Hi All,

I am a mechanical engineer looking to buy a new 3D printer. I am not sure this is the right place to ask these questions. I am not doing RepRap. I want something which will work out of the box and be up gradable.

My application is simple design validation b4 paying the big bucks to have something professionally printed, machined, or cast. I don't intend to make a business from 3D printing.

My time-frame is within 3 months of having a functional and tuned machine. I can put in about 40 hrs in this time.

Machine Specifications:
+/-.005 tolerance
decent industry standard finish
low cost materials (want to stay away from stratsys)
machinable material (the option to do clear and colored light pipes and soft gaskets would be nice but not a requirement)
8x8x8 print area
Open source electronics and software (I don't really need open hardware. I can easily re-engineer and build this.)
Price range up to $3k USD

I have a lot of existing machinery. My business is based on design and building 1 offs/1st prototype. I have designed and built a lot of machinery/mechanisms. I have outsourced a lot of 3D printed parts (mostly strasys). Mechanically, I will not need much support. I will need support on electronics and software. I am, however, not a newbie.

The printers I am now targeting are the Felix 3, Bukobot 8, and the Lulzbot TAZ. In that order. I am open to other suggestions. I do want something which will be running parts within 10-20 hours. Tweaking can take longer. I cannot wait 3 months for delivery of the machine.

I would be open to suggestions from experienced 3D printers.

Ultimaker 2. All the way.

Edited 1 time(s). Last edit at 06/12/2014 01:13PM by MrDoctorDIV.

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus

Just to make sure, you are looing for +/- 5micron (0.005mm) tolerance or +/- 0.005in tolerance?

Quote

I am not sure this is the right place to ask these questions. I am not doing RepRap

closed source printers offer services like printed samples. If you will change your mind, and choose to build a RepRap: The forum will offer lots of material about RepRaps

Quote
MrDoctorDIV
Ultimaker 2. All the way.

Thanks! That one looks pretty good. According to the Make Guide its right up there with the Makerbot Replicator 2 but open source. EZ to heat as well (I live in San Francisco).

I have heard good things about the Ultimaker (internet). It looks like they have a good forum as well for support.

When people say you need to tune lower cost machines to get them to print like the expensive ones what do they mean? I assume this has to do with tuning the motors and some process controls.

Quote
Polynom

Quote

I am not sure this is the right place to ask these questions. I am not doing RepRap

closed source printers offer services like printed samples. If you will change your mind, and choose to build a RepRap: The forum will offer lots of material about RepRaps

I may make one, one day. I just won't make one which makes one. I also just prefer steel holding my machines together. It is generally steel on the machines, aluminum on the motorcycles. Maybe that will change some day?

I wouldn't give the Replicator 2 any kind of praise, they are infamously troublesome and lacking in actual quality, especially for the price. Prints they show for showing off their printer look like garbage to what some of the 1/5th priced printers here can print. Unreliable. That's all just what I've read, though. I've never seen or used one, but it seems they've taken people's money before making a truely ready product, all too common in the 3D printing world.

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus

I think the the Taz is a good printer to get going fast looks like a large amount of it is preassembled. I have not personally used one but a looks like a pretty good one to buy and like I say pretty fast to set up I have bought other things through the company before and they were pretty good. Some people have had problems with their hot ends and nozzles I would switch to a all metal hot end anyway

Edited 2 time(s). Last edit at 06/13/2014 11:53AM by cnc dick.

I have used a Taz at our hacker space and we have had nothing but issues with it for 2 months. It is now working but we had to change the hot end out completely to a J head. Even now that the head is working, the printer screws up every time its moved around the space due to the frame not being rigid enough and it requires re calibration.

Ultimaker 2 is still the way to go. If you have the money ($15k or so) then you can lease a small SLA or SLS machine to do some high end stuff as well.

I am liking the Ultimaker 2 more and more. I really like the open source.

The Replicator never made the A-list. Part of my reasoning for doing this is to eliminate some of the prototypes tolls from Stratasys. I feel I pay a high price for inconsistent results. Maybe I am wrong?

Where would I get materials for the Ultimaker?

Directly on their site.
Or if you're going to build and source yourself you can find the source files.

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus

Go over to Techshop in SOMA (Howard and Mary) and check out Type A Machines. They've got space at Techshop and assembly in San Leandro. Last I was in the space Techshop was using the older machines but in the Tyope A office/space there were plenty of the new printers. Lead time on any larger prosumer or pro machine is going to be a month or so at least, perhaps more from almost every vendor. The reason I haven't bought ready built machines is that by the time they get here I could have made a few at about a third the cost for materials. My choices would be Ultimaker, Type A or Lulz in no particular order. I've seen them all print in person and all appear to me to be good quality and have good support.

