Hey! This thing is still a Work in Progress. Files, instructions, and other stuff might change!

Enlarged Rostock

by kolergy, published

Enlarged Rostock by kolergy Apr 26, 2013

Thing Info

25784Views 7855Downloads Found in 3D Printers
Report Thing


This is a linear delta 3D printer based on Johann's Rostock which has been enlarged to acomodate roughly double the print volume.

All of the 3D printed & laser cut parts have been made thanks to the Artilect Fablab Toulouse.

This is still work in progress so below you will find a blog like story of the build & set-up:

  • 30/08/2013 Discovered that one of the Idler_end started to develop cracks, therefore the design of the idler_end was updated to be stronger and as well for a better installation of the end stops.
    Updated Idler End design (V1.2) is there: http://www.thingiverse.com/thing:143186

  • 27/08/2013 Found an interesting improvement by Antonio Bencini which adapted magnetic bearings to the enlarged Rostock, this is really cool!

  • 23/08/2013 Back from holidays...

    • Following Zoltan's comment I realized that I had forgotten to put the platform STL, this is corrected now. Anyway it is unchanged from the Original Rostock one.
  • 09/07/2013 Presentation of the Enlarged rostock development (in French, sorry for the English speakers) http://prezi.com/slg0ddtggc2_/?utm_campaign=share&utm_medium=copy

  • 15/06/2013 Refined calibration

  • 09/06/2013 Some Updates

    • Demonstrated the printer at the Artilect Fablab Toulouse on the 03/06 (Picture of it being re assembled there). http://artilect.fr/
    • Added cable tentioner to allow speed increase without shacking
    • Insulated the hotend to be able to perform efficient cooling without affecting hotend temperature http://www.thingiverse.com/thing:101104
    • Chased play in the wade extruder drive.
  • 25/05/2013 First Perfect Print !

    • As on small part there were thermal issues I decided to test larger prints.
    • I went for the 6 sided Vase from
      MaakMijnIdee http://www.thingiverse.com/thing:18672
    • Printed with blue PLA 0.2mm layer thickness, 207degC for first layer & 197degC for the others, speed 40% of prusa std settings (5hour print time)
    • The result excellent :) It will be time to accelerate...
  • 22/05/2013 Biger & More precise but stil thermal issues

    • Slightly tightened the belts to improve acuracy
    • Went for the large Lucy the cat (in blue) with no ventilation & still at low velocity
    • Low parts where excellent with a good prescision of the layers
    • One can notice a problem around the head of the cat which give it a frog attitude. This was due to excessive softening of the material & partial colapse.
  • 21/05/2013 Post mortem & Back to life

    • Went to see the E-Motion Tech people with my dead J-Head and after post mortem analysis they figured out that the termistor reading was not correct it should be around 90k Ohms and was 120k. They where very kind & helped me to restore the head.
    • Now back to life calibrating properly the thermistor.
  • 20/05/2013 The J-Head lost it's head!

    • Trying to print another lucy the cat after tightening the belt to get a bit more prescision, Extrusion was not going well again. I tried to do a manual extrude, temperature was set @ 210degC, there it felt strange the extruder stepper's seemed to force significantely jumping some steps but the PLA was only going out lazily.
    • Sudently I the head started to move & poped out before I could stop the mains power.
  • 19/05/2013 Fighting with the Cat

    • Today I tried to print something beter than a cube but still relatively small & simple So I went for Lucy the cat http://www.thingiverse.com/thing:24255 which is small & rather low difficulty.
    • Yesturday wen printing the cube I realised that it required cooling so I installed a PC fan (which was the cause of follow up difficulties)
    • As you can see on the second row of the picture there has been several failures before getting something that looks like the STL. To make a long story short temperature is measured on the top part of the Brass part of the J-Head hotend, which hsa been a bit isolated with layers of kapton tape. however the tip of the hotend is not insulated and would get some degrees colder suficiently to reach very hygh viscosity of the PLA dramatically reducing the flow.
      and at the same time cloging the teeth of the filament pushing axis.
    • Increasing temperature by 10degC & cleaning the extruder was the solution.
    • Looking at the last Lucy printed it is not yet perfect the layers are not well alligned, so my next task will be to fight the play to ensure everything is alligned.
    • In addition I need to work on the extruder as for the moment the Wade's extruder sit on top of the platform limiting the movment of the printers and I have not yet ordered the tube to make the bowden extruder which was initially planned. As I start to realise that booth systems have different advantage & weekness. I guess the solution could be to devise an interchangable system able to do booth.
  • 18/05/2013 Geting a bit better...

