A 50 micron Z resolution FDM printer for newbie?

I am new to the scene of 3D printing, and I intend to print 3D parts that are subsequently used as machine parts (for model engineering), either directly, or as a master pattern, so I need a high resolution 3D printer. I know stereolithography printers can have a high quality, but they are expensive (both machine, and resin), so at the moment, I want to stick with FDM printers. Here are some requirements for the 3D printer:

- Have a Z-layer resolution of 50 microns
- be very reliable (don't easily break down, and safe to be left running overnight)
- have a relatively large print platform
- able to print in at least PLA and ABS
- mechanical parts (including stepper motor, transmission parts such as timing belts or gears) and electronic parts such as controller boards have ready replacement sources
- open source firmware and software

In this case, which printers I should look for? I have looked at MendelMax 2, but it is really expensive (USD 1400). Also what is the quality of Robo-3D? And how does Robo-3D compare to Reprap printers, in terms of build quality, and functionality? Robo-3D only costs USD 700 (fully assembled with heated bed for ABS), which is half of the MendelMax 2. But will the Robo-3D be able to satisfy the above requirements?


Is it really difficult to get the Z-layer resolution to 50 microns? What limits the Z-layer resolution? nozzle, or stepper motor, or both? Would the electromechanical parts of a modern reprap be able to do 50 microns given the right setting?

And if 50 microns is asking too much, then it is not necessary a deal-breaker and I can settle at 100 microns, but it could severely limit the quality of the product, which is used as master patterns for machine parts.

I had used the UP! Plus printer at 200 micron thickness, apart from the rough side layers, even on a flat surface, the cris-crossing of tubes also creates substantial roughness on the surface. Can I reduce this roughness by reducing the layer thickness, and/or making the head run faster to decrease the filament thickness of the nozzle?


Thanks!

Mark

50 micron layer thickness isn't too hard to achieve,doing it well is a different challenge! It's layer consistency that's hard and that's a problem at any thickness, just more noticeable as you go thinner. This comes down to to extruder calibration and mechanical rigidity of the whole setup.

Also when printing 100 micron layers I get problems bridging which is only going to get worse at 50!

To be honest going from 200 to 100 microns is a very noticeable difference in surface smoothness, I don't find going any further makes much difference.

A mendlemax is perfectly capable of these resolutions, but you will have to invest a lot of time calibrating and making your frame perfectly aligned. I don't think there is any advantage with print quality on the version 2,maybe the 1. 5 is cheaper (1400 seems like a lot, a mendlemax shouldn't be much more than a mendel!)

Mendel 90 makes lovely prints too, and I hear is very easy to construct.

So will 100 micron create parts that are smooth enough for master patterns of machine parts?

Also like you said, calibrating and making the frame perfectly aligned is critical in the success of 100 micron. This is why I really prefer cast / machined parts instead of parts made of a plastic printed from a 3D printer. In this application, reliability is crucial and the philosophy of self-reproducability is not important. If someone offers metal-cast versions of the 3D printed parts, that would be great!!!

The next question: Is 100-micron Z resolution with good layer consistency difficult to achieve?


The requirement of mechanical rigidity prompted me to look at MendelMax 2.0. I had a look at both MendelMax 1.5 and 2.0. The hardware kit cost are:

1.5: USD 490 2.0: USD 800

However the full 2.0 kit costs USD 1400. This means 1.5 is only USD 310 cheaper than 2.0, and the electromechanical parts costs USD 600, which I need to spend on top of both 1.5 and 2.0 anyway. This means 1.5 will cost USD 1090 altogether.........


Also what do you think about the specs of the robo 3D printer?

[www.robo3dprinter.com]


And finally, if I want an open-source printer that works well for 100 microns without too much manual calibration, then where should I go?


Thanks!

Mark

Every DIY printer requires calibration to get good prints out of it, if you've done it before and know what you're looking at it's not overly painful.
If it's your first printer it can be an exercise in frustration.
None of the DIY hobby printers are ever really click and print, they can be close once properly setup, but I tweak print parameters all the time,depending on what I'm printing.
The Robo3D AFAIK hasn't shipped yet, it looks a lot like a printrbot in a plastic case to me, I'll go out on a limb and say that you won't be able to buy one for a while because they'll be trying to fulfill the Kickstarter backlog, and I'd wait until real users have them so you can get some feedback.

> Every DIY printer requires calibration to get good
> prints out of it, if you've done it before and
> know what you're looking at it's not overly
> painful.
> If it's your first printer it can be an exercise
> in frustration.
> None of the DIY hobby printers are ever really
> click and print, they can be close once properly
> setup, but I tweak print parameters all the
> time,depending on what I'm printing.

I have a second thought, if the calibration is only one-off after assembly, then this is fine for me..... a learning process. However, if I need to finetune everytime I use it, then this will be problematic for me.

Also what is the extent of calibration? The startup automatic calibration of xyz? Or do I need to tweak the firmware variables a lot? Or just trying out different settings in the reprap software? However I am happy to experiment in things such as layer thickness, printed tube thickness and support type / density etc or nozzle / heated bed temperature that can be controlled in the software.

If I have to tweak the firmware of everything from the stepper motor controller, heater, extruder and the main controller for hours, then I am a bit worried that this may be a bit too much......

Like what others said, mechanical rigidity is crucial for good quality 100 micron prints. Therefore on this ground, wouldn't MendelMax 2 be most suitable? But what makes it so expensive? The aluminium frame or the electromechanical parts? After MendelMax 2 design is publicly released, will there be potentials to follow the design and reduce the cost somehow? Also to maximize rigidity, would it be possible by increasing the thickness of the extrusions, welding the extrusions together instead of bolting, and replace printed plastic parts with cast metal parts?


Though like what you said, I will wait until the MendelMax 2 and Robo 3D be released and I will read their reviews.....



> The Robo3D AFAIK hasn't shipped yet, it looks a
> lot like a printrbot in a plastic case to me, I'll
> go out on a limb and say that you won't be able to
> buy one for a while because they'll be trying to
> fulfill the Kickstarter backlog, and I'd wait
> until real users have them so you can get some
> feedback.

Calibration is mostly a one off, but it can be very involved.

You're making a printer from a bag of parts, there is measuring and typing numbers in, there is trial and error and there is debugging the things that aren't working as expected.

Perhaps a somewhat extreme example is I had some slight Z Wobble on my second printer, a mendel max 1.whatever, to fix it, I replaced a lot of the original parts and ended up printing different Z ends, the issue was apparent I thought I knew what was causing it, but I ended up replacing many parts over a period of weeks, it prints great now and I rarely do much ore than adjust the ZOffset in the firmware.

To print high quality parts at low layer heights, everything has to be working as expected, and getting to that point can take substantial time investment.

My 2c If I were only interested in production work I wouldn't buy one, if I were interested in it and had the opportunity to do some production work with it, I'd give a different answer. The question I guess is what is your time worth to you and how much value do you place in the experience of getting it working,

I plan on doing production with repraps. I don't care if I need ten of em I will do it. :-)

Polygonhell Wrote:
-------------------------------------------------------
>
> To print high quality parts at low layer heights,
> everything has to be working as expected, and
> getting to that point can take substantial time
> investment.
>
> My 2c If I were only interested in production work
> I wouldn't buy one, if I were interested in it and
> had the opportunity to do some production work
> with it, I'd give a different answer. The question
> I guess is what is your time worth to you and how
> much value do you place in the experience of
> getting it working,

95% of my interest is to do production work, as I really want to get a printer that works, and I only need to experiment with high-level things such as different support structures, optimize for the smoothness of the surface of the printed plastic and nozzle temperature for different materials etc.

As I am new to the 3D printing, I really want to get a working example first, and then learn how things work and how the system is calibrated etc..... though I do place a lot of value on experience of getting it working, but my time is pretty much limited.


So where can I get a reliable and sturdy PLA and ABS 100 micron 3D printer that doesn't cost too much?

I don't think there is a single FDM (FFF) machine in the world that can run that easily all the time. Even the most expensive professional machines have service contacts attached to them and a maintenance guy in every month to keep them going. Yet they still break down sometimes.

Although I've had a few different repraps that have happily run for months with very little to no maintenance after the teething period.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Depends what you mean by good enough for master pattern of machine parts. What you will have is a significantly better finish than would be possible with say sand casting, but it wouldn't be as good as a die cast part.

I presume you plan to use your 3D printed part and somehow make it metal. What process will you use? Is machining the bearing surfaces on a part after casting acceptable?

If you are expecting a perfect surface finish that needs no machining on bearing surfaces, then no, I don't think any rapid prototyping technique would be good enough. If you are looking for a decent sand cast finish and plan to machine critical surfaces afterwards, then 100 micron layers will be fine.

konwiddak Wrote:
-------------------------------------------------------
> Depends what you mean by good enough for master
> pattern of machine parts. What you will have is a
> significantly better finish than would be possible
> with say sand casting, but it wouldn't be as good
> as a die cast part.
>
> I presume you plan to use your 3D printed part and
> somehow make it metal. What process will you use?
> Is machining the bearing surfaces on a part after
> casting acceptable?
>
> If you are expecting a perfect surface finish that
> needs no machining on bearing surfaces, then no, I
> don't think any rapid prototyping technique would
> be good enough. If you are looking for a decent
> sand cast finish and plan to machine critical
> surfaces afterwards, then 100 micron layers will
> be fine.

Most of the printed models will serve as master patterns for the wax models for lost-wax metal casting, and in case of complex models with a lot of internal cavities, the printed models will be directly used in lost-PLA casting.

I am looking for a decent sand cast / lost-wax cast finish, though I prefer no machining on non-bearing parts.

By the way, when I print on 200 microns using the UP! Plus printer, people from a local model engineering club say that the surface is way too rough for sand cast as the rough surface does not allow me to extract the master pattern from the sand mould easily. Will the 100-micron thickness fix this problem?


Also, how can I calculate the necessary allowance of shrinkage in either case sand casting and lost wax casting? And in the case of direct lost-PLA casting, then how can I take the heat expansion of the PLA into account? Or does it cancel out the shrinkage of the final metal product?


Finally, I saw a wish for a more affordable mass-produced Reprap MendelMax 2.0 at the wishlist at seeedstudio, please have a look at vote for it:

[www.seeedstudio.com]

I haven't got much experience with casting (its something I am just starting to get into) but yes, I believe you need a smooth/shiny finish to successfully release from the sand. 0.1mm layers are considerably smoother than 0.2mm layers and it is very hard to see visible layers with the naked eye, however it is still a matte finish, not a shiny one.

At 0.1 mm layers, a quick scrub with a sheet of 100 grit sandpaper makes the layers invisible. With a sanding disk in a power drill or dremel, this would take no time. If you need a shiny surface, a generous coat of varnish would then give you a very shiny surface. In a lost-PLA casting, you probably don't even need to varnish and a quick sand will give you a nice finish.

The best suggestion I have is that you take a piece from the UP printed at 0.2mm layers (which I believe is a pretty decent printer), see how much effort it is for you to sand it a bit and apply a coat of varnish. If this gives you an acceptable finish then worst case scenario you will have to do this on all parts, best case you can print perfectly at 100/50 microns and use the parts are fine straight off.

I believe the UP uses a raft, while many people print directly onto glass, the surface printed directly onto glass gives you a super smooth shiny surface, so this can help reduce your work.

If you build a reprap from a kit, you can expect to spend a considerable amount of time getting things keyed in right. It took me about a week after completing mine to finally get to a point where I was getting consistent results, the most critical factors in achieving this being a completely level print surface, an accurate extrusion width over hight ratio, and proper hot-end temperature. All of these variables are going to influence to quality of your print, and all of them are unique to each printer, so there is some time investment to get them right. As for print quality in general - most printed parts require a fair amount of clean-up after begin printed. sometimes plastic doesn't get to where it needs to be and you end up having small amounts of filament that need to be removed, especially when it comes to bridging or overhangs. You also have to watch out for warping caused by uneven cooling, even on a heated print bed. My advice would be this: If you have time (and money) to spare, then a rep-rap or related style printer could get you what you want, but not without a lot of failure first. Expect to do lots and lots of test prints. I mean lots! what kind of machine parts would you be making and what is their accuracy requirement?

For machine parts, I plan to print gear box housing, axle box and perhaps flanged wheels and gears for model engineering. I am not sure about accuracy requirements, but I think the tolerance should be between 50 and 100 microns (correct me if I am wrong)

Anyway, if you want a printer that can do consistent 100 micron layers (or better, 50 micron layers) after calibration, which printer would you get?

i think when this is ready, i am going to kill myself calibrating it ... LOL
[3roomlabs.blogspot.sg]

The mendel 90 kit from nophead can produce excellent prints out of the box, mendel90 print 3rd print from machine without calibration

I can definitely say the Nop head Mendel 90 is superb. I have one the first print that came off it was amazing.

My blog details the build and some of prints.

[cylonmike.blogspot.co.uk]

---

Time to play every day... Why be bored and old

Been following this, any updates on what printer you end up going with>?

any thoughts on the mendelmax 2.0?

If you do acetone postprocessing on ABS either dip or vapor and coat with high solvent lacquer you get a finish like this picture of a Yoda I postprocessed this way and a bottle opener with no postprocessing