Basic considerations about Delta printers
Posted by fma
|
Re: Basic considerations about Delta printers January 05, 2016 04:40AM |
No, the gonio is under the bed, not on the effector... About the spindle, I'm thinking using a little brushless motor, like used in quadcopters...
Anyway, these are possible options only, and I really want to build a Delta
For now, I need to determine the global size, so arms geometry is an important factor.
Frédéric
Anyway, these are possible options only, and I really want to build a Delta
For now, I need to determine the global size, so arms geometry is an important factor.
Frédéric
|
Re: Basic considerations about Delta printers January 05, 2016 05:21AM |
Quote
fma
No, the gonio is under the bed, not on the effector... About the spindle, I'm thinking using a little brushless motor, like used in quadcopters...
Anyway, these are possible options only, and I really want to build a Delta
For now, I need to determine the global size, so arms geometry is an important factor.
The Arms should not be less than 20 degrees to the bed at the farthest extent of there travel and should be as wide apart as you can get but aim for no less than 45mm (that is around 40 mm wide fixing points on the effector/carriages which is the std for the Kossel if you can go wider then you will have greater rigidity)
Doug
|
Re: Basic considerations about Delta printers January 05, 2016 05:28AM |
In my opinion, the only reason to prefer delta to coreXY is the high printing speed. Otherwise delta has only disadvantages. It needs more powerful/expensive controller, it is harder to calibrate, and it is taller. But if you want a delta for the fun/challenge of it then go on 
Make your design so that you have an option to avoid the long bowden. The bowden is a major hurdle for high quality of prints which need to retract very often (e.g. models which contain a lot of thin towers). If you are after really high quality of prints then you need get rid of bowden. The options are:
Here are some thoughts about the diagonal rod length: [forums.reprap.org]
Make your design so that you have an option to avoid the long bowden. The bowden is a major hurdle for high quality of prints which need to retract very often (e.g. models which contain a lot of thin towers). If you are after really high quality of prints then you need get rid of bowden. The options are:
- mount motor on the platform just next to hotend (you need stiffer structure and stronger steppers for this since the platform will be heavier)
- use a thin flexible hard steel cardan/shaft and a speed reduction at the platform - you can design your own or buy e.g. flex3drive
- use the flying extruder mod (search this forum if you do not know what it is)
Here are some thoughts about the diagonal rod length: [forums.reprap.org]
|
Re: Basic considerations about Delta printers January 05, 2016 06:56AM |
Quote
hercek
In my opinion, the only reason to prefer delta to coreXY is the high printing speed. Otherwise delta has only disadvantages. It needs more powerful/expensive controller, it is harder to calibrate, and it is taller. But if you want a delta for the fun/challenge of it then go on
Well, I find it simpler to design: only 1 axis, repeated 3 times. About controller, I'm waiting for a Replicape... Very nice to add new features (mostly written in Python).
Why is it harder to calibrate?
Quote
Make your design so that you have an option to avoid the long bowden. The bowden is a major hurdle for high quality of prints which need to retract very often (e.g. models which contain a lot of thin towers). If you are after really high quality of prints then you need get rid of bowden. The options are:
- mount motor on the platform just next to hotend (you need stiffer structure and stronger steppers for this since the platform will be heavier)
- use a thin flexible hard steel cardan/shaft and a speed reduction at the platform - you can design your own or buy e.g. flex3drive
- use the flying extruder mod (search this forum if you do not know what it is)
Yes, this is what I planned to do; I designed a light but strong extruder, using a geared nema 14 motor. On the other hand, for dual extrusion, it will be difficult to embedded 2 extruders :/
I printed for a long time with bowden, and for the parts I'm printing, I don't have any issue; I switched back to a standard config to be able to extrude Filaflex. But I recently tried Flex PLA, and find it more interesting for my needs. So, I switched back again to bowden a few days ago, and I'm making more tests to see if I can still get good results (have to try Flex PLA).
BTW, I will use Nema 23 motors, and I also need to find strong belt, to avoid elongation under tension. Any suggestion? Does GT2 exist in 10mm width?
Quote
Place the equidistant diagonal rod as far away from each other as possible. Use high quality angle ball joints.
What ball joints do you suggest? metal ones? Or plastic (IGUS?) ones?
Thanks! Feel free to send me on old threads if you have some in mind (but don't make specific search for me!).
Frédéric
|
Re: Basic considerations about Delta printers January 06, 2016 12:30PM |
Quote
dc42
Quote
o_lampe
Quote
hercek
Place the equidistant diagonal rod as far away from each other as possible.
I don't fully agree with that. Wider rod distance also leads to a longer lever for any play or wobble in the carrier trucks.
-Olaf
And I don't fully agree with that either. The longer lever at the carriage end is compensated by the longer lever at the effector end. A wobble of a given angle at the carriage will lead to exactly the same angular wobble at the effector, no matter what rod spacing you use. However, I agree that if you do have carriage wobble about a vertical axis when using wheeled carriages, using wider extrusion will reduce it.
Everybody has his own opinion. But did you notice fma wants eventually to place a spindle on the platform? That means there will be significant forces trying to tilt the carriage. If these forces are compensated by bigger lever then it means less loading on the ball joint and the rods/joints do not need to be that strong. Also if the rod end have a small bit of play then the play will be less significant if the diagonal rods are spaced more far away.
I say put them as far away as your platform allows. Your platform will be on the bigger side if you want to fit some spindle there. It may not be wise to increase platform size just to increase the diagonal rod distance. But to be sure you need to compute what the forces will be. My guess is that a Ø 7 mm plastic ball angle joint will handle about 40N. The distance between rods will tell you whether you can stay with plastic (rods far away) or you need to go for POS 5 (rods near to each other).
On the other end, it will not help him much if he plans to mount it at only 4 cm wide aluminium extrusion on simple rollers. Fma, just make sure your rollers have high quality bearings which do not have any noticeable play and ... heh ... maybe delrin will be stiff enough. I would go with either Ø16 mm smooth rod with good quality linear bearings (not a Chinese cheap crap which has play around 0.1 mm) or good linear guides (the specification for the linear guide will tell you what forces it can withstand and under what errors). I would go with thick linear rods (they are cheaper in my location). But I wish you will go with simple delrin rollers
Let us know how it behaves when you put a spindle on the platform and try some soft wood or a PCB!Edited 1 time(s). Last edit at 01/06/2016 12:31PM by hercek.
Sorry, only registered users may post in this forum.