comparisons

I'm new to this world but am fascinated by the possibilities. However, I'm a bit overwhelmed by the options and choices to make. Is there a comparison chart about the features, pros and cons, capabilities of the different 3d printing solutions like reprap, netfabb and makerboot? As I understand, makerbot should be fairly similar to reprap, but I couldn't find much about reprap vs netfabb.

Hi polikire,

you try to compare apples and oranges. Reprap, makerbot and RapMan are to compare. That are 3D printers.
netfabb is a low cost software to produce the toolpath for this kind of machines.
Before you can print anything on a system you need to have a "3D printer driver" to convert 3D STL files into the toolpath to control the machine.
If you want to compare netfabb you have to compare it with skeinforge, which is an open source.
You can find more details about netfabb Engine for RapMan on [www.netfabb.com]

Having an actual chart would be a good marketing move. But us at RepRap tend to systematically avoid marketing.

I believe Mendel is somewhat cheaper than Makerbot, and has a larger build volume. Someone else may have the precise numbers, or you can quickly google them. Besides that, there are not really significant differences. Many people buy a Makerbot in order to print a Mendel, for example.

Cost-wise, remember you are welcome to build a Mendel, and then sell several sets of parts:
[forums.reprap.org]
Many of us enjoy doing this, both because it is fun, and because then the machine actually ends up paying for itself, which is nice if you are cost-sensitive.

Edited 1 time(s). Last edit at 04/19/2010 08:04PM by SebastienBailard.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

Thanks for the quick responses, I meant [www.fabathome.org] instead of netfabb, my mistake. I'm familiar with the general process. I know for example that makerbot prints only ABS, but reprap can also use PLA, but I don't know how much of a difference that makes in practice. Also, the fabathome uses a syringe tool, which they claim is better (more accurate?) than the extruder that makerbot and reprap are using. I like the fact that fabathome can print with a lot of materials.

What I'd like to see for example is the same reference design printed on different machines. To me it seems like makerbot is the easiest option, even if it not the most feature packed.

How much finetuning does a makerbot need after assembly? I'm seriously considering buying one (probably around batch 20).

Also, the fabathome uses a syringe tool, which they claim is better (more accurate?) than the extruder that makerbot and reprap are using. I like the fact that fabathome can print with a lot of materials.

I'd suggest building a Mendel and then adding on a syringe toolhead then.

I know for example that makerbot prints only ABS, but reprap can also use PLA, but I don't know how much of a difference that makes in practice.
I'll let someone else answer this and your makerbot query.

polikimre,
I took a look at the fab@home site. Their model 2 (equivalent to the Mendel here) costs over $1600 to make. And it is not clear if built it can make any of its own parts, thus failing the reprap goal self replication, resulting in reduced cost to own and exponential growth. The electronics alone is nearly $1000 because they use very expensive, specialized motors. The basic motor is very simple and cheap, a DC brushed motor with planetary gearbox and low resolution optical encoder and built in PIC based servo controller. They should be easy to design and build for $20-$30 bucks, but when bought from their supplier at $160 per motor, 5 motors needed, you have spent the cost of all the parts of a Mendel on just 5 motors! Their controller electronics are simpler, so only $160 for the box that up to 6 motors connect to, then the computer running the printer does some of the work that the reprap printers do on their internal motherboard.

The syringes might give them more precision is squirting out very small amounts of plastic, but you would have to constantly replace or refill the syringes to make a single modest sized piece. To make all the plastic parts of a Mendel might take hundreds of syringes, at a much higher cost. They do allow several syringes with different materials to be used at once, but I think repraps will soon be migrating to this capability too. The fab@home has no facility for a heated bed. In fact, the syringes look unheated too, which means the materials that can be worked with probably are limited to room temperature or only slightly above that. Working with high strength at room temperature thermoplastics like ABS and nylon are probably beyond the ability of the syringes used to dispense materials.

Mike

---

Team Open Air
Blog Team Open Air
rocket scientists think LIGHTYEARS outside the box!

Their controller electronics are simpler, so only $160 for the box that up to 6 motors connect to, then the computer running the printer does some of the work that the reprap printers do on their internal motherboard.
One flavor of the RepRap toolchain does this as well:
[reprap.org]

Thanks, this is exactly the kind of comparison I was looking for. I'm a mathematician, not a mechanical engineer. I think I'll just buy a makerbot, that way I get a solid unit but can still do some tinkering.

Edited 1 time(s). Last edit at 04/22/2010 09:09AM by polikimre.

SebastienBailard Wrote:
-------------------------------------------------------
> Their controller electronics are simpler, so only
> $160 for the box that up to 6 motors connect to,
> then the computer running the printer does some of
> the work that the reprap printers do on their
> internal motherboard.
> One flavor of the RepRap toolchain does this as
> well:
> [reprap.org]

I have not yet seen that tool chain. Do they also use the expensive motors? Any specifics on the name of the toolchain so that I can try and look it up in the Wiki? Not that I intend to use it. I am much more comfortable programming ATMegas to do my dirty work, and have designed (but not yet gotten working) my on MOSFET H-bridges for an underwater robot, so putting optical encoders on geared motors is not scary to me. However, I have not had a lot of luck with PID controllers yet.

Mike

---

Team Open Air
Blog Team Open Air
rocket scientists think LIGHTYEARS outside the box!

Thanks, this is exactly the kind of comparison I was looking for. I'm a mathematician, not a mechanical engineering. I think I'll just buy a makerbot, that way I get a solid unit but can still do some tinkering.

