Building a Rapid Prototyper

Hey Folks,

I currently have a CNC machine that I use primarily for making printed circuit boards. It has a cutting area of approximately 12" x 18" x 3". I am very much interested in building or buying an extruder head and heater bed so that I could use my machine for both CNC and plastic extrusion. My machine is driven right now by a 3axis CNC board with a controlled relay for the spindle, but I think it would not be too difficult for me to build an addon board that would allow the computer to control the heaters and extrusion motor as well. I currently use Mach3, but I've been thinking about trying out EMC2.

Looking on the RepRap site, I see numerous different machines and extruders; how do I know which one I should try and get? Is there a de-facto standard extruder that is best? Are some extruders better at different sized components? I'd mostly like to try making some small gear mechanisms, or plastic frames/enclosures for robotics or other gadgets that I'd like to build.

Also, I'd be interested in a detailed overview of the exact process of how the rep-rap works.
I understand that it takes a coil of plastic, heats up the end and extrudes melted plastic onto the work surface, but why does the work surface need to be heated? Is this to prevent it from sticking? What temperatures are involved and how are these typically controlled? I saw somewhere that the bed was heated to only ~45C and the extruder was set to ~205C when extruding. If you get too hot does the plastic become runny? Too cold and its sticky/chunky? Does the extruder filament burn out frequently, and if so, roughly how much extrusion can I expect to get done before I need to replace it, are they expensive? What is a good source for the plastic filaments of various colour, is there a common size? I live in Canada.

Still very new to the rapid prototyping scene, but it sounds like fun and I think with my CNC and a RepRap there is very little I could not create at home. The other option that I've been considering is a cutting laser, but these seem overly expensive and dangerous compared to something like RepRap. Anyways, all advice is appreciated.

Cheers.

Basically exactly what I am doing. I made desktop mill, could run with EMC2.
Now converting to a repstrap. Mind you, it's still in progress, I have just finished ordering everything - maybe the recipe doesn't work =)

Here is what needed.
Electronics -
1. Controllers (See wiki). I am using Arduino Mega (the RAMPS).
2. Fan out board to connect to your existing stepper driver (just DIY one, mostly connecting the wires)
3. MOSFETs circuit for driving the heated bed and extruder heat block.
4. One more stepper driver axis for the extruder. (I have a L298 driver laying around, I don't have L297 though...going to hack the firmware)
5. Some resistors, connection headers, wiring for the new extruder block, connecting to the thermistor, etc
Machanic -
1. Heated bed - Precisely the heating element. I ordered a locally fab'd MK1.
2. Mounting for the heat bed. I am having the workshop making a few PEEK support, so that I could mount on top of the milling bed. PTFE will do. I bet metal support will do the job too, as long as you don't make a big fat one to conduct all the heat away.
3. A glass (2~3mm), clamped on top of the heating element. (Cheap and superb flat thing for the bed)
4. Extruder + Hot end. AFAIK the current de-facto standard is J Head hot end, plus Wade's Geared Extruder. One could made it with a lathe, which I don't. For the extruder, source one from another reprapper. (Did I mention of course you will need motor, thermistor/thermocouple, heating resistor for the extruder?)
5. Mounting bracket for the extruder, replacing the spindle. As I don't have any scrap metal I just made the workshop do that for me too.

That's pretty much all the additional hardware needed.

I won't be interfacing the machine with Mach3 or EMC2, but with the Arduino Mega. From the prospective of executing the GCode path, Mach3/EMC2 is more superior, but to power the reprap, it needs to - control the temperature of the extruder, and one additional axis which isn't quite an ordinary linear axis, plus understanding the special reprap GCode (heater control). I and some other people have tried to hack the EMC2, but it was long time ago and my extruder failed before I got any remarkable result. If you want to start printing ASAP, an independent microcontroller is the way to go.

Interesting. I was thinking of jus t doing all the electronics myself, maybe even the heater bed. It is, after all, just a precisely sized copper trace routed around a large flat area. Shouldn't be tough to make. As for the heater controls, I was thinking just a simple microcontroller with a couple buttons for setting the desired bed/extruder temp. Then I should be able to connect it to an output on the driver board I have now so that it just turns on/off the extruder at appropriate times. The gcode may have to be modified to use a different command to turn on/off the heater, but I figure this may be simpler than trying to get it to pass the temperature directly. Or maybe I could hack into it so that it transmits the heater temperature somehow with the spindle speed options.

After looking around a bit, it may be possible for me to mill some stuff for the extruder myself, and then I just need to worry about getting the hot end. Does it take a lot of force to extrude the plastic? I have a stepper and driver kicking around from an old scanner that I might like to use, but it may not be powerful enough. Maybe I'll just have to buy an NEMA17 stepper like everybody else.

For the source of materila, we and even commercial reprap variants are using 3mm or 1.75mm filament, usually ABS or PLA.
There is a list here you might want to check it out: [reprap.org]

Temperature: We want to limit the hot zone as small as possible. Running too hot will resulted in a long meltzone, which will grab the wall of the extruder and get stuck. Contrary to what you think, those plastic in liquid state is more sticky and troublesome. Especially for PLA, usually heatsink, fan is need to keep the top of the extruder cool.