Sparkfun says they have TAZ4 in stock, ready to ship. It's about the only one I've seen without a lead time and that announcement came in this mornings emal from Sparkfun.

Quote
Iceman086
I have used a Taz at our hacker space and we have had nothing but issues with it for 2 months. It is now working but we had to change the hot end out completely to a J head. Even now that the head is working, the printer screws up every time its moved around the space due to the frame not being rigid enough and it requires re calibration.

Can you let me know which hackerspace this is, so we can get things straightened out? We certainly have many, many hackerspaces running various lulzbots successfully, so I'm sure we can get yours going too.

Thanks,

-Jeff

As I am not an officer of the hacker space and merely a member, I would rather not disclose the location in a public forum. Please PM me and we can talk further.

I will say that the Taz is now running well but with a J head and not a Buda schnozzle.

Quote
vegasloki
Go over to Techshop in SOMA (Howard and Mary) and check out Type A Machines. They've got space at Techshop and assembly in San Leandro. Last I was in the space Techshop was using the older machines but in the Tyope A office/space there were plenty of the new printers. Lead time on any larger prosumer or pro machine is going to be a month or so at least, perhaps more from almost every vendor. The reason I haven't bought ready built machines is that by the time they get here I could have made a few at about a third the cost for materials. My choices would be Ultimaker, Type A or Lulz in no particular order. I've seen them all print in person and all appear to me to be good quality and have good support.

Thanks for that one. I forgot about them. They are a few bart stops away.

Quote
vegasloki
Lead time on any larger prosumer or pro machine is going to be a month or so at least, perhaps more from almost every vendor. The reason I haven't bought ready built machines is that by the time they get here I could have made a few at about a third the cost for materials.

What if I just bought something similar to this: [www.ebay.com]

I'd weld some pads on steel tubing and machine it true (3 ballscrew gantry design) to attach them to. I would hook up some Nema steppers with a power supply and controller. Now I can move in a 12" cube (planned machine size) repeatability to within +/-.003". I can manually measure and tune the software to get this more precise. It will still have good repeatability. Seeing some of the other set-ups, I would guess this is well within tolerance. I am not 100% sure on how to measure it besides just print quality?

I would then need slicing software. I know there are several available (Slic3r, Cura). I assume this outputs Gcode.

The Gcode would talk to my controller software. I am not sure what to use here (linux CNC)?

I will then need a extruder nozzle. Any suggestions.

I will also need a heated plate and I assume float glass on top. Any suggestions?

I could later do a box around it and control the internal temp with a light bulb. I could preheat and dry my plastic this way.

Not to say that I am going this way, but it is good to have open options.

I don't suggest building fromscratch for a first build even if one does have CNC experience . Build something like an i3 Rework first to get the hang of things. The software tool chain is complete though not as mature as other CNC devices. Read up in wiki for new users and that will give you a good starting place.

If you want reliability and properly sized prints, don't go into the Slic3r world I have escaped from. I bought myself Simplify3D, and although I can't say anything about it vs KISSlicer since I haven't used it, I can say it is very well done on the slicing side. It can control your printer as well, but I still very much prefer using Repetier Host for controlling sofware.
E3D is a popular hotend, I use one myself, the v5 experiencing the common PLA problems. The v6 has fixed these problems and according to some people is unjammable. I don't entirely believe that, but it says something.

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus

For slicing software I'd highly recomend Cura over Slic3r, kisslice and skeinforge all of which I have tried. Cura has a good work flow and is straight forward to use and tweak. Highly polished. Brought to you by the same people who make the ultimaker: [software.ultimaker.com]

EDIT: Also, Cura chooses better toolpaths than Slic3r which I have used extensively. So there you go

Edited 1 time(s). Last edit at 06/16/2014 09:31AM by TheTechnicalNoob.

I too would suggest the Ultimaker line.
I've seen and dealt with enough machines, and the Ultimaker line stands out as being stable, and consistent. The Lulz is a nice machine, I hate the hot end, and it hates being moved, just as the others you chose, but otherwise, it's a nice machine. Change the head (if it bothers you) and leave it/them on a solid desk they will do fine. I wouldn't go near a Makerbot.There are a few others that work okay, some have minor bugs here or there (some fixable), but if I was buying new, complete printer, my first choice would be the Ultimaker line. They're pretty much the most consistent, stable printer I've been around besides my own design.

