Dual extrusion delta

Hi,

Last year I built an dual extrusion delta, sort of a dual kossel.
I used a larger effector, a rotating head so one hotend was close to the bed and the other was lifted.
The mechanics of that worked well, but I had many other issues such as:
-I felt like getting offset Z correct for both nozzles was too complicated
-I used a heated enclosure and even with corners printed in nylon these started to deform when the enclosure reached +50 degrees celcius.
-to avoid "drip" I had to cool down the hotend that wasn't in use, this made printing very slow.
-the amount of waste that came with every dual color print was signifcant.
-The printer was very large, and I don't actually have that much space,
because of the size and shape of printer the enclosure also needed to be large which resulted in a high power consumption,
-Because of the size the thing wasn't very rigid or easy to work on.
-I had some issues with rust

In the end the work in keeping it running wasn't worth it, I think it would've worked well if I could have avoided the enclosure, but for larger abs and nylon prints this is a bit of a must.
I also 3D print outside to avoid the fumes and the smell, which also requires an enclosure.

I'm now working on converting my first 3D printer (a simple design using igus twe bearings, ideal for printing outside where there is condensation etc.) to use 2 nozzles that are lifted 5cm in the air when not in use (again use compressed air as that worked very well), this will also allow setting the Z height manually with a setscrew (so head 1 & 2 have the exact same height from the bed)
That printer has served me very well for 4 years, I don't think the dual kossel would have ever gotten to that point.

I've put all footage I found online so others can avoid the mistakes I made.
[youtu.be]

Edited 1 time(s). Last edit at 07/18/2021 10:07AM by brampie.

Quote
brampie
.........................................
-I felt like getting offset Z correct for both nozzles was too complicated
....................................

I can only help with one of your items. See [www.youtube.com]

I like your idea of using a pneumatic lift to lift the inactive head above the active one and may even do some quick experiments to see if I can bring anything to it.

Mike

Hi,

Nice printer, is it still a work in progress >

My current idea is to use two 3mm rods for every "slider", both cylinders could be mounted in the center for easier access.
The pneumatics worked very well on the printer, very fast and easy, adjust strength with air pressure, requires just one output pin, doesn't consume power to keep it's position and can take the heat reasonably well.
With this new design I was thinking:
-I could wipe the head after every layer or so by moving the inactive head up, move over some kind of wiper, back down to wipe and back up.
Not sure about how well that would work in preventing drip would it would certainly be possible and not slow down printing that much.

-For leveling I could lift both heads and mount a microswitch to the carriage, no need for babystepping, I can adjust the Z height of each head with a screw or so.

If you want I can send you the openscad files, the slider would be printed in igus material or nylon to minimize wear,
as the cylinder pushes down and the adjustment screw functions as an endstop this should be precise enough I think.

thx
Bram

Hi brampie, It is sort of a work in progress. Mostly though, I use it to explore better ways of doing things: Can it be done cheaper, more reliable, simpler, or more accurate. Mostly though, I try to explore the neglected technologies in FFF printers and whether commonly held assumptions are true. For example, can a wirewound resistor be a better choice for a hotend heater than a heater cartridge. It turns out that it can, but only by using the right resistor and comparing it with the popular cheap cartridges made for DIY 3D printers.

At the moment, that printer is being used to try to get to grips with the causes of inaccuracy in underbed nozzle contact sensors and what can be done to make them accurate and reliable.

Mike