    • Just made two more print reducing temperature & extruder feed and now it start to look as a very small cube 10mmx10mm. (see the picture)
  • 17/05/13 First Print!!!

    • After fighting without sucess to have Repetier working I guess I will need to get an atmega1284 to replace the 644 that came with the Gen7 but which lacks memory for Delta printers.
    • I finnaly manage to get a 10uH self to replace the 100uH original one and got rid of the brown outs
    • Over the week I had to transport the printer a few hundred kilometers by car and apparently it did notlike it as play started to appear on some components and notably the joints.
    • Very late tonight I finally manage to attempt a first print: a 10mm cube, there was to much plastic feed and the hot end was too hot, so the result is not great see the picture and the video http://youtu.be/8jBBFH91Mn0.
    • Still some work to do to have high quality prints. Any way I'm very happy to be finnaly able to have a first print.
  • 12/05/13 A bit frustrating

    • I finnally managed to have the repeter firmware heating the hot end (the temperature measurement will need calibration)
    • For some strange reason it needs to be on the bed sensor plug & not the extruder, I guess I'l need to go & see in the Pin layout of the firmware.
    • But with repetier I do not manage to get the axis moving !
    • So I have the axis moving on the Marlin & extruder heating on Repetier...
  • 11/05/13 Trying to get it hot

    • Assambled the hot end
    • Assambled the extruder (Wade) initial plan was tu use it deported (Bowden) with a Teflon tube to link it to the hot end but I do not yer have the tube so it will be mounted directly on the platform. limiting speed & amplitude in the mean time.
    • With a friend we assembled the hot end to the extruder and then tryed it on the board. We were quite disapointed to see that there was still a 0degC temperature reading. I just realosed now that on the 06/05 the quick fix that I used to stop the broun out problem disables the temperatures. and I could not yet purchace the termistance to definately correct the brownout problem.
    • Therfore I tryed the Repetier-firmware with the lower memory settings & it seems to work. (it still needs to be configured properly for the rest.
    • Tryed again the heater with the Repetier-Firmware and there it was displaying -55degC which felt strange as earlier we tested the thermistor resistance & it was around 114k Ohms.
  • 10/05/13 Plenty of problems & some solutions:

    • After all the problem described bellow it finaly made some first decent drawings see the video: http://youtu.be/04nmR4zaZQw
    • Marlin Configuration Files & Test STL used for the drawings added to the downloadable files.
    • Found the solution for the homming problems from yesturday Had to reduce Max speed in the firmware.
    • Still one of the axis was not moving properly. At first I thought it was losing steps so I tried to increase the Driver's Current without sucess, I even switched the Drivers without results, to finnaly find that it was the pulley that was not tight enough.
    • Once that was solved I went back to drawing experiments and for some reason the drawings were to small & the overal movements where shorter than expected, after verifying all the set-up of the firmware was more or less correct found one error on the DELTA_CARRIAGE_OFFSET which was 22 on the original rostock & now is 28.0mm but this did not solve the main problem. Therfore I started to measure the belt pitch which was 2.5mm as expected and I went to count the numbers of teeth and there I understood where the problem was comming from: The pulley which I thought was an 18teeth one was in fact a 16teeth one. Once this corrected in the firmware It gave much better results.
    • Still some issues of steps lost from time to time
    • Glonally it looks rather sensitive to the speed the 400 steps per revolution & the 16 teeth T2.5 pulley means that it is hitting the maximum steps/s limits of the arduino
    • On top of that there is some issues of repetability of the starting height due to the endstops set-up which is not yet perfect
    • With all this I did not get a chance to try again the Repeter Firmware.
  • 09/05/13

    • Installation of end-stops
    • Start of calibration
    • It made it's first drawings (It is not great for the time being)
    • Some issues with homming: when moves comands are sent it moves correctely but when homming the steppers vibrates and nothing moves. I assume it is a error in my ste-up of the firmware, or too low current in the stepers drivers TBC.
    • Had some feedback on the issue with Repetier http://forums.reprap.org/read.php?267,206320 which should come from an excess usage of memory so I will try to lower the memory useage as advised and I will let you know.
  • 07/05/13 Platform fitted & moving