Mendel is solid, thank you very much and will meet your needs. We just don't advertise at CES or name our machines after confectioneries. You can buy the parts for it here if you like:
[dev.forums.reprap.org]
I warn you that assembling one may require use of an allen wrench.

---

I have not yet seen that tool chain. Do they also use the expensive motors? Any specifics on the name of the toolchain so that I can try and look it up in the Wiki?
Yes, it is at EMC RepStrap
[reprap.org]
[forums.reprap.org]

This is where people use a $1000 laptop or a $50 used white box pc as a machine controller, running emc2 [linuxcnc.org] on it. emc2 is a FLOSS industrial strength cnc controller which people use to control small benchtop mills and lathes but is intended for 'big iron'; large milling machines and the like. It came out of NIST.

The electronics are bit-banged stepper controllers or servo controllers hanging off the parallel port, although there is some usb stuff floating around.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

polikimre, you may want to see if you have a local RepRap Users Group that can help you with stuff. (Sorry if I was a little snippy just now.)

[forums.reprap.org]

-Sebastien

No problem Sebastien, but you must have a very impressive Allen wrench :-) I'm not in a hurry, I have some other DIY projects filling out my spare time, so I'll just wait and see. To see for example if that James guy is worth doing business with.

One question: I watched a few makerbot videos where they explain that reprap moves the extruder in XY and the build table in Z, while makerbot does the opposite. Any comment on the impact of this difference on, e.g., precision?

Another question (way off-topic here, I guess): is there any optimization going on to minimize the number of (sharp) turns in the generated GCode? I saw some patterns and they seemed very much zigzagging, which (I think) is not very good.

polikimre Wrote:
-------------------------------------------------------
> One question: I watched a few makerbot videos
> where they explain that reprap moves the extruder
> in XY and the build table in Z, while makerbot
> does the opposite. Any comment on the impact of
> this difference on, e.g., precision?

Darwin - Extruder in XY, and table Z.
Mendal - Extruder in XZ, and table Y.


Makerbot - Extruder in Z, and table XY.

Other combinations are possible too, including using a fixed extruder and an XYZ table, or vice versa. These tend to incur the greatest mechanical complexity because of routing electronics and motor placement, as well as greater stiffness requirements for the outermost axes.

Other than that, the differences in build area, precision does not tie directly to the axis topology. It usually is chosen based on constraints for mechanical design and build area, and designer preference.


> Another question (way off-topic here, I guess): is
> there any optimization going on to minimize the
> number of (sharp) turns in the generated GCode? I
> saw some patterns and they seemed very much
> zigzagging, which (I think) is not very good.

Not that I am aware of. Some amount of zig-zag is necessary. I could certainly see some optimization that, for example, adjust the infill angle slightly to try and align the infill more with a longer cross section if possible. Currently, I think it just alters the angle 90 degrees between layers. Structurally, it should seem possible to choose from a range of angles instead to optimize the zig-zagging, E.G, choosing the optimum among the path generated on a layer path for +70, +80, +90, +100, and +110 degrees for infill.

In terms of it's effect, the only drawback/advantage to doing so would be the extra time and/or vibration induced by lots of sharp corners.

probably not a question for here but ..

What makes the precision of the reprap...
is it the x,Y and Z movement or is it more the extruder part that makes all the difference.

I see so many differences in quality that I would like to know of what it is depending!

Is it also the slower the better or not? or is this again the extruder that makes all the difference!

mhensen Wrote:
-------------------------------------------------------
> What makes the precision of the reprap...

If by precision, you mean how accurately can the positioning system can select target points in the model space (I.E is the point at 1.003 selectable separately from the point at 1.002? If so, you have precision to at least 0.001), then only a one factor contributes for stepper base motion -- how far does a 'step' on the motor take the axis in an ideal perfect machine (by 'step', it considers a half-step to be a 'step' if your electronics have been enabled for halfstepping your motor.) Precision is adjusted by stepper resolution, gearing, and strategies such as half-stepping.

If by precision (Usually termed 'accuracy' rather than precision), you mean how accurately can the positioning system select one particular specific point, then many factors and variables will affect it: Precision, stiffness of the frame, motor speed (steppers skip have a tendency to skip more steps at higher RPM), backlash, quality and tolerances on linear bearings and guide structures, wear on parts, etc.

I believe the RepRap Darwin and Mendal design goals (and achieved, I believe) are around 0.1mm for both precision and accuracy or so, which means, it can select any 0.1mm cubic 'pixil' in model space, and put a blob of plastic about .5mm in diameter there. Some machines may have greater precision than accuracy, others may have greater accuracy than precision.

> Is it also the slower the better or not? or is
> this again the extruder that makes all the
> difference!

Generally, if you slow down your motions mechanically, using gearing, then you'll increase precision. You may or may not increase accuracy, depending on all the other factors, and the error the gearing introduces.

Quote
mhensen
is it the x,Y and Z movement or is it more the extruder part that makes all the difference. I see so many differences in quality that I would like to know of what it is depending!

Probably deserving its own thread. But anyway, all of these can make a difference, and probably the limiting factor varies from one machine to the next. I think that usually the extruder is the limiting factor. Also for any one machine there's usually a tradeoff between print quality/precision and print speed.

It might be interesting to do a full theoretical/experimental analysis of Mendel's printing errors to find ways to minimize them.