A different multi-extrusion approach - UM tool / printhead changer

[macua85 52]

Wow, simplify3d isn't cheap! Did you guys all buy a copy of this?

[Labern 774]

I find XT only brittle when its overheated.

[jonnybischof 60]

You need to make very thick walls with XT (actually, I always make 2 shells & 100% infill) to get "unbreakable" parts. Also, you need to find the temperature sweet-spot. I found that I have to print XT as hot as possible without getting too much stringing / melting. It needs to really melt and bond with the previous layers.

[ultiarjan 1,223]

I find XT only brittle when its overheated.

[/qruote]

I used 250 at 35 speed. .. too hot you think? I think i need to do some testing...

[ultiarjan 1,223]

Wow, simplify3d isn't cheap! Did you guys all buy a copy of this?

 

I did..

[ultiarjan 1,223]

A quick photo of my (now tested and working!) setup with one extruder for now. I'm printing the height adjustable unit tonight which will sit on the back right of the machine.

 

I like the way you integrated the fan in the pickup head. Curious to see your results.

[neotko 1,417]

You rotated the hangar so the clamp astops it from pushing it out. I tryed that. Try to check if the parked head stays in place when printing something that makes the printer 'shake'. On my test the force of the bowden and the angle of the hangar always made it end pushed out.

The head size looks very nice btw

[macua85 52]

Ultiarjan, the fan seems to work well. The hangar was printed with the new headsystem and came out really nicely. I'll print a robot tonight to show you guys. I also think the e3d prints much better than the OEM ultimaker nozzles, not sure why, maybe to do with the funky shoulder it has.

Neotko, i'll keep an eye out for that. The head is a pretty snug fit in the holder so hoping it wont be a problem. If it does seem to be, i'll put a slight angle on the holder (small enough that the magnets can still engage ok), to hold it in. The head for the other side will be slightly large as it has the height adjustment mechanism in it, an additional 6mm or so of depth.

Edited July 14, 2015 by Guest

[macua85 52]

This is the adjustable unit. An allen wrench can be put through the hole in the top, and the bolt (shown installed) turned to make the whole unit move up and down in relation to the fixed portion (that houses the magnets.

The fixed part houses a nut in the slot (tight fit)

Slide up from the bottom to assemble. Although I've not printed and tested it yet, I can't see why it wouldn't work as long as its all a snug fit.

Edited July 14, 2015 by Guest

[jonnybischof 60]

Theoretically you could calibrate the heights in software instead of making an adjustable printhead. You can adjust the Z-platform when changing printheads. The other printheads are parked, so they don't get in the way.

The question is, how to do it with Marlin...

[neotko 1,417]

Your design looks nice, I like the way you moved the annoying parts to the sides so you can use the most of the print area.

The problem I think you might find with a tight fit it's the balance of the magnets strength. I mean..

If it's too tight the change movement will need to be slower. You can always simulate a ramp so the head change it's fast and also goes slow when it's near the head change (to try to speed it). The problem I found doing this on my second version was that if the head doesn't just 'flow' to the clamp, there's a friction force that affects the x/y position of the head, that after a few head changes will make the printer to fail to repeat the coordinates, so you will end having a print like a Tower of Pisa. To solve this on the second version, and specially on the third and final version, I had to balance the clamp strength with the 'smooth-ness' of the movement. On my current version (I need to get some time and update my pdf) I changed the x/y slightly so the parked head it's the one that slides to meet the head changer. By doing this I removed any friction on the moving part and I can leave the printer for hours without any x/y fail. That's why I think that even changing the angle of the hangar it might flow out by a series of fast x/y movements. Another advantage by having the hangar like me or ultiarjan it's that the x metal bar does the job of 'pushing' in the parked head before it it's retrieved. By changing the angle you will have to use a metal 'thingy' like the one Foehsturm shows on this page post

In my experience if the head 'hits' the hangar, 1/5 times the x/y will loose alignment. So it's really important to have a smooth slide and to have the parked head on the same position when doing the pickup.

[macua85 52]

Thanks for the feedback guys. I'm trying the test print again now with the gcode supplied by tinkergnome. I had already written a command to slow the head down for the drop off and the pick-up as I thought that could be a problem. This test is 50 layers, will report back in 20 mins with the results!

