Ultimaker Mark2 - The Smart Dual-Extrusion Upgrade

[RudydG 62]

Regarding the printmaterial to use in the critical parts, i experienced:

Polymaker PLA is very good in tolerances, but not temperature stable, i got troubles after a long print (60hrs), the magnets in the coupler came out of their place.

Polymaker Polymax, is a better material imho, but there i had to re-adjust some dimensions in the drawings to get 100% play free coupling. (succeeded in this)

The dimensions i adjusted are the following:

-Coupler: lower magnet holder, reduced the slot where the magnets goes in with 0.1mm (was 6.00, is now 5.90) No play anymore;-)

-Parts you glue on the coupler, i reduced the holes to 3mm diameter, and added 0.25mm in the part with the hole in it (...) so there is less play in this part also.

And ofcourse, i ordered the greentec filament to be safe for the future ;-)

I must admid that i am a picky guy, the dual extruder parts that i print at this moment are already very nice. Maybe the tolerances i want to archieve are not neccesary, but i want it just to be good!

At the same time my S3D UM mk2 profile seems to get better and better.

[ultiarjan 1,223]

 

@RudydG can you share you latest S3D profile, I like to give it try...

[bob-hepple 59]

Decided to Order Ngen to print my Mark 2 head ???

 

PLA tec has a higher temperature tolerance... I would personally not use Ngen for this.

 

Thanks for your Honest Opinion, one of the reasons for not taking advise (neotko) was the link shown for the white was not in English so didn't want to just jump into the Dark.. I Found this on a sight I have purchased from before

https://www.3djake.com/extrudr/green-tec-red

https://www.3djake.com/extrudr/green-tec-white

I presume this is the same Material if so I will order , as always thanks for the comments

[MarcoUbtz 15]

@RudydG can you share you latest S3D profile, I like to give it try...

 

I am working on a S3D profile too, I'm focusing on tool change script with idle temperature and a soft acceleration/jerk control.

The latest script is good but not yet perfect.

My current version works good. There was some ooze issues after the printer reads T0/T1 beacuse it unretract immediately the new extruder.

With regard to the expected time and the real one, the maximum incongruity was +8 minutes.

If anyone will try it, let me know if there will be some problems.

https://www.dropbox.com/s/g3bs5chtbjl1qlk/Mark2_1.1.5.fff?dl=0

[ultiarjan 1,223]

https://www.3djake.com/extrudr/green-tec-white

I presume this is the same Material if so I will order , as always thanks for the comments

 

Thanks for that link.

I see they have free shipping to the Netherlands, will order here next time as the shipping extrudr is using (post.at) is crap, last order took over 2 weeks...

[neotko 1,417]

 

https://www.3djake.com/extrudr/green-tec-white

I presume this is the same Material if so I will order , as always thanks for the comments

 

Thanks for that link.

I see they have free shipping to the Netherlands, will order here next time as the shipping extrudr is using (post.at) is crap, last order took over 2 weeks...

 

2 weeks! Wow. I also order most from 3djake, free delivery to spain for 45€ or more, it has save me more than once when I get a sudden big order with an almost depleted spool.

Edited May 26, 2017 by Guest

[chiz 20]

Today, I have replaced the head with a new one (printed using Polymaker PC-MAX) with a new set of magnets that's suppose to be N52 rated. Personally, I don't see any difference from the previous magnets that I used. I've also replaced the feeder with the adjustable version. However, my 2nd hotend still falls off when the printhead goes near the extruder (rear right of the frame). I guess I'll be needing a longer bowden tube, or perhaps a less stiffer replacement.

Anyways, I calibrate the hotends by eye as I don't pretty much get how calibration works as described in the calibration section in the Mark2 documentation site. I get that the calibration gcode is sort of a "ruler" but I don't get which number I should be keying in the extruder offset. How does -2 value for X corresponds to the circled line in the pic?

Edited May 27, 2017 by Guest

[MarcoUbtz 15]

I get that the calibration gcode is sort of a "ruler" but I don't get which number I should be keying in the extruder offset. How does -2 value for X corresponds to the circled line in the pic?

Check this https://github.com/foehnsturm/Mark2/blob/master/calibration/Mark2-XY-Calibration.pdf

It is +14 lines in Y direction and -2 lines in X one.

[MarcoUbtz 15]

Can any of the developers or beta testers help me understand how these lines of code are to be interpreted?

