New to RepRap

I am new to the forums and reprap in general. I am getting involved in rapid prototyping and decided a while back to invest in CNC machines/lathes for also supporting small production runs if needed. However I need a 3D printer and I don;t really want to invest another 30-40K in a professional one. I am considering building a reprap, but would make all of the parts our of AL since I don't have one to start with, besides need the work for my CNC machines.

I have downloaded the files and started looking at what I need to do to make one. I do have some questions though as I go through.

What should I look for if I want to improve the overall quality of the finished parts. I realize the nature of the process will have a rough finish, but I am sure there are improvements that can help out. Some of the pictures I have seen, seem to have a better overall finish. Perhaps this is a function of the type nozzle or other things. again keep in mind I do have a lot of research/reading and digging to do.

I understand the UP! 3d printer has some of the best print quality currently.

Many people attribute it to the linear bearings. you can find several posts and wiki pages related to linear bearings.

I really like the stuff that prusajr is working on with simplifying the parts list, this might suit you well when casting or milling parts from Al.

precision, extrusion width (.5 vs .3 etc), speed, extrusion consistency, heated platform / chamber, skeinforge settings all have an impact.

personally, I have bene printing mainly on a mcwire repstrap, is extremely precise, extremely slow, and the main factors for my print quality are warping and keeping the extruder consistently flowing.

I know others have better luck with the extruders than I do (I have Wades geared NEMA 17) but I understand the Makerbot Mark5 is very solid. Working on testing others. I hope that helps.

@Anthong Redbeard

Thanks for the info. After posting I realized I kinda left my questions wide open. I have been reading and finding that the key is in the extruder as well as the accuracy of the parts made as you mention for linear bearings and such. I have also looked at the U print and the Z corp printers and while I could use them I think the cost and price of material would not be cost effective for me.

I have not looked to deep into mcwire and prusajr's info but will do that, thanks for the tips.

I also found the other forums by scrolling down.. DOH! lots more info in the others. I am interested in finding out if controllers such as the Gecko 540 can work. If they do it would mean I would have same controller and perhaps stepping motors for my CNC lathe later. Would only have to keep one set of parts on hand.

If anything seems it will be a fun project and get me familiar with setting up stepping motors and such before I start digging into that huge lathe anyway.

Thanks!

The G540 is nothing more than 4 opto isolated micro stepping drivers, so given the right hookup it will work fine. I'm using one for my build, but I'm not driving it with "standard" electronics, so it is a little different. The geckos are awesome drivers, midband comp is great.

Thanks, I did find it listed as verified working so that made me feel a little better, good to know someone else is using them without problems. They are highly recommended for DIY CNC machines, so I definitely want to go that route.

I run a CNC machine with one, I can hit 250IPM in rapid if I wanted, but the jerk forces are rather high, and I have no need to go that fast. With the reduced mass, and no cutting forces involved running a printer at speed should be cake.
A nice feature of the G540 is that the motor current set resistor is part of the cabling of the motor itself. When you make up a motor cable, you solder a resistor into the 9 pin connector, 1k resistance value per amp of current. No adjustments to mess with, and if you want to make another machine with one, plugging it in sets up a motor, making it easy to transport. I'm in the process of building an all in one power supply, G540 and smooth stepper box that I can swap between machines.
It's also protected against a lot of common failure modes, there's a fast blow fuse inside for power supply protection. You still want to power down before yanking motor cables (as with any drive) though.

250 IPM would probably shake everything apart unless you make it out of some heavy material. Even my CNC only pushes 60-90 but that is about all the rapids you need when making one off parts like I do. The time on part is the largest time. Good to know though, I may go ahead and buy one. I do want to look more at the electronics, I don't want to do a lot of soldering so would prefer off the shelf solutions if possible.

I'm right now testing different settings and trying to find the best quality preferences with a decent build time. So far I've had pretty good results, my first 10 parts were terrible. Where I've been adjusting the speed is under the global preferences tab, playing with the X and Y max feedrates. Also, I've been considering bumping the extruder steps per MM in the firmware. I'm getting quality parts now that I would feel comfortable selling on ebay or using to make another reprap. Reduce the infill gap amount under the extruder tab to something like .3 to take away infill gaps for a stronger part. Also, make sure to send the files to gcode with the proper prefs. Trying to adjust the prefs after you've sent the .stl to a gcode file does not work.

Are you still using the belt drive. Just curious if faster speeds are possible if you used a lead screw type system.

