Δ3D Build-log

Δ3D and Why

Hello people!
Meet the Δ3D (as it looks at this point)



Now that I (hopefully) got your attention, all the boring stuff
I love to design things and tinker, I always have! The last couple of years I have been using 3D printers quite a lot and gathered quite some experience in the field. My first experience with 3D printers was a MakerBot Replicator 2. Later I got myself a Prusa i3. I learned a lot from this machine and it is still producing really nayz prints for the cost of the machine. My latest 3D printer experience though is the most.. well.. interesting one. An AirWolf 3D AXIOM printer.. Luckily I didn’t buy it, or advise any to do so, I just happened to get stuck with it. However, over the last two years this machine have taught me so much about 3D printing.. at least how it should NOT be done!

I have been tempted to build a 3D printer for a long time.. just for the fun, learning and experience of it. A few months ago, I really started designing my own printer and now it is actually getting somewhere. So I decided to create this blog before I got to far for it to make sense! I create this blog mainly because I love these forums, I love to get inspiration from other people’s builds – Thus I want to share my build in case others find it inspiring. The project is a spare time project, I will be posting news when there are any, but it may take a while before the printer is finished.. Meanwhile, enjoy this build-log of my homemade delta style 3D printer. For now, I call it Δ3D. Very original, I know, but I am really bad with finding names..

LETS GET ON WITH IT!

Edited 5 time(s). Last edit at 08/06/2017 10:49AM by DannerD3H.

1. My design

I have chosen the Delta platform and I’m not going to lie about it.. I choose it because I think they look F****** amazing! Especially when printing!!
I have been reading a lot about Delta printers and DIY projects of such. I have gathered inspiration from many printer designs out there, but among others are the Kossel, Hexagon V2, Hexagon XL, Hexagon by Protoprinter and the ATOM printer to make a few honorable mentioning’s.

Before I begun I put down a few design requirements:
1. Silence! – The singing from a MakerBot stepper motors or PSU fan noise of the AXIOM is killing me!
2. The visual is more important than economy(within reason), after all I will only be building one.
3. It must be easy to modify – The first design is undoubtable going to be modified..
4. Quality over speed – The goal is to compete with Ultimaker print quality.. One gotta aim high!
5. Build area must be bigger than my Prusa I do not care much about build height, never used the Prusa’s 200 mm of height anyway but often needed one dimension above 200 mm.
6. Build area must be easily enclosed without affecting build area or hurt the design.
7. Dual extrusion must be possible in full build area also if build chamber is enclosed. Dual extrusion is only intended for support material, I won’t do much if any multicolor printing.
8. The printer must be as small as possible without compromising any of the above – or in other words, our apartment is not huge so it must be Girlfriend friendly

I have access to two working 3D printers, a workshop with a manual lathe and milling machine. At occasions, I also have access to a CNC mill. The design will mostly relay on 3D printing and manual work, but a few critical parts will be in aluminum from the CNC.

Some of my general design choices explained:
1. Hexagon shaped printer – Smaller footprint than triangle when considering enclosure.
2. 20x20 extrusion – I don’t want a super tall machine so stiffness should be okay as I have 5 towers in the worst case I will brace the frame with diagonal carbon rods.
3. Ø310 mm print bed – Bigger than the Prusa, availability and for now I use the heatbed of the Rostock Max V3, but the design will be prepared for a Mains silicone heater as an iteration.
4. 0.9 degree stepper motors – Higher accuracy over speed
5. 24V power supply – Fast heating times, Faster and stronger stepper motors and lower currents in general.
6. Smoothieboard V2 (will hopefully be released soon) – Just sounds like a great 32 bit controller! And suuper high microstepping 1/256. This will greatly reduce stepper motor noise! The controller also easily allows for added features like laser engraving.
7. Safety – No exposed WIRES (especially not mains). I am clumsy and already replaced my Prusas Ramps controller twice because of my clumsiness and stupidity..
8. Single nozzle dual extrusion Prusa/ATOM style! – No oozing issues, no Z calibration of two nozzles. Downside is more wasted material though.

My biggest remaining dilemma is the extruder.. I want to make the Bowden tube as short as possible. I just have not figured out if I should use a stationary extruder as close to the hotend as possible, which I have shown in my first post, or if I should use a flying extruder. If I go with the stationary extruder, I could mount both in the chamber. If I choose a flying extruder I would only have the main extruder flying. The support extruder I would keep remote with a longer Bowden tube.

The design will be explained in more detail as the build gets along!