I know they represent the shadow area occupied by the docking stations and the non-printable area because of the tool offset.

I understood the way they work but I can't go further.

[-115,  112.5], [ -10,  112.5], [ -10,  72.5], [-115,  72.5]],

[ 115,  112.5], [ 115,  72.5], [ 15,  72.5], [ 15,  112.5]],

[-115, -112.5], [-115, -87.5], [ 115, -87.5], [ 115, -112.5]],

[-115, 72.5], [-97, 72.5], [-97, -112.5], [-115, -112.5]]

Edit_

I understood, question solved.

Edited May 27, 2017 by Guest
Solved question

[foehnsturm 969]

@chiz, Can you make a picture or video, showing the rear right corner issue?

[chiz 20]

@chiz, Can you make a picture or video, showing the rear right corner issue?

Here. (Sorry for the awkward angle. We are moving in a few days and we have boxes with address in the background so I had to rotate the printer preventing the boxes to in the video.)

And here's how I mounted the adjustable feeder holder. I even trimmed the bottom screw portion to maximize the angle.
[attachment=46958:name]

Will probably get back with fixing this issue once we are done with the relocation. Sadly, after I have the Mark2 mod done, I now need to knock down the printer for the move.

Edited May 27, 2017 by Guest

[RudydG 62]

@chiz, Can you make a picture or video, showing the rear right corner issue?

Here. (Sorry for the awkward angle. We are moving in a few days and we have boxes with address in the background so I had to rotate the printer preventing the boxes to in the video.)

And here's how I mounted the adjustable feeder holder. I even trimmed the bottom screw portion to maximize the angle.

Will probably get back with fixing this issue once we are done with the relocation. Sadly, after I have the Mark2 mod done, I now need to knock down the printer for the move.

Are you sure the magnets are strong enough? seems to me they are not. i have to pull real hard to seperate them. pulling force of 0.9 kg is what is advised.

did you try feeding from another direction? so that the naturely bend from the spool is in the same direction as the bend in the bowden tube?

[chiz 20]

I'm actually wondering how I'll be able to measure the magnet's pulling force just to be sure. Here's the link to the N52 magnets I bought:

http://www.ebay.com/itm/162027521920?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[RudydG 62]

I'm actually wondering how I'll be able to measure the magnet's pulling force just to be sure. Here's the link to the N52 magnets I bought:

http://www.ebay.com/itm/162027521920?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Did you placed all the magnets (they are always pairs) in the same direction? It was not stated in the build instructions, but my believe is that you can actualy lose some pulling force when they are not all in the same North south directions.

I tried how much force i needed to seperate the 2 heads in the upper right corner, and to be honest, the force to undo is not big, so i can imagine, that if the magnets are just a little weaker, you can get in trouble.

Did you try unloading the filament and do a "dry run"

These are the magnets i bought:

https://www.magnetexpert.com/circular-disc-rod-magnets-c34/6mm-dia-x-3mm-thick-n42-neodymium-magnet-0-9kg-pull-p3566

Beatifull packed in a plastic box, with safety instructions. All indications of a quality product.

[chiz 20]

hmmm... Considering they're only N42 (as the page says and not N52), they're pretty impressive for their holding power.

Anyways, yes I ensured that the poles are opposite. I even ensured that the magnet pairs touch each other completely while making sure there's the "air gap" between the head and the coupler.

I did do a dry run with the bowden only and without the filament yet. But with the combined bowden+filament, it has become more stiff. I bought a replacement bowden that I've been using on my UMO and has not failed me so far. They're less transparent than what Ultimaker supplies but they're more slippery and less stiff. The tube comes 5ft long so I'll cut the tube slightly longer than stock so it will bend easier.

Edited May 28, 2017 by Guest

[RudydG 62]

hmmm... Considering they're only N42 (as the page says and not N52), they're pretty impressive for their holding power.

Anyways, yes I ensured that the poles are opposite. I even ensured that the magnet pairs touch each other completely while making sure there's the "air gap" between the head and the coupler.

I did do a dry run with the bowden only and without the filament yet. But with the combined bowden+filament, it has become more stiff. I bought a replacement bowden that I've been using on my UMO and has not failed me so far. They're less transparent than what Ultimaker supplies but they're more slippery and less stiff. The tube comes 5ft long so I'll cut the tube slightly longer than stock so it will bend easier.