[foehnsturm 969]

Hey guys.

Thank you all for your great work! So this new forum thing has at least one interesting hacking thread alive ...

(Don't want to offend anyone ;-)

[pm_dude 27]

After much wait, I'm about to finally jump in with a new UM2 mostly for that.

The question is: Where do I get the list of things I need to get started in this 44 pages thread?

[ultiarjan 1,223]

After much wait, I'm about to finally jump in with a new UM2 mostly for that.

The question is: Where do I get the list of things I need to get started in this 44 pages thread?

 

You first have to decide what hot-end you want as a second.. will you go for a UM2-like setup, or and E3D or something totally different.... direct drive ??

you basically need (bowden option)

- print a new feeder, get a feeder motor + knurled wheel+bowden

- get all the UM2 parts for a hotend, fanduct, fans, cables etc (all part nr's on github) or go for the alternative choice.

My main project summary is on page 27

[macua85 52]

So the accuracy test turned out very well, with the exception of mad stringing due to my poor retraction settings and generally poor gcode. It's become clear that this is a big part of the project, not just the mechanical aspects.

And cleaned up:

So the xy layer accuracy looks good. I need to do a couple of tweaks to adjustable design, then I'll crack on with printing and installing that. Looks like a purchase of simplify3d is required here too.

And for ultiarjan, here's a robot I just printed with this setup with the integral fan. 0.1 layers, 65mm/s, no infill:

Sorry about orientation of photos, they are fine on phone but upload wrong.

[ultiarjan 1,223]

@Macua85 Congratulations !! That looks very good !!

That robot looks really good, i would have thought 2 fans would be required for that kind of quality ... (thats what you get if you started with an UM2...) I'd love to see a stl/step of your fanmount just to have a better look at it.

[ultiarjan 1,223]

and here's the setup I'm printing right now...

[macua85 52]

Hey Ultiarjan, i've got around to uploading the stl files to youmagine so that you can have a look at the main module design. Not uploaded the adjustable e3d holder yet as I want to test it first. Is the gantry geometry the same on the UM2 as the UMO? If so this design may work on yours too if you want to try it out.

Link: https://www.youmagine.com/designs/e3d-v6-modular-print-system-magnetic-final

Edited July 15, 2015 by Guest

[ultiarjan 1,223]

UM2 & UMO are different, the distance between XY rods is the same but the XY are switched in height, so the design will be rotated 90 degrees on UM2

[macua85 52]

Ah yes, that is quite a difference! Looks like this design is for UMO only at the moment. If there is any interest, I can work at swapping the axis around on the design for it to suit UM2, but I have no way of testing it for other bugs.

[macua85 52]

How did the parts turn out by the way for the above design?

[ultiarjan 1,223]

Parts printed ok now, made the wall that clamps around the heatsink ribbons thinner (like the original E3D) so they can flex more.

For now I stay with the double fan approach, like to be able to swap easily between the double UM2 heads vs UM2 + E3D head. Actually made a breakout connector (just a screw terminal) to change a complete head+cable without the need to get to the electronics board. Did not test it jet so I hope a screw terminal between the heater & sensor cables will not create any negative side effects...

[macua85 52]

Yea I did the same too, all cables on mine are plug and play. Used RC xt batter connectors for the heaters, and smaller ones for fans etc.

[macua85 52]

Printed the adjustable head last night, assemble it roughly and the mechanism is going to work nicely, all clearances good. The only cock up I made was to leave 'extensive stitching' option on, so all of my holes have been printed over! Nightmare, but i'll just drill them out instead tonight.

I do have a query though. With my setup the front portion of the machine lies in unprintable territory. How do I account for this in Cura so that my models sit in the right relative place when I prepare prints? I understand I can adjust the print area in the Y direction, but how does it adjust it? Will it make it smaller based on the centre line of the x axis, or will it just trim it off the front?

It would be nice if I was able to set the extruder 1 offset from the, for example, X gantry centre line and Y gantry centreline......and for cura to work out the printable area from that (and based on the second extruder offsets from the first).

Edited July 16, 2015 by Guest