The heated surface is to prevent the printout from wrapping during print. It could somewhat make the printout sticks to the bed during the process too.

Heating of the extruder is done with the resistor. It doesn't fail easily. I have heard that the gear of the extruder would be the first thing to fail due to wear, after few hundreds to thousand hours of printing. The heating resistor itself is actually very cheap - maybe 50 cents to 1 USD if I recall correctly? The shipping and handling is expensive.

So what are the characteristics of each type of plastic filament? I assume that the thinner filaments are a bit easier to melt and can be used to make smaller parts, but what is the main difference to PLA and ABS? Why would I use one rather than the other?

As for the hot end, I've also seen they often seem to be encased in what I assume is a thermally conductive (or perhaps insulating??) putty/epoxy. Any idea where I can get some of that ?

Ironically, as part of my masters thesis, I just finished designing and building a heater controller for a 600W industrial heating element that uses PID control and a thermocouple to heat up an aluminum plate inside a vacuum chamber. It was able to set a temperature from ~50C up to 400C and maintain the temperature within about 0.5C. So I think I could probably use the same design almost without change to build a controller for the hot end and heating bed. Just have to add some power mosfets to source a couple amps through the heaters. Sounds like a really fun project.

I think the putty thing you are refering is fire cement or JB-Weld, once a popular design a year or two ago, but today most modern design just goes with heater block. Because fire cement and JB-Weld are not really design for 24/7 continuous heat at 250C, they fail easily.
Have a look at this: [reprap.org]

Manually tuning the heater and pause/resume the GCode might work, but I guess that's too much work. Printing usually takes hours to finish...And initially it will also takes hours to calibrate - different temperature, different speed, etc.
There are so many things in the chain, I think it's better to focus on just getting one thing to work, instead of spending all the energy to re-invent the wheel. That's why I am dumping EMC2/MACH3 all together to get a working thing, and for the electronics and software stack just rely on what's already proved working.

For ABS/PLA differences, I bet someone else more experienced could chime in =)

Since you already have a CNC mill, you can cut the mechanical parts of the extruder yourself like I did for my first 3D printer. I built a light CNC mill then converted it to a 3D printer using itself to make the extruder parts :
[forums.reprap.org]
[forums.reprap.org]

EMC2 can perfectly handle the extruder motor, just wire the extruder stepper driver on the A axis output and calibrate it to have 1° = 1mm of filament. With the alteration plugin in skeinforge you can modify the Gcode output to be compatible with EMC2 (replace E by A, remove the temperatures and extruder Mcodes)
But for the temperature management, you will need to do it with a dedicated device. A commercial PID controller will do the job, I personally built my own temperature controller for the extruder and the bed : [forums.reprap.org]

The extruder and bed temperatures doesn't need to change during printing, you can use a higher temperature on the first layer for better adhesion but it is optional and I get good prints using the same temperature for the entire print.

About PLA vs ABS, PLA needs a lower temperature (I set my extruder at 180°C), it warps much less, so you can print big parts, but it's more brittle and molten at relatively low temperature (60°c)

ABS need higher temperature (I use 230°C), is stronger, more flexible than PLA but it warps a lot (corner of the part are lifting) a heated bed minimizes the warping but it is still here on big parts.

So I've been playing around with my scanner head. It has a small stepper motor, larger than a floppy drive but smaller than a NEMA17. It has a small metal gear that meshes with a larger plastic gear which then meshes once more with a slightly larger plastic gear (three in total). The final output gear has a small gear on top which I was thinking about trying to use for extrusion. The only thing I would need to change is to add an idler bearing.

As a test, I removed the connector from the scanners driver board and figured out which pin was which. I then hooked it up to a PIC microcontroller and am now able to control the motion of the motor. What I was hoping that someone could tell me is what kind of force output is needed for extrusion? Right when spinning I can barely stop the large gear with my hand. Is this likely to be enough force? If you think this could work, then I will continue down this path by modifying the PIC controller to translate the step/dir signals from Mach3 into the correct phase/current signals to the motor and then adding a spring loaded bearing to see if I can feed some weed trimmer line.

Cheers.

3mm filament needs up to about 50N to feed, although it depends massively on the hot end design, plastic, speed and nozzle aperture.

Edited 2 time(s). Last edit at 07/13/2012 11:22AM by nophead.

---

[www.hydraraptor.blogspot.com]

Wow, thats quite a bit. Any idea if it would be a bit less for 1.7mm ?

50N = m * 9.8
m =~ 5.1kg = 11.24 pounds

So I guess an accurate test of my feeder then, would be to see if I can take my trimmer line and tie it to a 5Kg weight and see if it can lift it.

1.7mm is about 3 times less but needs feeding about 3 times faster for the same flow rate. (1.7/3)^2

---

[www.hydraraptor.blogspot.com]