Quote
DannerD3H
6. Smoothieboard V2 (will hopefully be released soon) – Just sounds like a great 32 bit controller! And suuper high microstepping 1/256. This will greatly reduce stepper motor noise!

I'm biased, but why not a Duet WiFi, which uses the same stepper motors that Smoothieboard 2 is supposed to be going to use, but has already been available for a year? Plus you get a better web interface and least squares delta auto calibration. Or the Duet Ethernet if you prefer a wired connection.

I can't fault most of your other design choices. The 5-verticals approach allows the possibility of adding 2 more carriages to mount the extruders on, so that they can move with the effector and allow shorter Bowden tubes. This is a configuration that we're planning to support soon in RepRapFirmware.

Edited 1 time(s). Last edit at 08/06/2017 03:56PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I was also biased on the smoothieboard
BUT you actually have a very good point. It is risky to use a brand new product compared to one that has been out there for a year. I will have a closer look at that controller that is for sure!

Oh that is interesting! Dind't know something like that was out there! will look out for that!

Quote
DannerD3H
I was also biased on the smoothieboard
BUT you actually have a very good point. It is risky to use a brand new product compared to one that has been out there for a year. I will have a closer look at that controller that is for sure!

Oh that is interesting! Dind't know something like that was out there! will look out for that!

More on the series 2 Duets can be found at [www.duet3d.com]. There is a support forum at [www.duet3d.com].

Don't get me wrong, I think Smoothieboard 1 was a good board for its time (I considered distributing it at one point) and I expect Smoothieboard 2 will be good when it eventually arrives, assuming that doesn't take so long that technology overtakes it first. Although for 3D printer control, I think Smoothieware has some catching up to do to equal RepRapFirmware.

Edited 2 time(s). Last edit at 08/07/2017 05:29PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

If your trying to build a top end 3dprinter don't skimp on the electronics. I have the Duet Wifi and could not be happier. I can't imaging building another printer without purchasing a DuetWfif. If your trying to save a bit of money, you can skip out on the Panel Due as the web interface that comes with the DuetWifi works seamlessly.

good design but for the frame you should realize by yourself

Hmm.. I really liked the documentation on the Smoothieboard V1, but i also didn't like that it wasn't reprap firmware, as that is what i am used to. I am still investigating further, i will still take some time before i have to choose
#Qdeathstar; Do you know if i can use the MKS tft32 display with the Duet? I would really like to use a touch screen with a pretty interface. I was also considering the Viki 2 LCD, but would prefer a touch screen.

#Titeuf007; what should i realize?

2. 3D print-ception

One of the best things about 3D printers are that they can produce parts for them self or other printers!


I have been printing parts for my delta printer for a while now. I have actually almost printed the first iteration of all parts., Quite a pile of plastic.. roughly 2 kg of PETG (including brims, rafts and so on).


As you can see, my design is not really suited for replication or certainly not mass production as the bare frame alone contains 30 3D printed parts. For the entire printer that number is closer to 70. Maybe not such a great design with all these printed parts, but I am still confident that the frame will be stiff enough due to the design and the three 4 mm aluminum plates I will have laser cut… somewhere.. will eliminate all flex from plastic corners.
I have also generally chosen to rather print more, smaller parts than fewer bigger parts. So compared to a Kossel which have 6 corner parts in total I have 24. I could have made 12 (hexagon) big ones instead of course, but by making 24 I have less risk of failing prints or at least shorter prints and more design/print flexibility. The frame alone (only structural parts) are roughly 120 hours of printing

Every structural part have been printed to optimize stiffness and strength with the print settings:
Material: PETG from DasFilament (some – the matte ones – are XT from ColorFabb though)
Layer height: 0.2 mm
Perimeters: 4
Top and bottom thickness: 1.6 mm
Infill: 35 %
Speed: 35 mm/s

All corner pieces are printed with brim and additional “Plates” to completely eliminated warping and ensure flatness. Before and after post processing picture of a top corner:


So far, I am missing four parts before I can assemble the frame! .. Yeah my priority of prints may not have been the best.. Should have finished the frame parts first.. but I didn’t have the 20X20 extrusions in the beginning, and wanted to ensure a good fit. They also have to be cut btw...

Print! Print! Print!..

Edited 1 time(s). Last edit at 08/08/2017 01:30PM by DannerD3H.

great
so impatient to see the result

Quote
DannerD3H
Hmm.. I really liked the documentation on the Smoothieboard V1, but i also didn't like that it wasn't reprap firmware, as that is what i am used to. I am still investigating further, i will still take some time before i have to choose
#Qdeathstar; Do you know if i can use the MKS tft32 display with the Duet? I would really like to use a touch screen with a pretty interface. I was also considering the Viki 2 LCD, but would prefer a touch screen.

