Large Delta speculation
-
Thanks all. Is there any sort of "Golden Ratio" of height vs diameter that works best, or is it largely immaterial? In terms of accuracy, reliability and ease of calibration, is there any difference between "tall and slim" vs "short and fat" if you get my meaning?
-
There is no golden ratio. Given a limitation on total printer height, you can trade printing height off against printing diameter however you wish. Some time ago I converted my Mini Kossel from 160mm print diameter to 300mm print diameter, using the same vertical extrusions. The cost of the increased print diameter was that the printable height dropped to around 90mm.
-
OK. Thanks for all that. I'll mull things over for while. I'm undecided whether to "upgrade" the CoreXY to single colour by making an X carriage adaptor to take something like and E3D volcano, or build a separate dedicated single colour printer.
-
Height isn't really a problem you can make them as tall as you want, but there is obviously a limit to how tall parts can be before they detach. As I'm sure you find with your current machine.
Greater width reduces available z height as you need longer arms.
Is there an optimum width to height ratio? A ratio which gives the greatest printable area for any given volume of delta printer? Is it the proportions of something like (the very popular) Kossel? Why were they chosen?
Also as size increases how do you plan to tackle the extrusion volume issue? Diamond 5 way and a 1.2mm nozzle? I always find these "big box" printers a bit ridiculous they can print a chair but only if you have 7 days and 20 reels of filament and are willing to swap them when necessary (maybe less often with diamond 5 way). Or go to pellet extruder and 2+mm nozzle?
Some sort of dual nozzle system (one massive and one small) for bulk infill and perimeters respectively, might give the most flexibility.
-
…..................................
Also as size increases how do you plan to tackle the extrusion volume issue? Diamond 5 way and a 1.2mm nozzle? I always find these "big box" printers a bit ridiculous they can print a chair but only if you have 7 days and 20 reels of filament and are willing to swap them when necessary (maybe less often with diamond 5 way). Or go to pellet extruder and 2+mm nozzle?A 0.9mm nozzle is fine but 1.2 might be fun and you don't need a Diamond 5 colour (I wouldn't even consider putting one on a Delta). I reckon a volcano might do. I can get filament in 2.3kg reels which last a while - wouldn't have to get up in the night to change it and I can always pause and resume.
I hate sitting on plastic chairs so wouldn't consider printing one :). The coffee table I did wasn't too bad - 1mm layer width and 0.7mm height took about 19 hrs each half https://www.youtube.com/watch?v=C4oljUKE6KM&t=205s.
Actually, I'm working on a design for a printable version of this sort of thing https://www.erzgebirgepalace.com/Christmas-Pyramids/4-tier-Pyramids/4-Tier-Christmas-Pyramid-Forest-Design-135-cm-53-inch::6983.html?utm_campaign=shopping&utm_source=merchantcenter&utm_medium=googleshopping&utm_content=google-uk&gclid=Cj0KCQiAkZHTBRCBARIsAMbXLhGa6bDNieFp9y5GSWdT3mZ3Zy1FcxQn5EtkEceQYyVQMTUXx0KwLd4aAipHEALw_wcB. Powered by a motor of course - candle power is maybe not such a good idea when it's made out of plastic.
For sure I could make it in sections but it might be fun to print in one go. It's easier to print the columns when they are attached to something. The entire thing is mostly hollow so it won't actually consume much filament but it'll need a "normal" size nozzle so for sure, print time might be horrendous.It's all still just the germ of an idea and I may not go ahead with the printer, but there are ways and means….....
-
Deltas are fun, mechanically simple (but they need to be built precisely), and great for tall prints. Here https://tractus3d.com/our-industrial-3d-printers/ are some delta printers of various sizes. The larger ones are powered by Duets, I don't know about the smaller ones.
-
Not sure it helps, but my Delta based on DC42's design is 1.7+ meters tall and has a 380mm dia bed, with about 350mm reachable. The extra size helps to keep the clips I use for the glass well out the way.
Here's the frame of the printer during the build on a table specially made for the printer, the spool is mounted under the table top on a lazy Susanne, I really must take some new pics of the finished printer.
-
An impressive looking machine and I love the table and spool holder, I have a 1m x 340mm delta myself, but can I ask two questions:
- at that height without cross bracing is the frame not somewhat flexible?
- what does all the extra height with the only slight extra width (compared with mine or DC's delta) allow you to do? Surely once a part is 800+mm tall it has a very high chance of detaching from the build plate just due to the leverage any impact with a blob or suchlike would generate.
If this machine were significantly wider as well as taller then I could see the obvious benefit.
Makes me wonder if you could fix attachments to the frame to print on a build plate at the bottom, then insert another build plate slightly above the first one, and immediately begin printing on that, then another etc.. With fast calibration and a good probe (smart effector/piezo) you could print almost straightaway. Its only a gimmick but it would be quite cool.
IMO a lot of height is rarely ever used, my 1m tall delta could be 600mm tall and it would have only prevented me printing about 2-3 objects in the last 12 months.
-
An impressive looking machine and I love the table and spool holder, I have a 1m x 340mm delta myself, but can I ask two questions:
- at that height without cross bracing is the frame not somewhat flexible?
- what does all the extra height with the only slight extra width (compared with mine or DC's delta) allow you to do? Surely once a part is 800+mm tall it has a very high chance of detaching from the build plate just due to the leverage any impact with a blob or suchlike would generate.
If this machine were significantly wider as well as taller then I could see the obvious benefit.
