Pinya3: a 3d Food Printer Platform




About: My name is Luis, I am from Barcelona and for the last years I've developed a curious obsession merging food and rapid prototyping. To have a unique and customized eating experience.

Pinya3 is an open 3d food printer platform design to fit in the kitchen and work with as many different types of food and mixtures extruders as possible.


  • Portable
  • Height under 50cm (~20")
  • Compatible with existing food compatible extruders
  • Open platform

Step 1: Frame Design

The printer configuration is based on a delta robot. Because it allows to have a fix printing platform, meaning that the food that is being printed does not move. The second reason is because this configuration permits easy access and disassembling moving parts (specially important for cleaning operation). But I can not fool anyone, the true reason is that I love how they move.

The desired printable area must have the diameter of a regular dish. But what does that really means? I am not sure so I head to the kitchen to measure different plates sizes.

ID (inner diameter) ranges from 1.6 to 9 inches (4-23 cm), OD (outer diameter) varies from 4 to 11 inches (10-28cm). The inner diameter corresponds to the flat surface on a dish, where we want to print. And the outer diameter is the opening size that the printer must have fit the plate inside.

The last one is going to set the dimensions for our printer because we want to be able to fit dishes easily into it! Knowing that the distance between delta towers must be at least 11 inches, we can abuse this awesome delta calculator to get a rough idea on the printer dimensions. If you feel like doing some kinematic calculation go ahead. I did this in the past, and turn out the building process is way more important to have a nice accuracy. Still food is about deliciousness not precision (yet).

Printer height is limited by the distance in Pier 9 kitchen between the counter and the cabinet around 50cm (20").

Step 2: Sourcing Parts

List of parts needed and where did I get them:

Attached to this steps are all the files for the custom parts that I used, I am still learning Fusion, do not expect pro design, I did what I had to do to get it done ;)

Step 3: Custom Printer Parts

There is a few parts that the printer uses that are made using rapid prototyping. Those are:

  • Upper corner
  • Lower corner
  • Angle corner
  • Electronics support plate
  • Endstop holder
  • Carriage
  • End effector

Being lucky enough to be at Pier 9 the the build. I wanted to try as many different options to make the different parts of the printer for fun and to keep learning. I used a laser cutter, a 3d printer, a milling machine and a water jet.

I laser cut the upper corners, carriages, end effector, endstops holders and the electronics support plate. By far is the fastest and simpler tool to use.

I used the water jet to cut the lower corners and angle corners out of aluminum. The lower corners were easy to cut, because the material that I had was already at a useful thickness.

I used a FDM type printer to test and compare angle corners with waterjet parts. I can not compare yet if its worth the effort to go for aluminum parts instead of plywood or PLA in terms of structure resilience. But I can tell that it takes loooooonger to make them. Cutting the stock, facing it, cut it with the water jet, file the tabs and drill the holes. I even had time for and small incident trying to smooth the smaller parts with the belt sander (no shortcuts) and rip off the belt. Also way harder to get a decent result due to the amount of different processes involved. Overall a great lecture.

Step 4: Building the Frame

First thing I did was to cut the V-Slots profile into 300 and 480 mm beams.

  1. Insert bolts and t-nuts into one angle corner
  2. Slide the two of the end t-nuts into one horizontal beam each
  3. Insert 2 t-nuts on each horizontal bean in the upper face
  4. Align those t-nuts with a upper corner plate and screw bolts into each one of them
  5. Flip the corner
  6. Insert a vertical beam into the last t-nut free from the angle corner
  7. Tight that bolt enough for the beam to be hold in place
  8. Flip the corner and insert the vertical bold that goes through the upper corner into the vertical beam
  9. Slide horizontal beams towards vertical beam until they touch and tight the angle corner bolds
  10. Tight the four screws holding the upper corner with the horizontal beams (pineapple face)
  11. Repeat from 1 to 10, for the other upper corners
  12. Assemble the wheels and the carriages following openbuilds indications
  13. Slide in 2 t-nuts into one of the vertical face inner faces. One for the endstop another for the belt idler
  14. Slide a carriage too
  15. For the lower triangle frame we will assemble all 3 corners at the same time
  16. Insert corner angles into the horizontal beams
  17. Slide t-nuts into the horizontal beams to hold in place the lower corners
  18. Place the lower corners and hold them with their screws
  19. Slide the angle corners free t-nuts into the vertical beams (place on top of the inverted stool structure)
  20. Thread in the nuts holding the lower plates and the vertical beams lower ends
  21. Tight the other screws holding lower plates with the horizontal beams make sure make sure the ends of the horizontal beams touch the vertical beams
  22. Last tight all the angle corner left screws
  23. Insert the t-nuts to hold in place the electronics support plate
  24. Align the electronics board and screw it to the upper triangle frame but into the interior face


I decided not to use the aluminum angle corners because I felt the 3d printer parts were more accurate and had enough strength for the structure.