Yes I'm using belts I have an Isaac Mendel I think you're referring to the Prusja design Mendel?

I have read some on the Prusja, but don't see the info right now but will check into it. Was really just thinking out load, but the more I read the more I can tell that belt drives may be faster anyway.

'prusa'!

all the reprap printers use a belt drive for XY and screw for Z. i think the advantage of belts is that you can run the printer faster

DOH! You mean the third Reprap option! Found it now thanks.

Speed depends on the pitch of the screws, and the diameter of the pulleys used with the belt. You cannot simply say "belts are faster than screws" without specifying pitches and or diameters. I do realize that most people refer to common UNC/UNF or metric standard thread pitches, but there are other screws more specifically tailored for use in motion control that will easily match the speed of belts and then some. http://www.roton.com/torqspline-screws-quiknuts.aspx?line=Torqspline are a good start, and not that expensive.

Normally when I think of a lead screw, I think of the type you have posted. I know my CNC pushes a lot weight around with a simple lead screw. Myself I think the installation would be easier than the belt.


Koko76 Wrote:
-------------------------------------------------------
> Speed depends on the pitch of the screws, and the
> diameter of the pulleys used with the belt. You
> cannot simply say "belts are faster than screws"
> without specifying pitches and or diameters. I
> do realize that most people refer to common
> UNC/UNF or metric standard thread pitches, but
> there are other screws more specifically tailored
> for use in motion control that will easily match
> the speed of belts and then some.
> [www.roton.com]
> px?line=Torqspline are a good start, and not that
> expensive.

those threads look pretty great

The one downside to using screws is the end bearing support. Lets assume in the simplest case you direct couple the screw to the stepper (not using any reduction, but a coupling that allows for misalignment). The bearings in the stepper motor are only designed to handle radial loads, not axial. Without correct bearing support for thrust, you get backlash, and can even have problems binding in the stepper. In the mill I built I use pairs of angular contact bearings. These are not the cheapest, but provide extremely low backlash. You could use ball thrust races to handle the thrust load and rely on regular bearings to carry the radial load. Just gets a bit complex.
Both methods do work, and are in use in a variety of commercial and industrial motion control applications. I'm still undecided as to which I'll use, but the torqsplines are high on my list.

That's not entirely true. Permanent magnet motors (tin can) generally use bushings and have lots of axial play but hybrid steppers like the ones we use have a pair of ball bearings and will take some axial load. The NEMA17s I use are rated for 10N axial and 28N radial.

---

[www.hydraraptor.blogspot.com]

That dynamic load or static? Exceeding 10N dynamic load isn't that tough. Besides that there are simply better solutions that yield better results. While I understand the act of adding things to the project adds to the cost it really seems like no one working on this project seems to want to " do it right" once and then start to work down from there to determine the minimum needed. Instead we have an army of folks saying "this is good enough" without having explored anywhere close to the upper bounds of the system.
I've built and more importantly measured a system that has <.0002" backlash on 3 axis. I don't think you need that level of accuracy with reprap but I know what it takes to get there and I can give numbers to support it. I have yet to see a report of anyone measuring backlash, actual resolution of the system (not calculated), actual bed height across the bed or any one of a number of measurements I would think are essential to know how a system works for reprap. I'll put my money where my mouth is as I build mine, and put those numbers out there, but it boggles my mind the number of folks who say "it's good enough" without being able to quantify and compare.

I can say for me I am looking to build a mendel from the plans, but see no reason to implement some of the upgrades and changes others may have found along the way.


Speaking of the builds, does anyone know if there are actual complete CAD assembly files of the Prusa?

[reprap.org]

mlagana Wrote:
-------------------------------------------------------
> [reprap.org]
> _Mendel


Thanks, I should have mentioned I found that, but I have not found a assembled set of CAD files. This page only has the updated parts.

However, I did find the ISAAC and it seems that may suite me. They have already made a few changes to make the parts easier to machine on CNC as opposed to printing. That is something I see a lot in design is that parts can easily be printed on a 3D printer, but try getting the part machined that way may cost you an arm and a leg. A few changes can make a difference.

Magnum,

What_Tooling_Do_You_Have?

---

-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!

SebastienBailard Wrote:
-------------------------------------------------------
> Magnum,
>
> What_Tooling_Do_You_Have?


Thanks for the info, I did go through all the info. But in the end opted for the Isaac and will start gathering info for the parts I need to complete it. Hopefully will have it all together by Christmas break.