    • Diagonal rods Cut from The carbon Arrows & glued to the Jaws
    • Got some extra MDF to rigidify the structure at some later point
    • Fitted Diagolal rods & Platform to the Printer
    • Tried some movements: it is not perfectely flat yet nor rapid but it is going the right way. Here is a video of some of the first moves http://youtu.be/qAyjTx0MV_U
    • The Z stepper seems a bit week I will need to adjust the power on the driver.
  • 06/05/13 IT MOVES!!!

    • Did not work on the structure today
    • Temporarly gave-up on Repieter Firmware and moved to the modified Marlin by Johann
    • Got Brown out issue on Marlin as described in the following thread https://github.com/ErikZalm/Marlin/issues/174 Got to change L1 on the Gen7 V1.4.1 as explained on the last post. I used the temporary fix from the first post as I do not require heating.
    • Had to invert the Pinning of the steppers for them to turn on the correct direction
    • For some strange reason the M80 comand needs to be entered manually
    • Now all the three axis moves corectly on delta mode!
  • 05/05/13 Now it looks like something

    • Installed The belts & Steppers on each axis
    • Assembled the Three Axes with the frame
    • Still fighting with Repetier Firmware to make it work in delta mode
  • 04/05/13 Start of the assembly

    • The three columns of the printer are now complete
    • All motors & endstops are properly wired & tested on cartesian mode
    • Configured & uploaded the delta mode of Repetier but for some reason the stepers do not move in delta mode. Still need to beter understand config...
  • 03/05/13 Back to the printer after a forced pause.
    Today I have been playing with the Gen7 & the stepers:
    • Ran the Test set-up with sucess
    • Installed the repetier firmware 0.82.2 & tested with sucess the cartesian mode steppers turning ok.
    • Started to set-up the delta mode of the repetier firmware but haven't had the time to test it today.

Enlarged version of Johann's Rostock adapted with the folowing features:

  • 400x400x500mm expected build area
  • 0.08mm theoretical vertical presision (400step/rev + 16 teeth T2.5 Pulley)
  • Carriages with self tensioning & play elimination

It is just the begining... Almost everything is available now it is time to start building.

To be Continued


Builded Parts:

Purchased parts:

More from 3D Printers

view more

Thing Info

25784Views 7855Downloads Found in 3D Printers
Report Thing


Liked By

View All


Enlarged Rostock by kolergy is licensed under the Creative Commons - Attribution license.

Give a Shout Out

If you print this Thing and display it in public proudly give attribution by printing and displaying this tag. Print Thing Tag

File Name



what went wrong during the first print?

Nov 11, 2015 - Modified Nov 11, 2015
kolergy - in reply to 3dNoScope

For the first print the filament extruder was not well calibrated meant that there was not enough material

The following issue was slippage in the extruder solved by increasing the pressure. The following design makes a better extruder -> http://www.thingiverse.com/thing:310705

Finally the most annoying problem is the retraction & how to have a strong attachment to the platter for the large parts. I want to try PTEG filament which should help for that

MiniStruder V0.3
by kolergy

Could someone laser cut the bottom and top plates ao i could build this? :)

The top & bottom part do not require a very high precision so you can make it with normal tools.

I do not know where you are in the world but you can look if there is a FabLab near you http://wiki.fablab.is/wiki/Portal:Labs

hi kolergy. I am currently printing parts for building this printer and about halfwat i noticed that the design uses 12mm smooth rods. i only have 8mm at hand. do you think its possible to use 8mm instead or would it become too unstable?

Hi 3dNoScope,

8 mmrods over a 1meter length, might be a bit flimsy, for the printer to be stand alone but if you make a support structure that rigidify the printer this might be ok at the end.

the rods are a meter each right?

anyone know the formula to find the measures of the arms in the event that the measures are not the same?

It is the tower radius * 1.44

measures which are arms?

Apr 2, 2015 - Modified Apr 2, 2015
kolergy - in reply to jhon_bishop

all the dimensions are in the configuration file, the rod length is 432mm for a 300mm print radius

cuales son las medidas de los brazos ?

do you have a github with all your revisions merged in?