Step 5: Ball Joints

The end effector is connected to the towers using arms. At the end of each arm a joint with at least two degrees of freedom needs to be used. For that I decided to use ball joints hold in placed with magnets. This allow to assemble/disassemble the end effector just by snapping the magnet ball joints. What is cool and fun, a WIN-WIN.

The joints have 3 parts, the magnet, the magnet socket and the ball stud.

How to build the ball stud?

Prepare the M3 10mm bolts and the 12 mm diameter chrome balls. First lightly sand their surfaces before we glue them together with epoxy. Sand the bolt head and a small area of the ball (the size of the bolt head area). Once they they are ready, I place the chrome ball on top of a magnet to hold it in place. Soak the bolt in well mixed epoxy and place it on top of the small sand area on the ball. The magnet will hold the bolt too and help pressing both together.

If you have the chance glue together some extra joints. Is not likely to happen but if one of them break is good to have a spare one around.

The magnet socket also has tight inserts for the carbon fiber rods. First I tried to be cool and print the first test using the Objet Connex 500. Nonsense. I went back to a simpler tool and used a MakerBot to iterate 5 times the part until I had the design that fit perfectly, both the magnet and the arms (in less than an hour)!

Great now we have the ball joints ready.

ALERT! If you use the fusion file to print your own, please make sure to tweak the dimensions to got perfect fits. This will completely depend on your printer tolerances.

Step 6: End Effector and Arms

First we will use three M3 nuts to hold the ball studs to the carriages. After that we will screw them to the metal carriages. We will follow the same procedure to attach the ball studs to the end effector.

To make the arms, first we cut the carbon fiber rod in 20cm parts. Before I did some straw tests to see if I was happy with the 20cm arm length, and I was :). After that I press fit the magnet sockets into both ends of the carbon fiber tubes.

Step 7: Towers Movement

First attach the motor bracket to the upper corner. Afterwards mount the motor on the bracket and insert the pulley into the motor shaft.

Insert the idler assembly into the lower t-nut of the vertical beams.

Mount the Endstops into their holders and fix the holders to the vertical beams using the upper t-nut within the vertical beams. Make sure to lock it all the way up on its closest position to the motor.

Cut the belt to around 40" (around a meter). Brace the lower belt gripper and hold the belt with your fingers. Lace it with a zip tie and lock it in place. Run the belt around the idler and through the motor pulley and repeat the zip tie operation with the upper belt gripper. Add the belt tensioner as closer as possible to the upper zip tie, to avoid collisions with the motor pulley.

Step 8: Issues Along the Way...

Some design issues appeared during the assembly face.

The first one was the carriages that holds the ball studs had no good way to grip and hold the belts, to solve that I designed a new carriages and 3d printed them instead of laser cut them.

The second one was the Smoothieboard orientation. It fit perfectly, but sooooooo perfectly that I could not plug the usb connector! I redesigned the electronics support plate, also adding some hand grippers to move the printer around easily.

The third one was the endstops holders. The distance between the endstop and the motor pulley was too small. I could had solved this using the laser cutter, but about that time I had no access to it, so I just went for a new 3d printed part.

The fourth was the order to mount some elements. Make sure to mount the motors on the upper corners before installing the upper corners otherwise its gonna be hard to screw everything together. Same for the electronics support plate. Is easier to hold the power supply and electronics first and later assembly the plate to the frame.

The fifth one, was the decision to press fit the magnet into the magnet sockets. Eventually they started to wear out and move. So I decided to glue them all.

Step 9: Electronics & Smoothieboard Configuration

The electronic part of pinya 3 is quite simple, in terms of connections. Make sure to match the motors connectors with the same axis connector for the endstop. Connect the smoothieboard to the power supply. A USB cable to the smoothieboard, and we are ready to have fun... setting up our smoothieboard parameters to fine tune the printer.

I could try to explain you how to adjust your delta values but it would never be as good as the Smoothieboard documentation for delta printers. Really, there you have all you need to make it work.

Step 10: Test It

For the first test, I was interested in using already available extruder (on the quest to make the printer as open as possible). That is why I decided to do it with a Discov3ry extruder. First I need it to set it up. Paying special attention to the steps per mm configuration value. For that we loaded the extruder with water and see how a software "push" of 10 mm translated into water displacement. Using a ruler I kept measuring the water displacement and adjusting the extruder steps per mm until we got the displacement right. I settled for a value of 6775 steps per mm. But this will totally depend on your printer setup, microsteps and so on.

Also if you are planning to use different types of extruders and motors pay attention to properly set up the current for the motors. Smoothieboard allows to change that through a GCode so for the Discov3ry extruder I set it up to 2A.

