Looking for input as I learn about and design my first 3D printer

Hey guys I've been looking into this seriously now for a few weeks. I've decided that to some degree I want to go the DIY route. I'm not looking for anyone to tell me every little detail up front but rather I'm looking for input and answers as I research. So far I think I've settled on using a Smoothieboard for the electronics and their firmware because I have an interest in laser cutting and CNC machining. I also think I've settled on a CoreXY setup using fishing line. I'm not set on a build volume but I've been thinking of something around a cubic foot for now. I'm planning on building this out of aluminum extrusions and right now I'm trying to figure out the bearings and rails. I have looked into LMXuu bearings and hardened rods but I'm trying to figure out more about using a plastic on the extrusion itself. I'm not trying to design something capable of doing CNC work yet but I do want to design this for speed and accuracy. Less noise would be nice also.

I anyone wants to go over things I've settled on I would appreciate it or if anyone wants to help me with the linear motion stuff I'm researching now. I haven't gotten into the extruder, motors, heat bed or other things yet. I do know I want to keep the extruder motor stationary to keep up with the CoreXY principle.

So far I'm looking into UHMW-PE, PAI and Nylon. I know from what I've read there is concern about the wear creating slop but I'm thinking that shouldn't be a problem with a self adjusting design. I know this is more complex but I'm getting the impression that one or more plastics may have more ideal characteristics than ball bearings and rods. I'm still trying to figure out if there is a plastic on plastic combination that is good but that is for down the road. Anyway thanks for your time and any input you guys might provide.

Hello, I am in a similar situation. I have started building a 3D printer from scratch and learning and making mistakes as I go along but enjoying every minute of the process. I have a post :

'My DIY 3D printer approach' which may be of interest and which is probably a couple of pages down at the moment. The main problem that I envisage at the moment, is weight of the Y motion.

I am also interested in the possibility of laser cutting in the future and am interested in the board that you mention in relation to this. I will initially be using arduino / ramps.

I've just started this journey. I had a box full of stepper motors found in the trash at a former job, but they were mostly small NEMA17's (about half the length of the ones I see in reprap photos), so probably too small. Yesterday I found some NEMA23 steppers at a hamfest, these are 2.7V @ 1.8A nameplate ratings with 3/8" long 1/4" dia knurled shafts. Hopefully I can find pulleys and/or shaft couplers to fit. From what I've read if I use two in the Z axis they need to be wired in series.

I want to build something cheaply as my budget is limited. I was looking at the wolfstrap, using drawer slides from Homedepot. The ball bearing types actually seem to operate quite smoothly but seem to have two issues. 1: the useful travel limit is one half the length of the slide, beyond that you get flex from unsupported overhang. This isn't a problem in the Z axis where the weight is hanging on the drive screw and the slides only guide the movement. 2: the slide has a redundant part that sticks way out beyond the carriage movement which increases the size of the unit. To get an 8x8 build area with an XY bed movement will require the printer be 24"x24" in size. I can halve that by making the X axis move the extruder, though the gantry will still sitck out beyond the Y axis bed.

I can get steel rod from homedepot, but only in "inch" sizes which don't play well with readily available (and cheap!) metric bearings. Also the depot ones have to be checked for not being bent! So now I'm wondering about using roller skate bearings (608's) to create homemade linear bearings to work with the fractional inch rod stock.

Clearly I have I lot of things to ponder as far as mechanical choices for a homebrew job. I have several atmega1284P development boards available, so building my own carrier for the stepper drivers might be the way to go. I've downloaded several of the available firmwares and still have to figure out how to set the configuration options for a "wolf strap" or other scrap built machine.

Thanks guys for your response. So far a low moisture absorbing nylon looks right. My design at this point is to have a completely stationary bed with a corexy that moves in the Z axis. The reason being that you start low and move up so that the center of mass starts low. Other corexy designs have the xy stationary and the bed moving in the Z plane but that means that all that weight starts up at the highest point. Anyway I'll keep things updated as I progress.

The type of design that moves the corexy will have the smallest footprint on the desk, which is nice.
However, this kind of design is much more complex mechanically.
I'm going with the more typical approach where the bed moves in the Y axis and having a "gantry" moving up and down in the Z axis. The gantry IS the X axis.
(the other variant is to have the gantry fixed and the Z axis rides along the X axis of the gantry).
I'm going to use drawer slides for my linear bearings, like the wolfstrap, but I'm going for an all metal build with steel or aluminum "L" girders and threaded rod.
I wonder if someone could print me up two of these: [www.thingiverse.com] to drive my Z axis? I emailed my daughter at BU, she might have access to a 3D printer there. (Typical chicken and egg problem!)

