and z, but am now planning on having the bed move in the y axis. I thought that the bed would be too heavy to move easily and accurately, but the extruder is now presenting the same difficulty. I had also considered a blower/vacuum fed pellet extruder where a bin of the material sits next to the printer, and it is sucked/blown through a tube to the extruder.