Documentation for the Duet is at [duet3d.com]. The Duet doesn't support the MKS display, instead it supports a display controller that can be used with a 4.3, 5 or 7 inch touch screen. You even get an on-screen keyboard so that you can enter gcode commands directly if you want. See [miscsolutions.wordpress.com].

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

#dc42
I have been reading a little about the Duet now, and i definitely also like that board and documentation. I still didn't decide yet though. My first impresion is that the Duet has more features/flexibility, but it also feel slightly more complex to use. But that is a first impression, I am still way more into the Smoothie way of doing it (not saying it is better).
Do you know if you can limit the PWM to the bed? I do plan on using a mains bed and a SSR eventually, but for a start i have the 12V PCB heater from the Rostock max V3 lying around, but using that one with 24V power requires to limit the bed PWM signal. And i am going to use 24V power.
Also, is it possible to modify the Paneldue's GUI? I like the features of it but certainly not the look of it..

#titeuf007
Not good.. it may take a while xD

Edited 1 time(s). Last edit at 08/09/2017 08:27AM by DannerD3H.

I'm a smoothie convert. Once you get used to specifying everything using the G-Code commands I'd say it is just as easy but yes there is some learning to switch to the Duet terminology and commands. I've found that once you are onboard with Duet most things are much easier to work out. David does a pretty good job of taking the concept and making it so you do not have to have a PhD to figure it out (delta leveling for instance). I found that while the write-ups in Smoothie were nice I never did get Delta calibrations and Z-Height adjustments work as well as I have them on the Duet.

Smoothie is a good setup also, I just found that my thinking seems to align better with RepRapFirmware's way of thinking about stuff. My end results have been better with less frustration since switching to RepRapFirmware.

Quote
DannerD3H
#dc42
I have been reading a little about the Duet now, and i definitely also like that board and documentation. I still didn't decide yet though. My first impresion is that the Duet has more features/flexibility, but it also feel slightly more complex to use. But that is a first impression, I am still way more into the Smoothie way of doing it (not saying it is better).

There is a learning curve involved in switching to any 3D printer firmware that you are not already familiar with. The Smoothie way of configuring everything in in a text file is good. RRF does the same, but the commands are all gcodes. One advantage of this is that you can make any configuration change you like in the fly (except change the network MAC address) to try it out, without having to reboot the board.

Quote
DannerD3H
Do you know if you can limit the PWM to the bed? I do plan on using a mains bed and a SSR eventually, but for a start i have the 12V PCB heater from the Rostock max V3 lying around, but using that one with 24V power requires to limit the bed PWM signal. And i am going to use 24V power.

Yes you can limit the PWM. The bed heater output of the Duet is rated at 18A RMS so take care not to exceed this if you use a 12V heater with a 24V supply.

Quote
DannerD3H
Also, is it possible to modify the Paneldue's GUI? I like the features of it but certainly not the look of it..

You can choose between light and dark colour schemes as standard. The firmware is open source, so if you are into C++ programming you can make whatever changes you like to it.

HTH David

Edited 2 time(s). Last edit at 08/09/2017 07:06PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Just a quick update..

I am really busy at the moment, but i have printed 3 of the remaining 4 parts and i have received the last aluminum extrusions. So i will soon cut the aluminum and assemble the frame!


Thanks for the inputs DC42 and WZ9V.

3. Aluminum extrusions

So… It has been a while already, sorry about that! It has been a busy month for me!

I finally cut all the pieces of 20 x 20 extrusions I need for my printer. A total of 43 pieces… luckily, most of the same length.

I know the only way to build a printer capable of making quality prints is with a quality build frame. I also know the limitation of my design will most likely be the high number of 3D printed parts. However, the fewer sources of errors always make the result better! Therefore, I tried to get the length of the extrusions as precise and, most importantly, similar as possible.
First I cut all pieces 5 mm to long on a band saw.


After that, I used a milling machine to cut all the pieces to length. I also did this to ensure that the pieces had a very angular end cut. I knew that the saw I used was cutting a bit off angle.


I first took roughly 2 mm of one end of each piece, and then I used a fixed end stop to ensure that all pieces of the same length would be very similar.


I managed to get all pieces within +- 0.1 mm, which I am happy with(33 of the pieces should be 180 mm).