Unfortunately not I have not been very tidy in revision control I just update a part when it is modified.

is this project still alive?

Jul 18, 2014 - Modified Dec 30, 2014
kolergy - in reply to cocktailyogi

Yes but now it is though the strong print project as I have used the frame of the Enlarged Rostock to do the Strong Print http://www.thingiverse.com/thing:329498. All the modifications made though the strong print are compatible with the Enlarged Rostock the only thing missing for a full update of the Enlarged Rostock with magnetic bearings is a new effector based on the one from the strong print but for a standard 3d printer head, I intend to do that once the strong print will start to work

StrongPrint - The DIY Metal 3D Printer
by kolergy

Hi kolergy,
your project looks very nice. Actually I am going to build a new Printer and I became attracted to your design because of the ground-level-build size. It looks very nice and stable. Before I build a Prusa I3 (unfortunately I sold it... it was my best printer). After that I have build a Reprap Morgan. That concept has to many drawbacks, which I realized as the machine was finished.... No I am building a new printer which shouls be my last one for now...
Whast do you think? What about the precision of the XY-plane? I am not talking about resolution, I mean the repeat accuracy. Is it good? Is it visible? Your pictures look good, but what do you think?
Are there any visible artefacts because of the kinemtic transformation?, like steps or arcs on straight lines?
I am fan of nonBowden-Extruders. Do you think it would be possible to mount a direct-driver extruder directly on the motion-effector? Or would it be too heavy?
Sorry for asking so much questions. But for me this makes the decision to participate in this project or not.
Greetings from the alps,

The Enlarged Rostock is very easy to build,
Concerning the prescision I was verry happy about the results, there is no directly visible curvature effect due to the delta transformation depending of your settings you can notice the apparition of moiré paterns on some perfectly flat surface wen doing high speed printing. I ptinted the Iphone holder towards what is the max speed of the printer http://www.thingiverse.com/thing:215369http://www.thingiverse.com/thi... you can see on the picture the moiré paterns. I think it is linked to the speed of printing as in the firmware there is a limit in number of commands that can be sent to the motors per seconds. The use of an Arduino Due based board like he Ramps-Fd should solve the issue as you can increase the number of secments per seconds.
Concerning the nom bowden it brings an advantage & two inconveninnts:

  • The volume of the motor might interact with the pinter arms reducing the build area
  • The weight of the motor & extrusion system result in slower printing speed
    • The weight of the motor & extrusion system result in all play of the head being set on one side with no apparent play on the head at resonable speed.
Iphone5 + Bumper Universal Adaptor
by kolergy

Hi there, great work on this re-adjustment.

I was wondering if there is a 3D printed version to the correct lengths I could use for the extruder motion rods?

Hi again! I've re-designed the vertical carriage to include a bolt tightened bearing sleeve, and upgrade to Traxxas ball bearings hinges!


I hope you like it!

Enlarged Rostock Clamp Carriage

Very good improvments thanks

HI there, could you please try designing a carriage to suit the traxxas ball hinges for the arms? There are a few designs, but they are for the standard 8mm linear bearings.


I'm working on a full system with magnetic ball hinges, however I have little time to devote to it therefore I'm slow. In the mean time you have the carriage modified by Antonio Bencini: http://www.thingiverse.com/thing:139264http://www.thingiverse.com/thi... Which has been done for the Enlarged Rostock.

enlarged rostock magnetic_carriage
by bencini

I will be making one in the future in Australia. I have priced it out to be about $520 worth. Very good value for the size print area!

Cool :)

As you see with my last post, I am working on platform and diagonal rods.
Why do you multiply diagonal rod length by 1.44 ( ie 300*1.44 in configuration.scad) ?
In original rostock, it is the same as tower radius...

Diving into delta cinematics, the reason could be to have a better precision, no ?
The longer the rods are, the lower moves of effector is (in latteral) for the same sliders moves...

The main benefit come from the fact that no axis is stacked on top of the other one with associated imprecision stacking.

The length of the rods has only an indirect effect on the precision. The length of the rods was chosen as such to maximize the build radius without increasing too much the tower diameter

For magnetic bearings, why not use geomag rod and ball ?

could be worth the try but I do not have any of these

I have work on a new bed:
same as you with rods embeded in plywood.

