Machine Shop Reprap - Advice?

So, I'm interested in building a RepRap after seeing several working at the Bay Area Maker Faire two weeks ago and am looking for some guidance.

My goal is to be able to make smaller (say <3" on any side) primarily mechanical parts like gears in a relatively reliable fashion. Speed is less important to me than precision, repeatability, and finish quality. The best RepRap-made parts I've seen are good enough for my uses, I would just like to maximize my odds of getting good output each run. Budget-wise less is better, but I'm ok with spending up to around $1000USD on this.

Given those goals, I am wondering whether there are areas where I should or could modify the Mendel/Huxley design. I have a CNC lathe and mill and supporting machinery, so I can manufacture mechanical parts pretty easily and cheaply. Some of my thoughts included:

- Making all the structural parts, brackets etc. out of aluminum or steel
- Using "real" linear motion bearings/bushings
- Using screws rather than belts

Anyway, I guess my question is, are any of these worth doing, or should I focus my resources on other parts of the machine, or should I just get a kit and build a stock Huxley?

Check the reprap wiki, thingiverse and the rest of this forum before you go about redesigning to make sure that no one has done it before. It could save you a lot of time. With the size of the parts that you want, any of the 3D printer kits or cheap plug-and-play like the Up! would be fine. The advantage if the bigger build platform on a Mendel versus the Huxley, U! or Makerbots is that you can print multiple parts at the same time.

Making all structural parts out of metal is probably not going to make much difference. But making all the gears out of metal might.

I would think that screws would be slower than belts. So, I don't think this is a good choice for the X and Y axis. Even if speed is not as important to you, you surely don't want to 10x longer just to print a part. The process is already slow as it is now.

In my experience with my current Mendel set-up with Adrian's geared extruder, the extruder is where one could make significant impact: specifically controlling the start and stop of extrusion to avoid ooze and blobs.

I posted another thread about this in the general forum which got more responses. Some people seemed to think that there would be some benefit to a more rigid frame like a T-slot extrusion approach. I'm not looking to re-invent anything for the sake of originality, but I don't have a problem inventing for a purpose. The Mendel is designed to maximize use of printed parts, which I personally don't care much about. My purpose is to maximize print quality, so I'm willing to compromise on requiring more precision-machined parts.

I have a set of NEMA23 steppers that can spin over 1000RPM on just 24V, so I should be able to get into the low hundreds of inches per minute pretty easily. Belts might be a good deal faster, but how fast is fast enough? I know 500IPM is not fast for a laser used for engraving, while for a milling machine that is very fast. I don't know what a reasonable feed speed for an extruder is.

Either way the point about focusing work on the extruder is taken. One or two people in my other thread said that was probably the area most worth focusing on, which makes sense. A machine tool is only as good as its spindle. Another person said that the firmware was also a source of trouble. I run EMC2 on my machine tools so I might try it on this first, since I am pretty comfortable with it.

Part of this too is that it might be easier/cheaper for me to make a machine from T-slot rails and machined parts. I have a mill and lathe so making parts that way is easy, but a lot of the Mendel parts are not easy to make on a conventional 3-axis machine tool. I've built 4 different CNC machines at this point so building a 3-axis system is not something that scares me anymore

Currently, typical machine movement speeds are somewhere between 30 and 100 mm/second. The problem with spindles is less the achievable speed, but the huge "mileage" your extruder has to move. Your machine will run at rapid speed for hours. Compare it to milling each part with a 0.5 mm mill bit.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Just in case anyone finds this thread in a search, I asked this question in another section of the forum and got some additional answers there:

[forums.reprap.org]

Hello did you ever build your 3-D printer I agree with your line of thought I use Mach 3 built aluminum extrusions and plate in my opinion the extruder and hot end the week point of reprap so I designed my own to be rocksolid reliable when I was happy with the results I started to build the machine the extruder and hot end was quite a large mass so I had to design the machine very strong and rigid to take acceleration and deceleration of the masses I use ballscrews with a high lead .600 per turn to get good speed here are a few pics of my machine hope you did build yours and post a couple pics