RappiDelta Large format delta 3d printer-join the revolution
Join the Delta Revolution! A large format 3d printer based on the delta concept from Johann Rocholl
The 3d revolution has started.
It is official, 3d printers are taking over the world and recently a departure from the more traditional cartesian 3d printers has started. Johann Rocholl has taken the printing world by storm with his innovative delta printer.
What I have done is taken this to the next level, and by using the principles of true delta operation I have been able to scale the delta printer beyond the bounds of the limitations imposed by the tower design. Truly thinking outside the box this delta printer can also print outside the box!
The project was inspired by my 7 year old son who started asking me to make accessories for his growing toy collection. He originally asked me to make them from Polymorph but this became fiddly and time consuming and I suggested that he would be better creating them with a 3d printer and the RappiDelta was born.
I like to be different so the advent of delta type printers appealed to me but I couldn't work out why they used towers rather than a true delta design. I made it my mission to create something new and unique and as a byproduct it became apparent that my design opened up some exciting developments.
Some early testing of the concept was performed with eggs, yes eggs. There was no printing involved, just accurate position control. Here is the video:
With the concept proven it was time to develop the frame, this was a simple affair as the design lends itself to a simple gallows construction, this will remain in production models giving a minimalist look and feel.
This design has a 400mmx400mmx300mm build area, but this will be increased to 400mmx400mmx400mm in the production model. It is an all aluminium construction printer which will allow for fast and efficient production as I will not need to wait for plastic parts to print. So far testing has been done with a Yoda bust printed at around 40mm/sec and a layer thickness of 100 microns.
The kit comes complete with optical enstops as my testing has encountered some issues with the mechanical endstops that seem to be widely used. For a delta robot it is critical that you know the position of all 3 arms at the start of the print and using the mechanical endstops seemed to give me variable results due to bending of the microswitch arms or complete failure of the switch itself. The decision to move to optical endstops was a logical one as they offer both better accuracy and repeatability.
Below is a time lapse video of another Yoda bust, this time printed at 400 microns. You can also see the full prototype design in the video including the top mounted spool holder and triangulation supports to stiffen the frame.
To show the versatility of the RappiDelta I have printed Yoda at a truly spectacular 35cmx30cmx28cm and not even approached the limits of the build volume.
Because of the unique way in which the delta printer works it is also possible to use the same build volume to place an object for scanning - something not possible with a cartesian printer and something not practical with existing delta type printers due to the constraints imposed by the towers. The scan capabilities will be close to those of the print volume by flipping traditional scanning on its head and using the delta design to move around the object rather than rotating the object on a platform. Simple, elegant, and obvious.
Funding from this project will be used to develop the scanning method as well as improve the design of the existing RappiDelta design.
Below is a 3d scan of my head that was created from a series of 2d photos using 123d Catch from Autodesk. This shows the principle in which the basic scanner will work, but rather than taking the images yourself the idea is that you can just push a button and the scanner will do the work.
As a comparison here is a partial scan of my head using David Laser Scanner, the detail is a lot better and as a bonus it proportions my head a little better. There are some artifacts that remain on this scan, but it is my first attempt and I have no doubt that a few tweaks will yield amazing results.
The scanner will utilise high resolution camera or a high resolution camera and miniature projector depending on the technology used. The assembly will attach to the print head and as the RappiDelta moves around the object that is being scanned servo motors will ensure that the 'scanner' is always pointing at and focussed on the object. This will all be controlled by the same control board that handles the printing process.
A prototype has been created with a webcam and 123d Catch to show the how it will work. The quality is not representative of what is possible but shows the basic principle of operation.
The scanning is done using a gcode file in Pronterface to move around the object and keep the camera pointing at the subject. This is used by connecting the servo to the spare 'Fan' pin and controlling it with the M106 code. It will also be possible to use the M240 code to trigger the camera to take a photo. The following demonstration shows that the webcam is mounted on an arm attached to the servo to get a greater distance from the object and I intend to add a tilt servo to capture the object from a wider range of angles so that 123D Catch has more to work with. I have found that providing a calibration pattern helps 123d Catch no end, and working within an enclsosed space also yields better, more controlable results. As a prototype it is still a bit rough around the edges but I am hoping that funding from this kickstarter will allow me to develop the scanner further so I can offer a cheap upgrade for RappiDelta users.