Enlarged Rostock by vicnet
by vicnet

Congratulations this is cool!

I try to use 1.8deg stepper instead 0.9 one that are more expensive.

How do you get expected z precision ?
16 teeth in T2.5 => pulley diameter = 12.7mm
0.9deg stepper => 1 step =r*sin a = d/2 x sin a => 0.098mm

PS: do you have my mail about your gen7 processor ?

You have to add the 1/16 Micro stepping on top of that.

Yes thank you I'l try to go today.

How micro stepping is applied ?
(I don't find any information on the net)
Could we say that the equivalent step is 0.9 degree / 16 ?

Yes, basically it divides the steps, you have half steps, quarter steps, 1/8 & 1/16 first it allows your motors to run smoother and it allows to reach greater resolution.

Is the beds (down and top) are in MDF 12mm ?
(I saw 12 in conf scad file)

The top & bottom plank are in 10mm MDF but since then we found out that MDF was very bad for the laser cutter so I would advise you to go for plywood (a good quality one so it dose not bend with time).

As I did not know how i will attach the printing bed I made sure I had a lot of Anchoring points, these are the round holes in the bottom wood.

I used the same wood plank for top & bottom which was much simpler.
There is no specific SVG for the top as you can just remove the rounded holes which are unnecessary for the top & cut it like that.


Which hot-end size (nozzle and filament) do you use ?

For end belt (axe opposite to the motor), what mechanical system do you use ?
I see that it is possible to use a ball bearing with 2 washers (or metal rings) tfor guiding. Do you use this system ?


For the moment I use 3mm filament with a 0.35mm nozzle But I have an other one able to accept 1.75mm filament. I will probably use the second one with the Bowden system (Extruder Fixed feeding the wire through a PTFE tube).

Apparently the Use of 1.75 wire have been reported to be more prone to hot-end clogging.

Yes with a standard roller bearing. However it is not the same diameter as the pulley, an improved solution would be to have the same diameter for the pulley & the idler.

About belt and pulleys in instructions, you said that you bought 2m of belt and 3 pulleys: 75.06€ + 18.63€ !
Is seems too expensive for such pieces :-) or some vendors are thiefs !

Is it (2m belt and 2 pulleys) x 3 ? if so, I have a total of 92€ TTC almost as you.

As I saw in pictures, it is aluminium pulleys, right ?
Do you think that SLS pulleys could do the job ?

  • It is (2m belt and 1 pulley) x 3 as you have a column length of 1m and you need to go & come back with your belt.
    You only need one pulley per column on the motor as on the other side you just want a free rolling surface like a bearing. I will need to make this clearer in the text.

  • I do not know what are SLS puleys but you want your pulleys to be well made ensuring a high precision as defects (centering, roundeness) will have a direct effect on the position of the head.

In Instruction, you said that you take 1 carraige from original model.
I think it is the platform instead.
For carbon rod, you tell me that you found kite rod to replace arrow. Could you add reference in instructions ?
You mention Extruder Jhead but I think it is a Hot End instead (or an extruder hot end ?), no ?

Which model of extruder do you choose ?
Thank to you to share your model and instructions !

Thank you Vicnet for your good remarks, I have implemented the corrections.

Hmm.. I did not understand what was the problem with the electronics? Can the combo ramps 1.4 and arduino mega 2560 do the job properly (I have the two boards assembled so was wondering can I use them?)

Hmm.. I did not understand what was the problem with the electronics? Can the combo ramps 1.4 and arduino mega 2560 do the job properly (I have the two boards assembled so was wondering can I use them?)

Yes there plenty of Delta with Ramps electronics, and the 2560 should have sufficient memory.

My issue arised because my board came with a 644P instead of a 1284 which has very low memory (but it is still useable though as I'm still using it)
My preference went to the Gen7 because it has a slightly higher clock rate 20Mhz instead of 16. Bit I'm not sure if it really makes a difference

gen7 is said cheaper thant Ramps 1.4.
Ramps has more connectivity: one motor, one fan, optional LCD...

Thank you very much. Unfortunately, I could not find the platform stl in the files. Could you, please, check it? I can hardly wait to finish the printer. The parts came out great, and it seems to get a very nice machine. Thank you again.

Oops for some reason it did not work the first time. Should be ok now

Yes. It is ok now. Thank you very much. I really appreciate it.