kscharf, for a Z axis driver solid enough to get you up and printing to the point you can print your own preferred coupler, take a length of fairly thick nylon tubing, put it in some hot water to soften, then thread about just under half of it onto your threaded rod. now wedge a small ball bearing ball about the same diameter as your threaded rod for a pivot, then slide the assembly over your NEMA 17 motor shaft and tighten a few zip ties down on either side (prefferably using a zip tie tensioner) You want the tube to be small enough that it is in contact with the nema motor shaft, but big enough you can still get it over your threaded rod. Some people swear by (or at) that particular setup. Either way it will be accurate and strong enough to get you printing and you can decide to keep or replace them from there. plus it's like $2 worth of tubing

I've actually experimented with this idea. I also wrapped some thin brass shim stock around the tubing and then clamped it to the shafts with the smallest hose clamps I could find. I need to connect a 1/4" shaft to a 5/16" rod. It also has to handle the down weight without pulling out. Might work with the addition of the hose clamps.

I have another question. Was thinking about building a bed heater. I found 10 pieces of 0.1 ohm 25 watt metal cased power resistors (the kind you can bolt to a heat sink) in the junk box. Connected in series (1 ohm) they would pull 12 amps from a 12 volt supply for 144 watts of power. This is more than the RAMPS board can control on board (unless I jumper out the fuse or replace it with a REAL fuse at a higher current.). How does the firmware regulate the bed temperature? Is it just on/off or does it use PWM? If it is just on/off I can simply let the board control a relay. If it is PWM I would need to use an external power fet or put a heat sink on the one on the board and replace the fuse.

Would this work?

So are there any suggestions regarding a good plastic to replace linear bearings?

Quote
MeltManBob
So are there any suggestions regarding a good plastic to replace linear bearings?

I'm designing a corexy variant of the Mondrian using fishing line. It going to use adjustable, printable x and y carriages to help with slop created from wear. The carriages also run on aluminum extrusion.

Check out the thread. I just in the beginning stages and still have to do some modeling and testing.

[forums.reprap.org]

---

greghoge.com

HUGE 3D PRINTER PARTS SALE!!!

gmh39 - thanks for the link.

Another question that's popped up and I can't seem to find an answer to is how much the Z axis needs to move in terms of acceleration and changing directions. I'm assuming that it just needs small steps for good resolution but doesn't need to move fast like the X and Y axis, is this correct or will there be a lot of small up and down movements?

You. Small steps. You can probably use the same kind of stepper for the z as the x and y. Its controlled by your electronics and firmware so as long as you have it calibrated correctly it should work fine

---

greghoge.com

HUGE 3D PRINTER PARTS SALE!!!

These motors are what I'm looking at, they probably are overkill. I've tried to read up on motors and I'm still not clear on picking them. These have an inductance of 16.2mh which from what I've read would have a high ideal operating voltage and the site doesn't show any torque curves. I'm actually thinking that it might be better to not worry about buying motors that are capable of light cnc work down the road because it would render my 3d printer useless while the motors were in the cnc setup. So really I just need to size them large enough to handle the build volume I'd like for 3d printing or laser cutting.

These are what I ended up with:
[www.mpja.com]

He had them for $5 each at a hamfest last weekend, but $7.95 isn't bad either.

Do you have any informaton on them?

MPJA didn't have too much data on them.
1.8 degree, 2.7 volts 1.8A 1.5 ohm coil 4 wire bipolar.

shaft is 1/4" 3/8" long knurled. Will take a pulley or a compression clamp type coupler.

Quote
kscharf
I have another question. Was thinking about building a bed heater. I found 10 pieces of 0.1 ohm 25 watt metal cased power resistors (the kind you can bolt to a heat sink) in the junk box. Connected in series (1 ohm) they would pull 12 amps from a 12 volt supply for 144 watts of power. This is more than the RAMPS board can control on board (unless I jumper out the fuse or replace it with a REAL fuse at a higher current.). How does the firmware regulate the bed temperature? Is it just on/off or does it use PWM? If it is just on/off I can simply let the board control a relay. If it is PWM I would need to use an external power fet or put a heat sink on the one on the board and replace the fuse.

Would this work?

Use a solidstate relay to switch the power to the heated bed. The electronics that normally control the heated bed is used to control the solidstate relay. In this way you are drawing very little power from the electronics and can switch high currents of power for the heated bed.
If you want to use mains power you need a DC-AC solidstate relay, if using a DC power supply that is remote from your electronics supply (secondary supply) use a DC-DC solidstate relay. Both can be found on ebay, they are available in several different amperage handling levels, get a larger one to be safe (same physical size)

---

[regpye.com.au]
"Experience is the mother of all knowledge." --Leonardo da Vinci

That's a good idea. I can run the heater on AC off of a high power filiament transformer left over from a HAM linear amp (10V @10A) and switch the primary with an AC solid state relay I also have in the junk box.