Well, maybe the stiffer bowden, (i use the original brand new bowden), combined with the direction of the bend which the filament has can cause your situation.

I am pretty sure, that, unless your magnets are to weak, it must be a to stiff bowden "combination"

[RudydG 62]

hmmm... Considering they're only N42 (as the page says and not N52), they're pretty impressive for their holding power.

Anyways, yes I ensured that the poles are opposite. I even ensured that the magnet pairs touch each other completely while making sure there's the "air gap" between the head and the coupler.

What i meant here, was are the pairs all in the same direction?

That they are all pairs is ofcourse a basic, but if they where not, you couldn't even pick the head up ;-)

[foehnsturm 969]

Most likely the head pivots around an axis, defined by the top left and the bottom magnet. This is the same movement as in the undocking process. The amount of possible movement is mostly defined by the gap shown in the picture. The bigger the gap the more easy the top right magnet can be pushed away.

[harmenschippers 11]

Hey Guys,

I'm almost done with adding the Mark2 modification on my Ultimaker 2+ extended. It's been going very well. I only have to do some wiring and then I hope I'm able to start calibration this week and do my first 2 color print :-).

Just wanted to say that I've printed every part with Ultimaker PC filament and everything went pretty well, but I had to print every part individually for good results. I also printed part "head-a" a little different. I turned it 90 degrees so that the 4 mounting holes are on the bed. I only added support for the big overhang.

I also noted that I don't have an extra hole in the back for the wiring so I will print a cable tray and route the wires through that to the underside of the printer (but first I will test the upgrade, because I can't wait any longer :-)). I will post it later on for people who may also need this cable tray.

[MarcoUbtz 15]

I would like to report a problem that I have found and solved.

For those who had wrong z positioning after a tool change [T0 to T1] with a gcode created in S3D, I found the cause, pay attention, you have to set to ZERO the "tool change retraction distance" in "advanced" tab.

[navigator 24]

I would like to report a problem that I have found and solved.

For those who had wrong z positioning after a tool change [T0 to T1] with a gcode created in S3D, I found the cause, pay attention, you have to set to ZERO the "tool change retraction distance" in "advanced" tab.

Nice find - been trying to sort out why that's been happening for ages now.

Update: Just checked and mine was at zero, still got Z height problem with S3D - Cura works fine :(

Edited May 30, 2017 by Guest

[neotko 1,417]

I would like to report a problem that I have found and solved.

For those who had wrong z positioning after a tool change [T0 to T1] with a gcode created in S3D, I found the cause, pay attention, you have to set to ZERO the "tool change retraction distance" in "advanced" tab.

The problem is related to how S3D uses processes as independent prints. Basically things (when using more than 1 process) that the tool has change

https://ultimaker.com/en/community/39416-simplify3d-retraction-bug-that-can-drive-you-nuts?page=last

[MarcoUbtz 15]

The problem is related to how S3D uses processes as independent prints. Basically things (when using more than 1 process) that the tool has change

My english is not good, so sorry if you will not understand me at 100%.

Practically, I seen there are some interferences between S3D's gcode with a tool change retraction/script and the script that is stored and beautifully implemented in the firmware of the mark2.

It happens that, before get/park the 2nd head the printer raises the head/s from the object, it moves to the wipe position [with the correct offset on the X-Y-Z direction] but after it heats and unretracts it doesn't reset the initial lift.

I don't know precisely the steps that the machine performs on a tool change, but from its behavior I can imagine it and I can adjust my script.

Now, without the toolchange retraction distance I found that it works -not good not bad, simply it works-, with a toolchange retraction in S3D you have to insert a script in the toolchange tab that remembers the machine to adjust the Z after a tool change, simply a "move to next position X Y Z".

However, it was right to alert other users of the possibility of a mistakes.

With my latest S3D's profile [not yet shared] these issues don't happen but with old ones it is possible.

[MarcoUbtz 15]

Update: Just checked and mine was at zero, still got Z height problem with S3D - Cura works fine :(

For the moment, leave the "toolchange" tab empty as had suggested tinkergnome, soon I will finish the new S3D's profile for the mark2

[ultiarjan 1,223]

For the moment, leave the "toolchange" tab empty as had suggested tinkergnome, soon I will finish the new S3D's profile for the mark2

I think I'll wait for a while and let you do the hard work