Great job. I have printed almost all parts. They came super. I am missing the platform for hot head. Do you have the stl? I do not know to use Openscad. Thank you.

Now it is included, you might want to have a look at a contribution from Bencini who modified the platform & the carriages to include magnetic bearings. I have add a link on the update of the 27/08/2013.

do you have a email adress ? i have quite a few questions about tjis build- mostly about the rods and bearings that i have trouble finding in my country. would using 10mm rods and bearigns make the structure more shaky/wobbly when printing? please leave a mail adress or some form of contact I am trying to build this printer and have some other questions.theank you

  • 10mm rods: well from 12 to 10mm the difference is not huge, but yes it would make it a bit more wobbly, in fact when I started to accelerate the prints I had to add crossed stiffening cables with tensioners to stiffen the printer
  • For the control board I use the Gen7 V1.4.1 (now you can get V1.5 which would be better and allow to avoid the self issue) this is a full Arduino compatible bord, therefore no additional Arduino bord is required.
    • I went for the Gen7 because it is the only board running at 20Mhz (all the others are running at 16Mhz)
    • You would need to make sure it comes equipped with an Atmega 1284 as it has much more memory
    • Delta printers require more cpu & memory.
  • The best resolution that works well with the printer up to now is 0.1mm, I have tried 0.05m but failed to see a difference.
  • I left you a message with my email address.


Beautiful. I'm so jealous! BTW have you seen the magnetic arms for the delta style printers that are suddenly floating around here?

You mean this http://www.thingiverse.com/thing:93042http://www.thingiverse.com/thi..., yes I think it is very cool and interesting for two things:

- Strictly no play
- Quick change (as you mentioned)
aerostock delta magnetic effector
by aeropic

So have you built a filament extruder to recycle all that plastic? Sorry but I can't help but laugh at some of those cats. Still I love seeing your updates and watching as it all comes together.

Maybe one day I will build a filament extruder :)

But you know nobody provide images of the bad prints but it is very educative and helps others not to do the same mistakes or to identify what is going wrong.

In fact this is a bit why I made this page a bit like a blog & describe all the good things & bad things I had to face to build this printer.

I think that's why I like your project so much. I'm seeing just how much work it takes to get one of these up and running. I'm somewhat ashamed I bought a prebuilt Solidoodle. Still, I think my next project will be one of these. The delta printer style is too sci-fi to pass up.

Awsome! Another enlarged Rostock!

I modified the plywood .scad file from original to allow parametric selection of delta radius. (Default is 175mm) which allowed me easily to scale up the bed/top to 200mm delta radius. I'm curently assembling the arms and so far it looks very neat. It should give me a print area of 230x230 or a radius of 285 vs original 250. Not a big scaleup but enough.

Btw, a lot of people seem to read and take for granted earlier blogs and such saying that long bearings have no merit vs short ones for the LMxxUU type guides.

I decided to check that and I ended up using long ones. I discovered the bearings are actually specced for around 1 degree slop! That basically means that they will wobble back and forth +/- 1 degree maximum. I measured and found that all things considered, like the distance from the center of bearing to the ball center for the delta arms, I only get a fraction of that but still it ammounts of around 0.13mm slop with the 22mm offset.

The long bearings have so little slop I can't measure it. It was really noticeable when I started printing with very thin layers and fast movement.

Those cubes look like melted frosting! How accurate do you think you'll be able to get your final prints once you get your settings figured out?

The theoretical maximum accuracy is 6.26 microns in the Z direction, X & Y should more accurate.

there, I have an issue on the temperature selection, the wire feed
& there is bit of play in the joints that must be eliminated.

Great combination of parts. My guess is that you'll be needing some side plates for bracing the unit. The rods alone won't provide enough support for the printer.

My Woodstock Rostock uses 12mm wall sections and metal corner braces for stability. See:http://www.thingiverse.com/make:21459http://www.thingiverse.com/mak...

Yes at terms I will need to do something

  • in the short term I was wandering if just diagonal cables would do the trick

  • In the long term I want to fully close it to control the inside temp to 50-100°C+ as I'm sure this is the trick to minimize deformation on large shapes.

  • I'm not sure wood is best for that due to it's propensity to deformation.

I like your printer build it is nice & very clean.