The assembly will have a positive mounting on the end effector for stability and repeatability and will also be easy to mount and unmount for fast transition from scanner to printer and vice versa.
The scanner is not yet available as a production ready model but I will continue development with the aim of being able to provide a cheap scanning solution for RappiDelta users in the near future.
Parts, where possible, will be sourced and manufactured locally to help small businesses in these tough times. I have experience of having parts laser cut and machined for previous projects and in a previous life I designed a kit car chassis that was cut on a CNC Plasma Cutter (Robin Hood Exmo) There will be several phases of product development:
- The top Y frame has been made in my garage with simple hand tools from flat aluminium stock. This will be replaced by laser cut aluminium offering precise alignment of the arm assemblies and motor mounts.
- The current design uses off the shelf bearing blocks to support the arms. These will be replaced with custom made more aesthetically appealing.
- The worm and wheel are currently cheap plastic items and although they are functional I am looking at various options for improvement, particularly by integrating the arm with the wheel which will again make alignment more precise.
- The end effector is somewhat functional at the moment but again in the name of precision this will be replaced with a laser cut aluminium item.
- The bed is currently heated making the printer suitable for PLA and ABS prints but development will continue.
- Power comes currently from a PC power supply, but this will be replaced with a high power 12v power supply.
- The final stages of development will focus on integrating the scanner. This will attach to the end effector allowing full 3d scanning capability. I will be developing 2 types of scanner, one that can use a normal webcam and 123d Catch from Autodesk and one that uses Structures Light Scanning and the David Laser Scanner software.
The printer uses a number of open source components and I will continue to use the J-Head nozzle, an airtripper extruder, and an arduino based controller with a derivative of the Marlin firmware that has been modified to work with the true delta design. I have modified the airtripper extruder and replaced the nut and bowden tube for push fit 4mm adapter which is easier to load, similarly I have used a 4mm push fitting on the J-Head nozzle.
Once funded, I will endeavour to get the rewards processed as soon as possible and although I have allowed significant time for the major rewards I would hope to fulfil them in advance of the dates provided. I have a number of suppliers that I plan to use and all rewards will be handled and shipped by me personally after ensuring that all orders are complete. However, there is a chance that I will get more demand than I can reasonably fulfil in short timescales on my own and although I have ideas on how to prepare for this eventuality I can not make firm plans.....put me to the test, I am up for a challenge!
By supporting this project you will not only be joining the delta revolution you will be helping to create it!
It is simple really, for most startups raising the capital is a big problem and in many cases is the reason that the average entrepreneur isn't able to get their project to market. The kickstarter platform allows cool projects like this to become a reality as consumers provide the funds up front so that bulk discounts on materials and parts can be obtained at the outset to provide better value products, the funding goal is set to ensure these discounts can be leveraged.
Because it is impossible to quote international shipping I have put indicative costs only. Please be assured I will work with each individual baccker to ensure that the best and most cost effective shipping quote is provided. I will only charge the cost price plus a small handling fee.
I dont claim to have invented this printer from the ground up, although I have introduced a few twists of my own. A lot of credit must go to the open source community who provide elements such as the Marlin Firmware for the controller, the sangulino controller, the Printrun suite of interface software and of course Johann Rocholl for his pioneering work with Delta printers.
Risks and challenges
In real life I am a Project Manager in a software environment and so I am familiar with the concept of risk!
This project is no different to any other and there will be risks. I have mitigated a number of them with the design and development of a working printer, but the addition of a scanner represents a significant risk.
I am positive that I will be able to overcome the risks and be able to present a viable solution, the only risk that will remain will be timescales. I have tried to be realistic with my timescales but I am aware that this type of project may receive more backing than I can possibly fulfil alone which will in turn lead to delivery delays. I am quite prepared to put plans in place if my funding goal is acheived and will publish regular updates.
- (30 days)