I'd also recommend Cura, it works well, it's easy to use and uses little resources. It's fast enough to use on netbooks, so long as you aren't slicing something large.
As for materials, you can get filament from many many places, you aren't stuck with just one brand.



I will second Vegasloki regarding building from scratch.
You said you have 40 hours, you can easily burn up a large percentage of that just tuning and learning how to get good results from a home built machine if you have never done it before, and easily much more designing and building your own by the time you fix mistakes and bad choices. They can be simple machines, but there are a lot of intricacies and gotchas. If you want to build one from scratch, follow someones plans and/or buy a kit. It will save you a TON of effort and money. If you do build a popular kit, look for a vitamin/minerals kit, that will save you a lot of time, money and hassle.


That said, if you want to build you own...
For nozzles, there are lots to choose from, some are better at some plastics than others. J-heads are well respected and sort of the jack of all trades, it works pretty well with just about anything, just make sure you get a good one (hotends.com), knockoffs have varying degrees of quality. You only need a heated bed for ABS, PLA can be printed on blue painters tape or Elmer's glue stick on glass. Some print right on a heated bed pcb, but most these days will recommend glass. Many just clamp the heater and glass together with clamp style paper clips. Use belts, no string/fishing line, it will save you a lot of hassle. If you build a heated chamber, leave the electronics and motors outside, you really only need one for ABS though. Many get by with just blocking airflow over the print surface when they need to do ABS.

Thanks for all the info. Another good link with build info: [geekhack.org]

I will most likely buy then build (if I ever do; no time). But building one does sound like a lot of fun.

I should probably ask my ballscrew/linear slide questions more directly. There are rolled ball screws available from China with pretty good accuracy for not a ton of money. China also makes lower cost profile THK type rail as seen on the Felix. This set-up will allow for a very accurate (repeatable to same location) although not a super precise (variances in rolled ballscrew) setup. The precision may be good enough for 3D printing off the bat. If not, you can tweak the precision via software and get a precise and repeatable machine in one calibration setup. Belt tension is not a factor. Why are there not more like this?

Ballscrews for this sort of design are overkill and don't offer an advantage compared to the price. Many if not most using these types of DIY designs/kits use M5 or M8 threaded rod for the Z and belts for the X and Y. I've got machines that use TR10 lead screws and I don't find that much more advantage to them. The commercially built machines are using lead screws and belts as well.

I've used all currently available slicing software. Slicing software (and the CAM toolchain in general) is one area that can use more development and the devs are certainly making strides toward that. I find Cura slices well but lacks some features and in some ways isn't flexible. The thing about Cura I don't like in particular is that the hot end heats prior to the bed. I haven't found a good way to change that except by editing the Gcode directly. Slic3r offers good utility and a strong feature set but some of the slicing can be problematic or not as consistent as some of the others. It looks like KISSlicer is back from the dead with Jonathan updating it. It slices as good or better than anything but lacks some features and some aren't full implemented. I think Simplfy 3D is a good concept and a solid program though the implementation could use improvement particularly for the steep price tag. For a DIY build it could be 25% or so of the cost of the build and I don't think it adds that much value compared to freely available solutions. It's easy enough to try any of the free solutions and use the one you like best or pony up the dollars and get Simplify 3D.

I'd caution against over thinking your design. If you are looking for a design with more precision components look at a Mendelmax 2. When looking at the design of a machine you are only as strong as your weakest link. That is if you use ballscrews on one part but drill rod and low spec linear bearings you'll be limited by the lower cost/spec performance. Additionaly, building your own at first you won't have a baseline for what the performance should be and will spend a great deal of time calibrating and reinventing the wheel. Reprap is pretty mature and offers the path to learning about these machines at a relatively low price point and depending on the skill of the builder/operator a farily quick learning curve though it may be 100 hours or so before you are proficient. It take others longer.

Edited 2 time(s). Last edit at 06/17/2014 02:39PM by vegasloki.

I received the Ultimaker 2 about a week and a half ago. It was stuck in customs for a couple of days since there was no value listed for the power cord. It was still very fast delivery.

I did not have time to open the box until a couple of days ago. This is my second print after the test robot.

I think this will be a great tool!

Thanks again for all the advice!

Quote
vegasloki
The thing about Cura I don't like in particular is that the hot end heats prior to the bed. I haven't found a good way to change that except by editing the Gcode directly.

If you put in your own commands for heating the bed and hotend in the start gcode setting in Cura, it will work like you want it to.

That's a beautiful second print. I can't get that quality until a week of retuning and configuring.

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus