by Stargrove1 Dec 4, 2016
Download All Files

Thing Apps Enabled

Please Login to Comment

I got this moving in pronterface perfectly. Now, do i need to get some specific slicer settings?

No special settings are required, just start with default and do a calibration like a standard cartesian machine

Had a look again at my original MicroCube, now testing 3d printed bushings (https://www.thingiverse.com/thing:2202854), so far they seem to work. I love the MicroCube as a platform for testing new ideas and concepts. It is cheap and simple, great design!

Customizable Linear Bushing / Bearing - LM8UU, LM8LUU, LM12UU, LM12LUU, LM16UU, ... IGUS like

I actually startet out with the idea to include the linear bearing as a printed part to the x, y and z axis. But first trials with the h-bot style, showed that they tend to stick and therefor jump. It would have saved me a lot of space and the design would have been moer clean and simple. The attached picture shows one of trial parts.

I actually use the MicroCube for exactly the same purpose, I wanted to test a 32Bit board without talking my Robo or Kossel apart, with the MicroCube, I have easy to access test bench for everything I'd like to try. Now I have added the 32Bit board to give smoothieware a try, I also use the LV8729 drivers, which are working great and silent. I also replaced the extruder by a bondtech clone. Now that I know the performance, I will update my old Robo with at least the drivers and dual gear extruder.

Somehow the linear bushings are working so much better compared to the commercial IGUS ones. I'm totally keeping them for now, print quality (with my new fan shroud https://www.thingiverse.com/thing:3569037) is great!

MicroCube alternate Fanshroud (Single 40mm Fan for Hotend and Part Cooling)

You may want to have a look at the x-carriage of the CXY version, I included the part fan to the carriage, by this the hotend and the part can be cooled separately.

Hello everyone, I'm new on 3d printing, I have ideas of building the hypercube evolution, but I want to learn the maxim I can. So I found this project and congratulations to Stargrove. Why I chose this project, not only for the price, but for simplicity and the basic material, that I sink is incredible to learn. Sorry if am gonna make too many questions. The goal is to star from the base the microcube, then make the upgrade parts with microcube, and demonstrate the evolution the curve of learning. I'm gonna make a make on the project. And I hope to have the collaboration of all. To Stargrove how can I contact you by email? Thanks

Hello, if you have any questions, you can post it here in the comments, by this, other users may help you as well.

I would suggest to use the Nema17 version with the corexy setup as it performs way better. You can find the link at the top of the Details section. I have also improved some parts like the z axis or the carriage with included fan

Oh you will love my current project, I'm merging the MicroCube and the HyperCube Evolution schematics together. I really like the simplicity and low cost of the MicroCube, but I want a more relieable movement system like the one from the HyperCube. Everything is still in prototype stage, but I have high hopes for my symbiosis of both projects. I'll have to redesign a few parts and improve on others, then I'm going to publish all the plans.

This looks amazing! I've already got an Evolution and would love to build this!

Your top x-carriage bearing seems to be misaligned here.

This is still a mockup to test the motion system and frame rigidness.

Is the frame rigid enough to run the belted corexy setup? When I created the corner blocks, I was aware that I will keep is small and lightweight, that's why I kept them quite small. In your case, I guess you have increased the length of all three dimensions. How is the performance, I found that the MicroCubeCXY has already a quite good quality as shown in the sample prints.

Printed a few samples like the gecko model https://www.thingiverse.com/thing:1363148 which worked even without a part cooling fan. The machine is precise, but rattles around like crazy. Funny enough, it already prints better than my Fabrikator II mini even in this prototype stage. I should definitely replace the A4988 drivers and reprint some parts for better alignment. Probably going to try a MKS Gen L board on this build aswell, didn't you use one for the CXY? Ordered myself a set of TMC2208 drivers, successor to the 2100, haven't heard of them so far but they were cheaper...


I saw a cheap MKS clone on Aliexpress, as this board has all included in one pcb, I got rid of the stack on RAMPS gaining a little z travel. I use the MicroCube as a little playground to test parts, I also exchanged the A4988 stepper diver by MKS LV8729 as everything is so easy to access, in my opinion, the hotend fan is the loudest part now. I also exchanged the extruder by clone of the Bondtech dual drive BMG extruder which also works pretty good.

Meanwhile I found a cheap 32 bit board compatible with smoothieware, this will be the next part to try.

I frankenstein'd together a working prototype. Attaching endstops with cable ties is innovative! Also butchered some pieces together with a 3d pen... I was curious about the performance of the movement system during a print situation, the video shows a PLA vase mode print with no part cooling fan. I'm very happy with the results of this rough prototype. Next step is to remodel the parts with all the stuff I learned from the test, invest in proper belts, pulleys and stepper drivers and reprint everything in metallic red ;-)

Nema 11 are enough to drive X, Y and Z. I chose Nema 17 again for the extruder, just to make sure I don't get any nozzle clogs.

Video: https://youtu.be/4UCG6auol24

I attached all HyperCube parts directly to the corner blocks, all belt forces are parallel to the rods. It is a very clever design and this way the belt tension is holding the structure together and not working against it. I'm very pleased with the first tests, still have to reinforce some parts and work on others like hotend and endstop placement. I'm also using Nema11 motors, which seem more than enough to drive the belts. I had trouble with layer shifts using your fishing line, this was my main concern with the MicroCubeCXY too.

So far the belt system of the HyperCube does an oustanding job and the frame is stable even without the reinforced diagonal rods, I recorded a small clip here https://youtu.be/Nn7XD-JAMPc

I attached some files, but heavy work in progress! This is all you need to build what I have right now + the parts from the original MicroCube and HyperCube projects. I will publish the final and tested design as a complete project.

send me your contact, if you need i can help on anything

Hello, I hope you will have a lot of fun and also learn a lot while building. You can post your questions in this comment section, but I will not provide my email address. Good luck

Hmm...just found these 39MM Hollow Shaft Hybrid Stepper Motor 4 Phase 5 Wire Square 1.8 Degrees Stepper Motor [Shaft is hollow, but has a 6mm shaft so externally screws can be attached]

It's wired like the smaller cheaper...but a full stepper (not geared)..and pretty darn cheap (cheaper than Nema 17's and 3-4mm smaller) more power, better speeds....
I will see if I can order a few...and maybe make a mod...still working on getting my MicroCube(XL) finished...got everything on it...finishing up the wiring...and hopefully will have it running before the next century is up....LOL

I checked out the stepper, I am missing the actual torque value which would be interesting for me. Resistance of the coil is way higher than the usual Nema ratings, I will also order some to check out, size and prize are very good.

Found some equivalent (and about the same price) from aliexpress

.13Nm calculates out to 18oz holding torque

From this link it shows .13N.m....but remember just like the 28BYJ-48 cut that red wire that splices between both coils the torque goes way up. (but this is just "estimated" and without physically having one, no way to tell)

They have a 5mm shaft like the Nemas, but I wonder if the torque is enough to act as y x axis driver. If you try them out, let us know about your experience

See the BOM, at least 10 x 200mm, if the diagonals are added to stabilize, additional 2 x ~250mm are required, for the included spool holder, again ~1200mm in sum

How much 8mm tube does it take to make this?

The idea of a belt driven MicroCube is haunting me ever since I tried with the H-Bot system which ended in failure. I came across this amazing coreXY design https://www.thingiverse.com/thing:2254103 and I'm thinking about a crossover between this and the MicroCube...

HyperCube Evolution

I never aimed to reach such a quality, size and speed as the HyperCube, my goal was a small, cheap and easy to built desktop printer.

However, since I saw your core xy version, I started a little redesign of the MicroCube. I will release a Core xy version of the MicroCube as soon as I finished the design. The x-carriage is finished, it uses core xy setup, is cleaned up and includes a better fan design. Next up is the y-axis design. Once this is done, you can simply update the MicroCube by reprinting those parts. Only point is, I still stick with the fishing line as it seems sufficient to me.

This is exciting news! I can't wait to give it a try. I'm currently playing around with some ideas, but haven't started building anything.

The Nema17 Extruder looks very clean compared to the double 28BYJ "construct", I'll attempt to build a version with the BYJ for X, Y and probably Z combined with a standard Mk8 Aluminium Extruder + Nema 17. Looking forward to this build, after building the 3Dtje printer. Alot less parts to print and 8mm Aluminium tubes are super cheap in the hw store! Cheers!

You are right, the extruder with two 28BY motors is a little fragile, but it was working better than I expected, the sample prints in green were printed with this extruder. I have no information about the durability since switched to the Nemas quite early. It was more of an experiment if it is possible and also to create a very low cost printer for students and education.

I would suggest to also spend the two Nemas for X/Y axis, the speed and accurace performance is better even though the 28BY also worked suprisingly well. But as I tried to keep it modular, you can easily start of with 28BYs and then just print the Nema brackets to update.

It was also my intention to reduce the amount of printed parts as much as possible, I had a look at the ToyRep before which consists of quite large parts, this, and the fact that I wanted to test the h-bot style lead me to design the Microcube.

Have fun, I would love to see your "built" when your done

I finally did it! The first prototype is working, using a mix of Nema17 (E0), Nema11 (Z) and 28BY (X/Y) motors. I plan to replace all 28BY with Nema11 once I get them from china, they are just amazingly small and have a nice amount of torque. I used drylin linear bearings and aluminium rods for everything. So far it seems to work, I have to replace my modified parts with the final versions and clean up the wiring. I'll upload my build, once I'm done with all the testing. So far it was a really fun build and an amazing design!

Nice, yes the Nemas are a better choise, that makes the printer more agile. I found that it could be actually used for small parts on the desktop as ist is small and silent.
Did you extend the z-axis? It's hard to tell from this perspective.

Comments deleted.

Very nice, somebody already asked about extending axis for bigger volume. Have printed something tall yet? I would be interested in some pictures of printed objects to see how it performs at the upper end of the print.
Mine is still pending, I want to update the z-axis motor and the whole bed as well. Unfortunately I don't have time right now for this.

I just finished the second attempt of the model Black Swan v2 in original scale, which is ~140mm tall. (PLA, 0.2mm, 3 shells, 3 bottom, 5 top layers, 190°C, no infill) I tried printing it with no infill to save time which (almost) worked. Apart from some minor missing support structures the layers look really nice. I attached pictures of the top for you, also a video of my first attempt printing is on youtube (https://youtu.be/ZYm4_e9loPM). I changed some things about the fishing line and reduced the printing speed to 20mm/s, also lowered X/Y acceleration settings. I get less layer shifting, but it was still present in the lower part of the model. Will have to investigate further, still waiting on my Nema11 to arrive. Maybe they work more relieable.

Wow, very nice, I am sure the Nemas will improve in a lot of points, mostly in possible speed but also in precision. I hope you turned the induction heater of before starting the print, since it tells "Zeit" and "Stufe", I guess you are also in Germany :-)

But even with the 28s, the model looks quite nice, great to see that an extended Z axis works that well.

I was tempted to use it as a heated bed... but we have nice warm weather here in germany, so it wasn't needed ;-) Currently printing for precision testing, 0.1mm looked flawless, now testing 0.05mm!

What a heat that would be :-) yes its still warm here, but thunder is coming.
I can't wait to see some pictures about the high resolution prints

I did some tests and I'm very happy with the potential! I attached pictures from my 0.1mm and 0.05mm prints. The prints suffer from some part cooling / retraction issues, but more importantly I keep getting layer shifts. I tried more and also less loops around the pulleys, different fishing line strength, higher and lower the printing speed, still shifting. I think it could be the 28BY motors or some wierd combination of movement which gets the fishing line stuck or slipping. I will try again once my motors arrive and report back.

As for the prints, I'm esp. blown away by the 0.05mm one. You can hardly see / feel the layer lines in reality. I never imagined a 0.4mm nozzle could do this. I have high hopes for this printer once the little problems are sorted!

Wow, I am very impressed about the quality even with the cheap motors. You might be right about skipping steps, problem is that the torque is pretty low, its also tricky to set the current to have enough power without overheating. I think with the Nemas, you will achieve better results pretty easy. I saw some oozing, at which temperature did you print? You may also try to increase the retraction setting since the printer moves slower due to the motors, it requires a little more retraction.

I was printing between 185C and 190C with 4mm retraction. The main problem is very likely the part cooling, I'm using a shroud adapter to redirect some air from the hotend fan to cool the part.

Is there a reason why you don't use the separat part cooling fan? I know it ain't pretty, but it improves the print quality, expecially when printing overhangs like the Benchy. I used a 30mm fan, but there is also a remixed version for a 40mm fan (https://www.thingiverse.com/thing:2538976). You can still add them, you only need to print two small parts and add them by M3 screws to the carriage.

When upgrading to Nemas, part cooling fan is even more important because of the higher speed.

Microcube 40mm Fan Connector

No particular reason, I'm still prototyping and this was the easiest solution. I will look into part cooling and print refinement once my mechanical setup is relieable, which it is not at the moment...

I have problems with the reliability of the fishing line and the drive wheels, it keeps slipping / jamming and shifting layers. I'm thinking about redesigning the upper part of the printer to use a GT2 belt instead. Just have to lift everything up to compensate for the wider belt and add some geared pulleys to X and Y motors.

I think a belt would be too much for the small frame, maybe try to tighten the string first, if this does not work, try an extra loop (or more) around the pulley. I saw printers with way more loops, this will surely increase the grip.

I tried using a belt with worse results, see https://youtu.be/ufwftvI2GNM

I was using Nema11 before but they were too weak to drive the belt, then upgraded to Nema17 and they had no problem moving it. But the stiffness of the frame wasn't enough and everything wobbled around. I'm now back to the fishing line setup with Nema17 and redesigned the pulleys for less movement of the line. This seems to work really nice, I will publish the files once I tested everything!

As mentioned, I was expecting that the frame and carriage is not rigid enough to be driven by a belt, in the beginning of the design phase, I wasn't even sure if it is good enough at all.
But I am glad that you shared the result of your trial, thank you.
Good to hear that you have a working setup, I don't know why different pulleys improve the print, but great that you got it solved.

I just found the following site .... seems someone risked to even convert it to a CoryXY (unfortunately, I can't read the result):

I converted the fishing line version to CoreXY inspired by your link and it works so much better than the H-Bot! The missing stiffness of the x-carriage doesn't compromise the movement system anymore, I have no "wobbling" or "bending" in the movement. I'm very happy with the result of this really easy conversion combined with my modified pulley design. I'm expecting great things from this little beauty, thank you again for this amazing design! Did you plan to use CoreXY all along? I noticed already perfectly aligned holes in the front corner parts.


MicroCube CoreXY conversion
MicroCube Nema17 Version, tighter Pulleys

Damn it, wish I did myself. Can I send you mine for the upgrade to core xy :-)

Would it be ok for you to include this in my original description and set of file, as it is to good to miss for the MicroCube if someone want to build it?

Please do, I would be honored! I like having all these options available on thingiverse. People can build any flavour of this design they want and improve on it - it is super addicting to have all those little side projects going. Next I have to fix my 3Dtje printer... blew my ramps board up :-P

Slightly off-topic, but my 3Dtje is back alive! I'll definitely try to build a tiny delta printer next. They look intriguing. https://www.thingiverse.com/make:521678

#3DBenchy - The jolly 3D printing torture-test by CreativeTools.se

That was also on my schedule, a tiny delta using 28BYJ motors for the axis, just like MicroCube very minimalistic. But meanwhile, I found a linear rail Kossel Delta on Aliexpress for ~200€ incl. shipping from Germany, and before picking up this idea again, I want to equip my MPCNC with the laser for engraving first.

I made some modifications, yes. I reduced the X/Y dimensions to 200x200mm (180mm rods) and extended the Z to 320mm (300mm rods). I saw the original build volume of roughly 120x120mm and decided to go for a super cheap 100x100mm mirror tile as printbed. I really felt inspired by the ultimaker 2 extended machines and opted for a larger Z build volume. I have usually very little space on my desks, but lots of unused space above the printers.

I plan to refine my changes and present a "I made this" page, once everything is tested and works. (Currently testing my remodelled buildplate holder and mount option for the extruder.) Extending the Z axis was easy and so far I did not run into problems with my Nema11 driving the axis. The fishing line motion systems works really precise too, I'm very happy with the mechanical side of things.

Did a ratio calculator
200mm -> 250mm (for the 8mm smooth rods and the metal rod)
So the 188mm 5mm threaded rod >- 235

200 : 250 = 188 : 235

Also did the calculation on the bed, 200 : 250 = 150 : 187.5 so the bed could be increased to say 185mm square vs the 150

I was wanting to remix to make MicrocubeXL (a little bit bigger square...not alot bigger just a little more beefy)

Would this work?

If I get you right, you want to increase the frame parts from 200mm to 250mm in all three dimensions, right? If so, you simply need to add the 50mm to M5 rod since the motor shaft remains the same, that means the length of the M5 threaded rod will be 238mm

Same thing for the bed, if you have 50mm more to travel on both axis, the bed should be 50mm bigger in both dimension, but please keep in mind that you need to remix the L shaped parts of the bed, otherwise the screws will be inside the print area.

Ok cool...was just wanting to make sure it was feasible...I have A TON of other projects in the mix already and I was just "dreaming" of making one of these...maybe with Nema 14 or Nema 11 motors

Thanks for the clarification

ok, I did a Nema 11 motor bracket...and now just did a 50 mm wider Z_axis_platform_mount(XL) by chopping stock bed up, adding 25mm long blocks then putting back on arms...so it should be able to fit a 250mm wide frame.

Stupid Nema 14 and 11's are more expensive than Nema 17's....fooey


Nice job. Can't wait to see a picture of the enlarged version, I hope it woll work out fine.

Due to the high production rate of the Nema17, the price is quite low. I hope that the smaller versions will follow this trend

Well figured out a few things while messing with the printer (trying desparately to get it working)
Found that the pinouts/colors default are all wrong...they need to be blue/yellow, pink/orange
otherwise the motors are too weak to move and just wiggle/jiggle back and forth...and heat up stupid fast!!!
Then I think I killed my cheap chinese Arduino Mega...not sure how...but the display now doesn't work
Think I figured it out...smoked the teeny tiny little 5vdc regulator...which then smoked the display...and/or just need to wire in a LM7805 or something...those small surface mount ones only provide a few hundred milliamps and can't/won't run enough..then I compounded it by plugging both USB and powersupply up to the ramps board creating a ground loop fault and really burned stuff out...yippie!
I think I have a spare somewhere...gonna fix it...fooey

after making the hack can i use the driver provide with motors it is based on unl2003an, and then from this driver wired to the ramps

It is required to do the bipolar hack, you need to open the housing or drill a small hole to interrupt the connection between the coils (see https://www.instructables.com/id/Supercharge-and-Convert-a-Unipolar-Stepper-to-Bipo/). Otherwise the drivers of RAMPs will not work. It is also a bit tricky to set the current for the small steppers correctly, too less means not enough torque, too much will end up in heating the motors until they fail. Take the values from my guide as a starting point.

Sorry to hear you board is broken, I once killed one from my Robo3D too. That happens :-)

Im crossing fingers that you get it running soon

Oh yeah...I meant to ask...do you use 1/16 stepping setting for steppers or no jumpers on the RAMPS board

Yes, microsteps with 1/16 is used

Oh I did...then figured out with the link above the colors on the pinouts are not correct (in stock form)
The correct coil patterns for bipolar are Blue/Yellow, and Pink/Orange in case anyone else has the problem I did.

Even in that link for the instructable...it doesn't mention the coil match or color codes
In fact with the normal driver http://www.geeetech.com/wiki/images/thumb/c/c7/Stepper_schematic.jpg/700px-Stepper_schematic.jpg
It shows Blue/Pink, Yellow/Orange as the coils...which is not correct
In the plug colors go Blue,Pink,Yellow,Orange,Red (for unipolar) bipolar the pins/colors need to be Blue/Yellow, Pink/Orange (Red cut/Empty)

I found these motors on aliexpress...both Nema 11 and Nema 14
The Nema 11 - https://goo.gl/P4HBRB
and Nema 14 - https://goo.gl/3BqFgd
I see the Nema 14 motor mount...is there a way you can create a Nema 11 motor mount
I see ort has done alot of remixing...maybe he can make the Nema 11 remix versions

How do i go about using a nema motor for the z axis? also could i swap the fishing line out for something a little more substantial, say gt2 belt?

I still want to create a full version, but I can't find the time at the moment. But others created a Nema14/11 solution you might want to check out:


There will not be enough space to simple switch the gt2 belt, you would need to redesign the y axis to have more space. I wanted to keep the printer small and simple, thats why I didn't use the belt.

NEMA 11 motor mount for Microcube Z Axis
by ort
MicroCube z-axis for NEMA 14 Reuses everything else

Don't know if I'm missing them

But can you put source files up?

The STLs are available and can be imported to change them. The actual source files are, I guess, not very well engineered since I am still new to 3D design, they can't be shared with the public ;-)

Don't worry about it. Very few here are experts, we're just hobbyists who do whatever we can.

Look at the OpenSCAD source which Johann used to make the original Rostock and Kossel, the code is awful and confusing, but it doesn't make the printer any less revolutionary or lessen his contributions in any way. Quite the opposite! The source is there so people can modify and remix as much as they want, and the design gets popular.

I look back on my own early CAD designs and shake my head. I'm sure in a year I'll look at my current designs and shake my head again. But that doesn't matter, the parts themselves worked and others can use them fully. That's what's important.

Don't be embarrassed about your designs, we all started somewhere. This printer is excellent and nothing will change that. Someone may even show you a better way of doing such a design, so you get a contribution to make your project better and you learn something and you grow personally as well. That's what open source is all about!

Also any source is better than no source and hacking together remixes of imprecise STL files then repairing the non-manifold mesh through NetFabb.

So I hope you'll reconsider releasing your source :)

I am currently building this printer and made lots of modifications. One thing I am curious about is how many fishing line windings should the nema17 pulleys include to not the fishing line to slip during printing process?

Hi S3rkan, I run the printer with one and a half loops as described in my guide, it works fine for me, even longer prints showed no sign of displacement due to slipping. I have seen other printers with more windings, but I was afraid that overlaps can happen with negative results as well.

Almost finished: https://www.dropbox.com/s/50611u75cgs8ygc/IMG_5717.m4v?dl=0 . One question though: How should I go about calibrating the stepper motors for this? It looks like everywhere I've looked for it's just an explanation of why CoreXY is better and HBot style printers should be updated to CoreXY.

I'm using Nema 14 for X, Y and Z axis and Nema 17 for the extruder. I know (I think) how to calibrate the extruder as I've already done it on my delta. Z-axis, I think I'll be able to manage a bit (however I've never dealt with lead screws or screw-based axis before). The main Issue I'm having is that the hotend doesn't travel the whole bed area (not even the 120x115 mm that it's supposed to). I've fiddled with steps per unit, but not sure if there's some math I should be doing instead of trying to guess.

Any help is appreciated!

Looks very nice, just one thing, I don't know if it comes from the video, but it seems that the pulleys at the motor are not 100% straight, this may result in non straight prints. I love the z axis, I hope find the time for a remix to improve mine too.

About your question, I recommend the 66.5 from my sample config for Nema, from this point calibrate it like described in many guides (http://reprap.org/wiki/Calibration section "X & Y scaling and steps/mm calculations"), control the axis for a defined value, measure the actual movement, calculate the deviaten in percent and add/subtract this percentage to the steps per unit. But I wonder why you are so far off, perhaps try to tighten the fishing line or add an extra loop around the pulley to avoid slipping.

About the calculation, here is the theoretical part:
Pulley diameter 15mm
Steps per rev 200 (1.8° motor)
Microsteps 16

Steps per unit = Steps per rev x Microsteps / Diameter x PI
Steps per unit = 200 x 16 / 15mm x 3.1416
Steps per unit = 67.91

I hope this helps you


Also, I took some pictures and it should be more clear that they're in fact not straight: https://www.dropbox.com/sh/b3yrs62hnjluium/AAAy6SGO5uIM82nzjfOnJ86Ka?dl=0. How is that happening and what can I do to fix it?

Also the fishing line I'm using currently stretches a lot but I'm waiting for another braided one, a proper one I hope.

Fishing line (or whatever you use) is not allowed to stretch at all, this may already be the root cause for slipping.

I could only guess why the pulley is not straight. The stl off course is designed to be straight, that means you have an issue with printing them properly, from my experience, it might be a problem with the cooling, at small parts it could happen that the previous layer is not cool enough before the next layer is printed, this leads to deformed results. You can try to improve cooling with a fan, reduce the print temperature (if possible), reduce the printing speed to cool down the layers or print the part multiple times, then each one has time to cool down while the same layer is printed at another part.

I think I found another issue I was having: It appears my Nema 14 are 0.9 degree stepper motors. I changed them for Nema 17s I had and the distance travelled with your marlin settings seemed to be on track.

The only problem now is the Z Axis which I need to calculate the steps per unit. I've tried multiplying by two (81516) steps / unit but it seems that I still get too large distance travels.

How did you come up with the 40758 steps/unit in your config?

If you have a different stepper angle, you have to recalculate the steps per unit, thats why the 28BYJ has so high values, the gears increase this value dew to the ratio of ~1:65.

About the z axis, in my example, the BYJ is still used for z-axis (I want to change that too if I find the time), so you have to recalculate by the formula:

Thread pitch = 0.8mm (M5)
Steps per rev Nema 17 = 200
Steps per rev Nema 14 = 400 (your case)
Steps per rev 28BYJ = 32606.1824

Steps per unit Nema 14 = Steps per rev Nema 14 / Thread pitch
Steps per unit Nema 14 = 400 / 0.8
Steps per unit Nema 14 = 500

You should also increase the speed and acceleration values

Oh man, I'm really impressed right now! This one is really cute! What an amazing job, congrats!

Thank you very much, you're far too kind

Just an idea I had if anyone wanted to make this a portable printer with heated bed.... would build one but half way through another build


Any way we can get the gerbers...and upload to OSHPark...seems that this is in British pounds vs American dollars
And/or can you increase the size of the heat area to 150x150?
This one is 120x120, but when you get larger boards from OSHPark, it get's expensive fast!

How is this 3D printer accurate?

Actually I can't tell, in the 28BYJ-48 version accuracy is limited due to the mechanical play of gears, with Nema17 accuracy is comparable to any other printer. Have a look at the sample prints. By calibrating the printer the results are very good, but I can't state any numbers

I want to make 3D assembly in solidworks. How I can convert models in STL. to SLDPRT(solidworks) or other.

Use the open dialog and chose .stl then go to options and chose volume model and open a part. That should work according to Google:

Nice!!! Sehr schön


Progress update (gallery with videos): https://www.instagram.com/p/BWlGUqoFUId/

Finished the design for the z-axis (wider, with 22mm bearing and NEMA 14 mount) and got the firmware (Teacup on an Arduino Uno with a CNC shield (to stay true to the whole concept)) running.

Comments deleted.

I have already been asked in the comments about using an Arduino Uno + CNC shield to further reduce costs. Would you share the firmware for easier use? I could add it to the MicroCube Thing or you can create a remix which I could point to.

Since I'm going with a belt-driven NEMA14 setup, I have to redesign basically every part (already finished a good chunk (using OpenSCAD)) - sharing the source files (as a remix) and the firmware was the plan all along. :)

But for the firmware/config files I'll be going with a git repository. This way it's easier to document and keep track of (possible) changes.

The last missing hardware parts I'm currently waiting for are the idler pulleys. Once they have arrived I can finish the axis and start tweaking the firmware settings (motionwise).

Ok, cool. I also have been asked about a belt driven version. Will you keep the h-bot style or move to core-xy?

Nice, I'm curious how the core xy will look like and how it will perform

I'm going to try a core xy setup (for reasons). We'll see how that goes.

What are you thoughts about using this printer design with less cheap components? I want to build one with a heated bed, an e3D Titan with the NEMA17 upgrade.

I am running the printer with the Nema17 motors for a while now and I would definitely recommend it, the speed of the 28BYJ-48 motors is simply to slow. Better components will lead to a better performance, the Titan might not reach it's full capacity, since the speed is limited by the h-bot style, but in this case you might use it in another project. I am running the printer with a Titan and E3D clone and it works pretty well.

Can you tell me what the spacing is for the carriage rods? I have a serviceable CoreXY printer at the moment, but the carriage design is lacking, and I rather like the one you've designed here.

The rods are 36mm apart from each other

Got all the parts, about to make one, but I have a question: Can you use the cheap 28-byj motors for the xyz movement, and use a nema motor for the extruder?

As OkurRo mentioned, yes you can. Actually, the first motor I replaced with a Nema17 was the extruder since the solution with two 28-BYJ motors is quite weak and slow. You need to adapt the steps and you can also increase the speed in the firmware.

how do i increase the speed in the firmware, which line?

Have a look at the description or Marlin in general:

define DEFAULT_MAX_FEEDRATE {30, 30, 0.25, 5} // X, Y, Z, E

For Nema motors, you also need to change the steps values and acceleration

Yes you can, only steps per mm will be diferent on extruder motor

Awwww its so cute!!! I might have to make on but make the Z axis taller.

Extending Z axis should be no big deal, simply extend the length of the vertical rods. But don't extend too much, rigidy is based on the small size.

yes i know, I'm keeping that in mind .

What size are you aiming for?

Z of about 175 or 200, i know that might cause some structural issues but ill think of a way to brace it.

I wish you did something for nema14

I thought of using Nema14, but they were quite expensive to get here in Germany. But if you like, I could add a bracket for the X/Y motion system, its pretty simple.

I edited the Nema17 holders, I added the holes for Nema14 motor, now you are able to mount both motors on the bracket, you can even exchange them without cutting a new fishing line. Please let me know if you give it a try, I am interested in the result since I thought of using the Nema14 myself

Thank you that would be very nice

Hi I have access to aluminium shaft (it's not hollow), can I use it? I asked because I saw a hole in the corners that seems to go straight into the tube, what is it for? Mounting the motor?

EDIT: Nvm, it's for the top bearing, so I guess the motor and bearing mount doesn't have nut trap?

Hi, yes you could use solid material as well as a hollow tube.
For the z motor no nut trap is used since the force is quite low to keep it in place. The bearing is held in place by the same screws + nuts which are used for z-axis rods.

I thought of lowering the price by using uno, cnc shield and relay for hotend. Do you think that it's possible?

Marlin is not compatible with the UNO, however there is a firmware which is designed to run on an Uno (and other controllers) https://github.com/Traumflug/Teacup_Firmware.
So yes, should be possible, but a little additional hardware needs to be added because the CNC shield does not provide a driver for the hotend.

Do you have any resources for adding a hotend to a CNC shield? It would be nice to have a cute small controller to go with the cute small printer :)

Also you could use carbon fiber tube for the frame and even the linear slides, they need either Igus bushings or printed PLA or PETG bearings.

Instead of adding the hotend to the CNC shield, I would suggest to use a MKS board which combines the Arduino Mega and the Ramps board in one single pcb (http://www.ebay.de/itm/New-MKS-Gen-L-3D-Printer-Controller-Board-Compatible-With-Ramps-1-4-Mega-2560-/162433300401), otherwise you need to copy the MOSFET schematic of the RAMPs board which will end up in a higher price and mor effort.

Carbon fiber is a nice idea, haven't thought of it yet. I decided to use the aluminum because it was easier to source and machine. I also wanted to keep it as cheap as possible.

I'm building one with CF rods and a few different parts:

I've got a flat Nema 17 (22mm length), 8mm threaded rod and would like to use gt2 pulleys and a single PCB from a Delta printer. I'm looking to mod the designs a bit and see if I can make it work, any chance you can release the source?

This is a picture of my progress so far (not a lot) https://www.dropbox.com/s/8jqulv1qizw4uuu/Photo%20May%2020%2C%2022%2011%2025.jpg?dl=0

Btw, I love the concept and your design! Thanks for sharing it!

Wow, looks cool, I like the colour scheme as well. I hope it works well with the carbon rods.

Which parts would you need for your redesign? I also got a pancake stepper, I would be interrested too to upgrade to a Nema17 z axis. I found that the platform is not very rigid, needs also an improvement.

For now I need the z axis motor mount, and the bed platform so I can fit the pancake Nema 17 with the bigger and heavier 8mm leadscrew. I still got some parts on their way from china so this is probably not going to be finished anytime soon, but I can let you know what other parts I need to modify as things arrive and I can progress.

As for rigidity, I fear the frame might not be heavy enough and that the center of mass being offset to the upper middle and close to the motors, due to most of the weight being on the back (as it appears in the photos).

I'm planning to move the motors to the front or filling in the front rods with sand trying to move the center of mass closer to the middle/bottom. Maybe some support pieces might also be needed.

Send me a private message with an email addresse to send the files. When adding the pancake, can you update the platform to accept both, M8 nut as well as T8 brass nut, I was planning to use one.

Rigidity of the frame itself is great, especially with the diagonal stabilizer, issue is the bed itself, I tried to reduce printed parts as much as possible, but the arms are to weak at the front.

Its also stands quite stable on the desk, of course, I don't know about the light carbon fibre version, I use a Laptop power supply which I want to integrate behind the display. This will make it more compact too.

Yeah I have MKS BASE and MKS GEN-L boards, but I wanted something even smaller to match the printer :) I was hoping there was some easy circuit to add a hotend to a CNC shield but I guess not. Oh well. A Melzi or RAMBo would also be a good choice. I like the all-in-one boards better than cheap RAMPS shields which tend to have poor quality.

Carbon fiber rod is getting pretty cheap these days due to quadcopter guys. HaoZhong Carbon Fiber on eBay reportedly have very good accuracy and consistency with sizes. It's cheaper for me in Australia to get carbon fiber than anodized aluminium. Cheaper still to get hardened steel smooth rod, but that's also very heavy.

I think a MKS or another single pcb would fit great compared to arduino combined with a shield which is more bulky.

Besides the price, I wanted to avoid cutting the carbon fiber, aluminum is easy to cut and file. But if you are going to build one, I would love to see a built with carbon fiber and a single board solution.

I also plan to add the Laptop power supply in the free area besides the Ramps board and an extension for the spool to get everything more compact.

I will definitely let you know if I do it. At the moment I've got the parts on order for a HyperCube, so I'll spend a few months getting that working well, then start on a new project.

I'd actually like to remix this design into a CoreXY, enlarge it slightly with 250mm tubes, and keep everything except the Bowden tube inside the frame. I've got a few other ideas like acrylic panels on bottom/back/sides for stability and for mounting things to and for the look, using a laptop power brick and buck converter, and having the build surface as boro glass with an easily removable clasp for convenience.

I'd like to use NEMA14s but high-torque ones are not cheap. It would make more sense to upgrade my main printer's motors, and use the old NEMA17s to build this thing.

Everyone wants to go big these days, here we are talking about smaller printers :)

Very cool. I will make this soon.

On the parts list you say -

6 x Smooth rod 8mm x 200mm
10 x Aluminium tubes 8mm x 200mm (or other material)

I have smooth rod, but what is the tubes? Where are they used and can you provide a source (aliexpress?) for them?

Thank you,

The tubes are used for frame. It can be any 8mm pipe. Alumminum are better for frame so just try hardware store, or some store for house and garden.

The smooth rods are used as part of the frame as well as for the linear motion of the x/y/z axis, for the other frame parts, I used cheap aluminium tubes with 8mm outer dimension from a local hardware store, but you could use any material you want (aluminium, steel, carbon fibre, etc.), solid or tube. Amazon link: https://www.amazon.de/gp/product/B00ZXPJ0QC/ref=ox_sc_act_title_1?ie=UTF8&psc=1&smid=A1P3ZJ39C3CUTD

Thanks. I am in the US. 8mm kind of hard to get. I look for conversion and looks like 5/16 is pretty close.
I will keep looking for 8mm.

K&S makes 8mm aluminum tubes. Check a local hobby shop for them (though those in my area only ever had 1 or 2 in stock), or you can get them online, from the likes of Tower Hobbies etc. They run ~$2 for one 300mm tube.

If you are in region where the imperial system is used, it is also possible to use 5/16 rods, the corner clamps should work as well as with 8mm

Maybe try call a metal company, I am getting the bar from an aluminium extruder manufacturer. Maybe this will be easy in the US. Tubes will be hard to find, but solid bar should be easy.

As I am from Germany, metric system was used. But I think the 5/16 should work well, there is enough tolerance, the corners are like clamps tightened by the screws. If it should not work, use some kapton tape around the rod to fill the gap.

I wish we would switch to the metric system....

Think I found something, I'll get these - http://www.ebay.com/itm/271607850532

Tubes are almost here, time to start printing the parts!

Printing last part now.. Looking over the instructions, are you consistent with your use of Tube versus Rod or did you mix the two?
For example, step 1, assembling the frame, 4 x Aluminum rod 200mm. Then it jumps to talking about Z axis rod, which should be smooth rod. I want the frame to be tube.

I did a search for Tube in the instructions and found something around step 12 or 13 and 16 only.

I can probably finger it out but thought I'd ask for clarification.

All parts of the frame which are not used as a linear guide, are made of aluminium tube. The linear guide requires smooth rods. Have a look at the pictures, you can see the difference between the two types.

Printer is coming together. Need to wait for more parts (bearings) to arrive. This is a nice and simple design, really happy with how easy it is coming together.

I broke the X carriage applying the zip ties on one side but not the other.

Couple questions:
Any sources for the print bed?
Regarding fishing line, it has been many many many years since I have been fishing, It seems that kite string might also work for this. I had some of that around here somewhere. In the event I can't find it, what type fishing line did you use?

I'd go with Spectra line.

There is nothing special about the printbed, I used a piece of polycarbonate, other material are fine too, I got it from the local hardware store.
I just ordered a fishing line from Amazon, I guess kite string would be good too, it is just important that the material is not flexible to achieve precise positioning.

MADE!!! I posted a make. This was a fun build and everything was working GREAT even dimensional accuracy, THEN I HIT RESTORE FAILSAFE SETTINGS in Marlin firmware and now the extruder does not seem to extrude enough material. Can someone help? I am using the NEMA17, 26t pinion. Using values documented in PDF file, Motor speed settings (Nema17 version):

#define DEFAULT_AXIS_STEPS_PER_UNIT {66.5,66.5,40758,96.1}
#define DEFAULT_MAX_FEEDRATE {120, 120, 0.2, 10}
#define DEFAULT_MAX_ACCELERATION {300,300,1,20}
#define DEFAULT_AXIS_STEPS_PER_UNIT {66.5,66.5,40758,96.1}
#define DEFAULT_MAX_FEEDRATE {120, 120, 0.2, 10}
#define DEFAULT_MAX_ACCELERATION {300,300,1,20}
#define DEFAULT_XYJERK 5.0 // (mm/sec)
#define DEFAULT_ZJERK 0.01 // (mm/sec)
#define DEFAULT_EJERK 1.0 // (mm/sec)

These values also appear when I look at the EEPROM, everything matches, wish I backed them up before I overwrote them. hehe

Fixed extruder by setting steps/mm to 100.00, after calibrating.

Still one thing that is wonky -

I have max X and Y set to 150mm.
In Cura (have not tested anything else yet), I have print bed set to 150 x 150.

When I save gcode and print the file, the output is not centered on the bed, but basically Y min and X min (top-right corner, opposite home).

See my make!

I don't know why the extruder setting stopped you printer from working, but I'm glad you solved it.

About the print bed center, I don't know which picture you mean, but I think the reason might be that the printer is only capable of printing at an area of 120mm x 115mm. If you set 150mm square, the calculated center will be off the actual center. Try to set the actual values of the printable area and check again.

Did the benchy test print work fine?

I thought of Nema Z axis as well, I bought an flat Nema17 already, but didn't find the time to implement yet

oh a pancake stepper, of course.

I will use 120x115 dimensions are retest.

The benchy prints GREAT on this printer. See my Make.

I tried printing a bigger object and the printer drifts to the right. Is there a bug in Marlin RC8 Bugfix?

I think I saw you were using RC6, maybe I will switch to this. Can you confirm what version you are on?

Actually it'S RC5 of Marlin, I also realized this shift, I increased the tension of the fishing line a bit, this reduced the shift. I suspect a slight slip at pulleys, maybe it would be better to use more loops at the pulley, it's also possible that the chamfer was no good idea. Still need to do some testing, but at the moment I focus on other things like finishing the MPCNC in the basement ;-)

I turn up current going to steppers and printer is fine!

Thank you for the feedback, guess I have to recheck the current. Glad to hear that it works fine now.

Very good project! I love small 3d printer
Is the h-bot sistem good? I read on internet that someone have problem with this configuration, but I think if you print slow there is no problem.
I will made this project with this cheap stepper https://www.aliexpress.com/item/2-Phase-4-Wire-35-Stepper-Motor-0-9-degree-20mm-3D-printer-Stepping-Motor-35H20HM/32677700359.html ( 4 of this cost only 20$ ), Do you think that they are good?

Thanks for sharing this project!

Thank you. I haven't had issues with the h-bot style, I also think the combination of slow speed and the small size makes it less vulnerable.

The steppers look good for x/y, but for the direct drive extruder it will be to weak. But perhaps you could use a different geared extruder, then it might work.

I'd love to see your result in the made section and also a comment about the small Nema17 motor performance

Hi, I have done the bipolar hack with steppers but only one is working rightly. Most of others are now moving only in one direction and moving only every third/secong step. The other one is moving right and left without me changing direction, but only in about 20 degrees. Did anything similar happened to you?

Hello, I didn't have such issues. I have done the hack with five motors and they all worked fine.

Try to check the following points:

  • Is the connection completely interrupted? Do you have a multi meter to check?
  • Are the coils connected correctly? Compare the wiring with the working one.
  • Have you flashed a firmware with correct steps/mm, to fast might end in unexpected movement
  • Have you set the current of the drivers to the values as described?

I hope this helps in finding the root cause.

Thank you, it was wiring and damaged driver.

Glad to hear that you fixed the issue. How far is your built? Is the guide detailed enough for setting up the printer?

I have added video of first tests.

Yes it's detailed enough. I'm still waiting for delivery of endstops and i don't have the buildplate yet. Right now i am placing fishing line.

im building a micro cube just wondering Motor_mount_Nema17_r file is there a left hand file as well or do i just use this one twice


I forgot to add to left one, please check the files, I added the second holder. You would to mirror the part, but they are not the same due to the mount holes. Thank you
I'd love to see your printer at the made section

What an incredibly engineered little bot! Thanks for sharing

Thank you for your kind words

Hi, could you please add some scheme about wiring of fishing line? And how come the fishing line isn't sliding freely around the motor wheel? Is there any specific one which is sticky or something?

Hi, I am about to write the setup guide, the assembly will include a step about it. The scheme is quite standard H-bot, you can find a lot of examples in the net. It's a standard fishing line, I simply wrap it one and a half time around the coupler and with the required tensioning it does not slip at all.

Edit 2016/12/27:
Have a look at the assembly guide, you can find the scheme of the fishing line in the description.

Hello, do you think it would be possible to NOT hack the 28BYJ-48 motor, but instead use drivers meant for this motor? http://www.ebay.com/itm/1Pc-ULN2003-chip-Module-Stepper-Motor-Driver-Board-Test-Board-For-Arduino-NEW-/272462871114?hash=item3f700d7a4a:g:ThMAAOSw5cNYOG0F
And if yes, do you know how to wire it into RAMPS 1,4?

Hi OkurRo,

I am afraid this will not work, there are two issues, first the stepper driver of RAMPs is designed for bipolar steppers only (motor is a unipolar model with a simple transistor driver, the ULN2003). The other issue is the fact that the stepper motors will not have enough torque, in unipolar mode, only half of the coil is used, while bipolar mode uses the whole coil by inverting the voltage. If you are afraid of hacking the motors, you don't have to, all you have to do is cut the connection between the two coils. It is very simple.

Comments deleted.

you could use acrylic sheet in the 3d printer as in enclosure and it will make it stronger

Yes you are right, but when I designed the printer, I tried to use only printable or easy to source parts to make it easy to build, acrylic parts should be cut on a cnc or laser cutter, I don't have access to like many others. I will try to add additional diagonal rods for the Nema17 version.to have more stability.

hay what is the name of the Ball bearings, 13mm outer, 11mm inner, 4mm hole, 4mm thick or from where you buy it

Hi, the bearing code is 604uu, I added the names of the bearings to the list. I found them in a German 3D printer online store, but you will find several stores in the internet selling them.

Hey, nice work! I don't know the limitations of these steppers but I guess that some lateral / back / bottom crossbeams would increase stiffness a lot and, maybe, allow some more speed.

+1 for video(s)

Thank you, the steppers are very slow because of a gear ratio of 65:1 they are already at their limit. For this speed the structure is well enough. But you are right, when I add the Nema17 upgrade, I will have to check the influence of the frame at higher speeds.

required nema 17 fixation and gt 2 belt fixation.
i m sorry bed english

As described, I will add a set of holders for the Nema17 for the XY axis. Belt would require a bigger rework which is not planned by me, I am not sure if the structure is strong enough, have a look at one of the existing core xy printers, they already offer the belt Nema17 combination.

Could you add a video showing the printer in action ?

Yes I will add a video along with some more information about how to setup the printer.