After all the milling and cutting, I also removed all sharp edges and cleaned the pieces… I will spare you a picture of that – it was ridiculously boring…

I chose the 20 x 20 based on some reading around on other people’s designs. In addition, I thought since my printer will not be as tall as other delta printers the 20 x 20 will be fine for the towers, I also have five instead of three. Now I am not so sure anymore though. I am truly surprised how flexible an 800 mm 20 x 20 actually is… It will be a really big design change to use other extrusions for the towers so I will see how the frame feels when assembled before considering that. The linear rails will also stiffen three of the towers so maybe it will be good enough.

During the last half month, I have managed to print all the remaining parts for the frame (and almost all the parts, which has been finalized in the design). I also received the 201 t-nuts(surprisingly many :O ) used to assemble the frame, so my next post(hopefully sooner than 18 days) will be on assembling the frame!

Looking forward to see the printer taking shape!

4. Frame build

Hmm I am not going to say anything about the long waiting time.. Spare time project, remember? and I may have used the time on building rather than updating you guys.. sorry! I am going to a Maker Fare here in Denmark the 14-15 in this month and I am planning to bring my Delta printer. My goal is to have a printer that can move (dry printing) at the fare. So I need to get going!

I have finally started the assembly, I will not have the aluminium plates laser cut before the fare so that will have to come later! So, this is the parts used for the frame:


Building the frame I used a flat stone block we have at work to ensure that the hexagons are assembled flat. In additions to the printed parts I use small internal steel angle brackets where possible to stiffen up the frame even more.



The final frame unfortunately didn’t turn out as stiff as I hoped.. The limitation is really the vertical 20x20 extrusions. The top and bottom compartments are VERY stiff. I hope that the long rails I plan to use will stiffen it further and that it will be stiff enough.. if not I will have to reprint corners to allow for bigger verticals.. let’s hope not xD I am planning to add some cross bracing on the back because of looks anyway 

PS if any of you are Danes, then come find me in Aarhus 14-15 th of October!

5. Drive system

I have chosen to use fully supported rails as I consider this the most high-end solution and the most correct type of linear guide for a 3D printer. That said I of course first tried to buy a cheap copy of the HIWIN MGN12 rails. However, they turned out to be too cheap.. I don’t know if there are other better ones, but these were horrible and looked and felt like they had been used for years.. RETURN TO SENDER! I then bought three original rails from HIWIN. Muuuuuch better! Smooth motion. I chose to re-grease them with some grease instead of oil as this greatly reduced the noise they produce. I hope this will not affect the lifetime of the rails!


I have also chosen quality belts from GATES. I really couldn’t find a uniform answer to which type of belt s better, steel or glass fiber reinforced.. So I have chosen glass fiber based on my own intuition.. Glass fiber is stiff, and it is much more flexible and I guess the loads will be limited in a 3D printer.
The motors are of course 0.9 degree stepper motors – the biggest I could find with a flat shaft (really don’t want a round shaft and worry that the pulleys may slip – even though they probably won’t if used correctly). I am planning to use 24 V to reach high speeds without loosing too much torque. For now the motors are mounted directly on a 3D printed motor bracket, I am considering to have these machined from aluminium to avoid them softening from heat. For now the motors are just mounted with a cork insulator – also to get rid of some horrible stepper noise


My carriage have fixed belt mounts. I chose this to make them as light as possible and simple for machining. I also found it easier to make a proper tensioner in the frame. Some may say that the top compartment is wasted space, but I would like to keep the spool there, at least as long as only one filament is used. My tensioners are simply pulling the idler pulley upwards. The idlers are mounted on a 5 mm shaft carried by two ball bearings.


The carriage are machined in aluminium. This is done for stiffness, and precision. The balls are mounted straight since my magnet joints allow for 180 degree of free movement. A picture says it all

 

beautifulll

I like what Im seeing but I did not read anything about delta arms so here is my proposal. You should use this [www.duet3d.com] you already have mounting points for carriage just order custom arms and a must magnetic balls

Thx. Titeuf007

Sorry Viod, but i am already way beyond that And i have 60 mm spacing on my arms so i sadly can not use your Smart PCB..

I have glued some arms already and manufactured some parts of my endeffector, so now i at least wnat to try my own design before trying something else

6. Arms and effector prototype

I have chosen to use magnet joints for my printer. I consider it the best solution since they by nature are backlash free and you can easily swap out the end effector. Additionally the serve as a safety feature, as the effector will just fall of one most types of collisions.
I bought a magnet and ball set, which I am greatly disappointed by.. One thing I am going to update very soon is the arms. Instead of using the magnet as a “bearing” I will use cheaper and stronger normal rod magnets and machine some endcaps like the Haydn Huntley’s Magball delta arms. More on that in the future!
I glued the arms in a jig to ensure that the arms had the same lengths, this process turned out to be more of a hassle than I thought it would have been.. I am pretty sure the arms are going to work just fine, but they sure isn’t pretty when you get up close..