After this basic configuration is time for some basic fun!

Using curv3s we designed a basic curvy shape.

Step 11: Stainless Steel Plate

After using the printer for upcoming instructables. I realized the importance of having a decent plate to print on. By decent I mean that is easy to clean, strong and food safe of course.

Having this awesome opportunity at Pier 9, I could not do anything else than go big and build a stainless steel plate.

I bought a sheet of 304 stainless steel 18"x18" with a thickness of 36 thousands.

  • Designed the plate with fusion 360 to match the lower triangle
  • Cut it with the water jet
  • Cut the holding tabs with the band saw
  • Filed the sharp edges
  • Mark and brake the corners
  • I finished the part polishing it

File added as dxf.

Step 12: Conclusions

It has been a great exercise to start truly thinking how to design a printer for the kitchen. Along the way I realized many of the decisions were technology based rather than scope wise (printing food in the kitchen). I would say that is one of the biggest lesson from Pinya 3. We could argue if this is a simple delta robot placed in the kitchen. I like to think that is not just about what it "looks" but also about the ideas that triggered along the building process. Lots of new sketches for Pinya 4 pop up along the way... But before that, first we should abuse Pinya 3 for a while, keep learning and having fun. That is what 3d food printing is about (for me)

If you read the whole thing, it means you must be interested in 3D food printing, if that is the case you can read more about what I do at 3DigitalCooks.

Thank you!



    • Sensors Contest

      Sensors Contest
    • Planter Challenge

      Planter Challenge
    • Growing Beyond Earth Maker Contest

      Growing Beyond Earth Maker Contest

    26 Discussions


    1 year ago

    Hi there,

    Just wondering how long this took you to make and how complexed the manufacturing process is?

    Thanks, Anna

    1 reply

    Hi! Thanks for your comment, I dont really remember, but putting together the printer once having all the parts would not take that long. Maybe a couple of days or so. I hope you give it a shot, at least to 3d food printing. It is fun :)


    2 years ago

    Hi there,

    i would like to know if you have printed anything else with it?

    what other ingredients can i print with it and what is it depends on?

    what are the limitations?

    thank you

    1 reply

    3 years ago

    May i ask about your instructablse journey? i see you join recently in jan last year, I am curious in how your passion lead you from Barcelona to SF Pier9 !! i am from Hong Kong and am applying for the coming summer internship, i am dying to pump out my 3d idea with the awesome CNC!! !! May i ask about the international presence too?


    3 years ago

    May i ask about your instructablse journey? i see you join recently in jan last year, I am curious in how your passion lead you from Barcelona to SF Pier9 !! i am from Hong Kong and am applying for the coming summer internship, i am dying to pump out my 3d idea with the awesome CNC!! !! May i ask about the international presence too?


    3 years ago

    How would you print something say like Chocolate? Or would the materials used for printing have to be semi-solid but squeeze-able through the extruder - and not temperature dependent?

    1 reply

    Reply 3 years ago

    Chocolate is tricky, also depends a lot on the type of chocolate. Controlling temperature is a necessary. But depends on what you want to achieve, I would look into other people printing chocolate and learn from them. That's the best way :)

    Thank you


    3 years ago

    What would you change to make this a FDM 3D Printer, for use with filament?

    1 reply

    Reply 3 years ago

    I would add a plastic extruder a heated bed. I would also add autoleveling, and I would had make it bigger. For me, I would had taken a different approach, but definetely could work.
    Thank you


    3 years ago

    Well, there is a 3d food printer that can do the same, but a lot faster and a lot cheaper.

    1 reply

    Reply 3 years ago

    I think beating the human hand motion is a tough one, specially if you add our amazing closed loop vision system! I would argue though about repeatability and the cheaper factor. I think this are too different approaches. For me one cool thing about 3dfood printing is that allow to more people getting involved designed for food, without having to be masters at piping or decoration. But I agree that there is different ways to do the same :) Thank you


    3 years ago

    You said nothing about what this uses for ingredients. Is it some sort of synthetic slurry, powder, or ??

    3 replies

    Reply 3 years ago

    After a cursory search for "3d food material" I found this-

    It's hummus paste. Seems that anything thick enough to not drip out of the extruder could be used, icing, flour and water mixtures, blended peas, refried beans, jello, jelly. But that's just my opinion, I haven't made any 3d printed foot at this time.


    Reply 3 years ago

    Hi! My belief is that anything can be printed if you have the right approach. That being said, liquids, pastes and powders are the common materials. The extruder needs to fit the food/mix properties. Hummus is by far one of the easiest ones to have fun with, easy to tweak and play with! Thank you.


    Reply 3 years ago

    Hi sorry, for the test I was using Jell-o, just because its easy to source and print with. Hopefully more test prints to come soon. Thank you