Super hard nozzles - Test pilots

[Izzy]

will do, PLA worked great on numerous makes, have tried ColorFabb XT White and looks good, I have found that I need to run about 5 degrees hotter than my previous settings, but getting better overhangs and on the robot the 'ears' and 'antenna' are better than previous brass nozzles only difference is the 5 degrees hotter, I'm away for the weekend but will try some more next week and try some flexi.
My wife asks are you also going to do sapphire and emerald too

[martin-bienz]

ok, I have printed my first 2 things. I am happy to report, the results are great. While I would also say I require a bit more heat (+5°) for the otherwise same settings, but that might be me paying to much attention to the print process. Overhangs are indeed fantastic. As the two objects don't really have much of a top layer, I don't feel there was a lot of difference. Retraction plays nice, almost no stringing.
I used blue PLA only so far. The only thing, it curled quite a bit more at the start than with my other nozzle (jet nozzle from 3dsolex) but I only adjusted the temperature afterwards.
Thanks again Anders, looks great!

[Dim3nsioneer]

I did some first prints as well.
Curling around the the tip does indeed seem to be an issue. With my ruby nozzle, it's independent of the temperature. So I get the effect that soft filament is collected in the gap between the ruby tip and the brass until it gets too much. It is then deposited as a blob on the print. Especially with XT-CF20 (otherwise a very nice material which I'm printing for the first time) this curling is very strong.
I had started my tests with sky blue PLA, imho the most difficult color to print as you see every tiny fluctuation in temperature or filament diameter at once. I printed the @Ultiarjan overhang test and got an excellent result, especially for the overhangs. The same is true for an overhang test with XT-CF20 although I had accidentially 107% flow and got some overextrusion.
One can minimize the curling/blob effect by very fine tuned bed levelling. However, the window between non-closing lines and too high blob-building is very narrow.
My first impression summarized: A very nice tool but not for starters.

[Neotko]

Imo carbon filaments tent to curl around the nozzle no matter the nozzle kind. I print the first layer slightly underlevel to avoid the first layer scratches going on the nozzle and I pick the nozzle build at 1h intervals. That happens also o e3d brass nozzles at the same rate. Carbon partially unsticks everytime the nozzle passover and leaves residue that the nozzle picks. Try your carbon on an old brass nozzle to check it aince a 2h prints already softens the tip even if it's new.

[martin-bienz]

That happens also o e3d brass nozzles at the same rate.

I can confirm, that using the e3d brass nozzles with my UMO also curled way more than with 3dsolex / the standard UMO nozzle.

[Anders Olsson]

Thank you for the reports so far!
I also thought I saw some more curling on the last batch when priming the nozzle, but I haven't actually compared them to earlier ruby nozzles yet. Some filaments are much more prone to curling too (as Neotko writes), so one has to have a consistent way of comparing them.

It is possible that there is some more curling as a result of variations in the thermal conductivity (which should be better on the 1.75 mm version) or because of varying dimensions of the brass components. The internal geometry did not come out exactly the same on all nozzles in this batch.

[Meduza]

XT-CF20 is the filament that has curled the most for me also, it always makes a mess of the outside of the nozzle (at least it did with the old sapphire nozzles)

[Amedee]

The postman brought the jewel Friday...

My daughter needs the UMO+ for a school project, so I can't touch it these days

I eventually decided to mount it on my old UMO, without heated bed I am limited in what I can test, but on the other hand I guess not a lot of people tested that configuration, so it might be worth trying, and definitely better than just looking at it.

So here here is my short report:

The device as a got it -- so we can make a before/after comparison after some time:

Ruby ruby ruby....

I followed the instructions to mount it...

Handle with care

I also used my preferred test filament, the Ultimaker blue

First thing, bed adhesion is a bit tricky. If the bed level is not spot on, it won't stick. Might be a challenge if there are a lot of islands in the print.

When it sticks it s quite amazing, the first layer is very flat, very nice:

First layer

But then it suffers from under-extrusion:

Under-extrusion

There are two things here:
- One thing is thermal conductivity -- I had to bump the temperature by about 10°C to get the same flow as I normally would have. This is not a big issue for PLA, but I am a bit more concerned for other material that I currently print in the 240-245 range: that would push me in the area where the PEEK (and the Teflon) are not too happy.
- The other part is purely the flow, where I added 10% -- I guess it is probably not new, but the better finish from the nozzle makes you look closer at small issues.

So we have now a nice surface:

Better!

Here is the first print finished -- don't look too much at it: I needed a shorter fan shroud for the shorter nozzle, so I took the 'official Ultimaker' one that I compressed, it is printed without support and without fan which is not ideal for that part -- even more when 'compressed'. Just happy it finished so I can print back with fan

Finished

More tests this week...

[Neotko]

Today I was testing my v3 um2-fancap-for-umo (symmetric yeahh) and did some Ruby vs E3D. Nothing new there, ruby rocks



LEFT fresh printed Ruby - RIGHT E3D nozzle

Both with the same gcode, and I'm quite proud of my fancap because that's printed at 50mm/s for all the lines, 3 perimeters, 0,1 layer, 2.5mm retract and 39% infill. Finally finally I'm almost on a production print speed with good overhangs. Because printing that at 30mm/s it's too easy, but at 50mm/s now that's something and a half

The only special option activated was Print islands sequentially without optimization (to avoid overheat from layer to layer).

Used my ugly but very nice to debug Coral PLA smartmaterials pla. It's soo good because it prints just like faber from the technical point (but ofc it's ugly coral color).

[LePaul]

Well the one on the right has a much tighter skirt (I know, you removed it and placed it there to compare!)

[LePaul]

Here is the first print finished -- don't look too much at it: I needed a shorter fan shroud for the shorter nozzle, so I took the 'official Ultimaker' one that I compressed, it is printed without support and without fan which is not ideal for that part -- even more when 'compressed'. Just happy it finished so I can print back with fan

More tests this week...

You know I never thought about that...the nozzle is shorter than what I have in my UMO so I will also need to (further) tweak my fan solution.

Your comments on temperature are noted...I wonder how long my PEEK would service using the CF filaments that require a lot more heat too.

[Neotko]

At 255C after 10h I started to see peek deformations near the base of the peek that touches the barrel. But at 245-250 the deformation stopped.

[Amedee]

I have printed many hours (200+) at 245°C without problem. 245°C should be OK for CF unless we need to bump the temperature due to not-so good thermal conductivity.

Unfortunately I can't test CF on my blue-tape UMO right now.

@Neokto, which temperature did you use for the robots?

[Neotko]

The heat was at 200C for e3d and 205 for ruby.

On umo I can usually print 5-7C under but hotarea of um2 hotend it's smaller and needs a bit more.

[Dim3nsioneer]

I ran an interesting comparison. It's Ultiarjan's overhang test with black PLA. I appologize for the sub-optimal picture quality (I'm better at 3D printing than at taking pictures).
On the left you can see a print I did some time ago. Overhangs were average but stringing was very low, independent of the temperature.
On the right you see a print done now with the Ruby nozzle. Overhangs came out much better but I got an immense stringing. Both prints were made from the very same PLA.

As I have printed XT-CF20 before which showed also nice overhangs but lot of stringing, I performed a series of cold pulls. Now, it's currently printing the overhang test again in order to check if it was an effect of the XT-CF20.

I have to mention that the Ruby vs. Brass nozzle is not the only difference between the two prints. The left print was printed with the stock UM2 fan shroud while the right one was printed with Labern's fan shroud design. However, an earlier test showed overhang quality was about the same for both fan shrouds but the same (Brass) nozzle.