My end effector is a story of it’s own. I have been reading and reading.. It is much easier to make an end effector with everything hanging under the arm attachment point like the Smart PCB one. I am sure this is also good enough! But, I still chose to try to create a balanced end effector with the COG at the same height as the ball joints rotation point. I think I managed to make a pretty and functional solution. This is what the design looks like:


I use two 35 mm blower fans for part cooling and one 35 mm fan (possibly a 40 mm Noctua) for hotend cooling. I use the BlTouch probe (This choice I made as I want to be able to probe ANY surface at ANY temperature equally well), an E3D V6 of course (I love the silicone socks, not for cleanliness but because they protect the heatblock from air from the cooling fans!). I plan to integrate single nozzle dual extrusion like the new Prusa and ATOM printers. The design really aims not to have anything sticking out of the sides, as this will minimize build space.
So far this is what my prototype looks like, this is what I can do for the fare next weekend. The only feature installed in the effector is the LED’s


And everything mounted on the printer



 

7. Cooling and Temporary electronics

I have already destroyed the electronics on my old prusa twice by being stupid and I had some stepper drivers overheat. I also have never heard such a noisy printer as the AXIOM at my work. This printers power supply sounds like a hairdryer! Therefore, I decided to do what I could to have effective quiet cooling of my electronics. The bottom compartment will have two 60x25 mm Noctua fans in either side. One side will push and the other will pull to force an airflow through the compartment. I received the fans and mounted them using the supplied vibration dampers from Noctua.


And I must say.. I just LOVE these fans! They move a lot of air and they have such a comfortable sound – Only a subtle wind noise! And they look super badass!
I would like to have the black panel’s laser cut as well but for now, I printed some dummies to protect my fans.
So, for the fare this weekend I really wanted the printer to be able to move. I have no hotend yet so I have no heaters to power. So I use a 12V 30W LED supply that I had lying around and a spare Ramps board I had for my ooold Prusa. This should be plenty to have the printer dry-printing at the fare! To keep everything slightly in order I have made a removable electronics-bay that I can pull from the front (I found another place for the display).




 

8. First movement test

So I finally got to see the printer move! And it looks super cool! I really love Delta robots!
But.. I am not happy.. Everything seems to work, except, that the printer sounds like a rattlesnake.. Listen for yourself.

[youtu.be]

It turns out the sound comes from my rails… because they do have a little play! This is really annoying.. I know now, that I should have bought rails with a slight preload, which these does not have. Well it will be fine for the fare in the weekend, but it will not be useful for printing! In addition, I really tried to make the printer Quiet, or at least not as annoying as my old Prusa and the AXIOM at work, but this is just as bad as both of those!
Does anyone here have an idea how to solve this without buying new rails?
Can the balls be replaced to fix this?

Edited 1 time(s). Last edit at 10/09/2017 04:33AM by DannerD3H.

I think you can adjust the preload yourself using the 2 little screws in the ends of the carriages but not definitely certain but worth a look?

Doug

I belive the two screws in the end are just to mount the wiper seal.. I don't belive the preload can be adjusted on the MGN12 blocks.. But i would love if it could!

So the fare was.. fun but most of the people had never seen a Delta printer before if even a normal 3D printer xD.


However, i did manage to have the aluminium plates lasercut anyway and barely got to mount them!. If one looks closely there is some tolerance issues.. it seems like my hexagons are slightly bigger than expected.. Anyway, I think it looks supper cool and especially the bottom plate which is screwed directly into the towers with a M6 bolt helped on the stifness! I think i might be good with the 20x20's anyway!


I cannot decide if i should keep the plates like they are, try to polish them or run them through a glass blaster. I don't want to paint them as i like the metal look.


I have also changed the design/location of the display as i have now chosen to use the duetwifi and the paneldue!
The effector design is also going to change as DjDemonD has convinced me to try the Nimble extruder!
More on those later!

Sadly i am out of fun-money for this month so i am afraid the building is gonna be on a pause until next month


PS Am i the only one where some of my images won't load? i have no idea what is wrong, as most pictures does load.. sorry about that!

Edited 2 time(s). Last edit at 10/18/2017 05:22AM by DannerD3H.

I can't see any of the images in your latest post, nor the ones in your previous 2 posts that have images - even if I click on "Open image URL".

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Okay.. i tried to share images from a google album, weirdly it works for the first few posts, but now i figured out how to do it properly. Sadly i can no longer edit the previous posts..