Makes me wonder if you could fix attachments to the frame to print on a build plate at the bottom, then insert another build plate slightly above the first one, and immediately begin printing on that, then another etc.. With fast calibration and a good probe (smart effector/piezo) you could print almost straightaway. Its only a gimmick but it would be quite cool.
IMO a lot of height is rarely ever used, my 1m tall delta could be 600mm tall and it would have only prevented me printing about 2-3 objects in the last 12 months.
I'm using the metal 20x40 corners and extrusion and so its stayed quite stiff, there was very little extra cost in going higher which is mainly why I went with the height with the option that I could always make it shorter if needed later. I wanted to use linear rail but the suppliers I found no longer had it long enough. And one seller on Ali kept asking for more money after agreeing a price, luckily I got a refund from them.
I've made some test tubes over a meter tall and they did not detach and I like the idea of making tall vases some day with it.
Heres a pic of it finished and running.
-
Dekingman, you may find this helpful. It is a calculator for Delta Printers. Still somewhat in Beta; very accurate for the RobotDigg 2040 corners (the default).
I'm building a largeish delta (see photos) around a DuetWifI. 1.5 Meter tall overall, 600mm build plate. This URL presets the calculator to my final sizes.
-
Danal, get some toothed idlers. You'll thank me later.
-
Danal, get some toothed idlers. You'll thank me later.
Will do.
Also, I forgot to mention, there will be a ton of diagonal bracing. Haven't decided exactly what/where; still very clear that it is needed.
-
-
-
I am going to see if we can find an optimum width to height ratio. Please contribute your deltas specs here:
http://forums.reprap.org/read.php?178,806859,806943#msg-806943
or here if you prefer:
https://groups.google.com/forum/#!topic/deltabot/Gc2Ee9y3lrM
Lets see if there is an optimum.
Is there a way to do this by mathematics/simulation? If you know how please let me know (on one of the threads above).
-
I really don't think there is an optimum ratio, but there are some design considerations:
- Taller vertical extrusions require thicker extrusions or cross bracing to maintain rigidity;
- A wider print area/larger delta radius needs longer arms, which may need thicker carbon fibre tubes to maintain arm rigidity;
- Taller and wider both increase the Bowden tube length if you use an extruder drive mounted on the frame. So for larger delta builds especially, consider a lightweight direct drive extruder mounted on the effector, or a remote direct drive extruder, or a flying extruder.
-
I really don't think there is an optimum ratio, but there are some design considerations:
- Taller vertical extrusions require thicker extrusions or cross bracing to maintain rigidity;
- A wider print area/larger delta radius needs longer arms, which may need thicker carbon fibre tubes to maintain arm rigidity;
- Taller and wider both increase the Bowden tube length if you use an extruder drive mounted on the frame. So for larger delta builds especially, consider a lightweight direct drive extruder mounted on the effector, or a remote direct drive extruder, or a flying extruder.
For my 1.7m Delta the e3d titan extruder I had just couldn't push the filament without fairly constant problems, however swapping to the Bondtech and capricorn ptfe tube fixed a lot of issues with extruding and reducing the amount of retract that was needed, although my Bondtech has to run backwards as the bowden tube can only be gripped by the adapter rather than the built in fitting.
As for optimal dimensions I think that depends on what you want to make, if you want to 3d print lots of rockets, tall and and narrow is good, if you want to print lots of fossil skulls then wider and shorter is fine.
-
There is a calculable ratio between all the horizontal elements of a Delta/Kossel printer. Meaning the horizontal extrusions, resulting bed size, etc. "Ratio" may not be the correct word because there are some fixed parts (the offsets in the 'corners' or 'brackets') and some variable parts (how long the horizontal extrusions are cut). Clearly "calculable".
Vertical? There are some lower ratios to horizontal that don't make much sense. Going upward is limited only by practicality of building, bracing, fitting into a home or other building, etc. Upward is (almost) unconstrained. Maybe eventually belts get too long, or homing takes forever or similar.
The diagonal rods fit mostly in the "calculate from the desired horizontal" category. They can have some constraints from a too-short vertical… ignoring that for a moment (because there are many other reasons not to be too short, like being able to print tall things!), then, mostly, diagonal rods are constrained by not wanting to be below 20 degrees angle at the maximum travel of the extruder (constraining angle will always be at the edge directly opposite any tower). Again, "calculable", either 'forward' or 'reverse'.
Back to the original question: No perfect 'ratio' between H and V. Maybe some lower limit on vertical, given a horizontal as a starting point.
There also needs to be some thought toward availability of given parts. As an example, there can be HUGE cost differences to get some of the components in 1.7 meter length vs. 1.5 or 2.0. This applies to many things that interact with each other, like the bed vs. the horizontals. Assuming a glass/aluminum/heater bed, my research indicates the glass is probably about the same cost in any diameter (per sq in or cm), the aluminum has HUGE step changes at various sizes (assuming cast plate, not sheet), and the heater has fewer step changes while still becoming quite expensive for anything beyond "standard hight volume sizes".
If we combine step changes in parts sizes, horizontal interactions that are calculable (and need the fixed offsets to precisely calculate a given build), verticals that are calculable, and a somewhat weak interaction between H and V... those are the thoughts that led me to create the calculator linked above.
Here is a repeat of the link:
-
Some thoughts on google groups about this:
https://groups.google.com/forum/#!topic/deltabot/Gc2Ee9y3lrM -
Here is a example of my printers. They have a print height of about 1200mm. I have no issue what do ever printing tubes that highe, and powered by Duet's.
Im 6'3 tall
