Lulzbot mini poor fine detail

[mrblac]

Hi, been trying to print this spitfire mast, which is really basic. But the mini keeps doing a crap job. Or maybe my programming of it in cura is crap too…
best print result I’ve had so far in PLA is speed: 20, heat 190 at a.09 layer thickness. But again, still not even coming out close on the mast as it gives really crap results. I’m thinking the temp is the main reason. so now its at 190, but how low can i go on heat with pla? thinking it would need 175 because there was a major jump of improvement from 205 to 190.
I am also printing this through octoprint too using the slicer in cura though. Had equally bad results when just doing it in cura too.
Anyway, I will add pics below of the settings I used in cura and how the mast has came out. Any help on getting the mini to do what it’s meant to would be awesome…

Cheers

[Graham]

Hi @Mrblac, can you share your STL?

Use this upload tool

[Sam]

@Aidan might be able to shed some light, he wants to marry his Mini.

[mrblac]

No I can’t sorry, I’m out all day. But if you goto thingiverse and just type in search “spitfire mast” it will come straight up. Also I wanted to upload 2 pics of the basic and advanced settings I was using in Cura, but your forum would only allow 1 pic upload which is kind of redundant in trying to give you enough good detail to what’s going on.

[mrblac]

[Graham]

Thanks, how large are you scaling it too?

[mrblac]

I can’t remember if that was the correct scale… But it should measure about 17 mm or 17.5 mm in height. Upfate, sorry height size “75 mm”

[mrblac]

Also what is the lowest heat temperature in PLA the mini can handle printing at?

[Graham]

Thanks! The Mini, as with all FDM 3D Printers on earth, have limitations. Not everything is “printable” and this is by no means a limitation of the Mini, rather, it’s just the way things are given the limitations of hot plastic and small objects. The Lulzbot Mini sets the standard incredibly high for FDM printers, making things like this much more possible, but not guaranteed.

The key limitation is that the max “angle” the printer can manage vertically; 45 degrees. At 45 degrees, each layer has just 50% of the previous layer to bind onto. Supports make overhangs such as these possible, but come at the cost of removal (which is sometimes very tough if the layer you are removing them from is very thin, such as this design).

All of objects highlighted below can’t be printed without supports enabled, and even then, it’s going to be incredibly difficult to remove those supports as some of the layers the supports are connected to are less than <1mm:

The STL has other constraints, such as the width of the column is <1mm in some spots. The recommended wall thickness for FDM is about 1mm. You could dial this down to the width of the nozzle (0.50mm), however, heat and other factors make this a bit tough to deal with. You could build a mast on opposite sides of the mini, which would double the amount of time each layer has to cool (while the printer jumps away and begins printing on the other mast).

Perhaps the take away is; just because something is on “Thingyverse” doesn’t mean it’s “printable”. The entire design approach should be built around the limitations of FDM printing. In this case, I’d lead to re-designing the model, making all surfaces at least 2mm thick, stepped contours that are stepped at the layer height you’ve chosen. I’d build manual supports that start from the heat bed and go “around” the objects such as the large panel on the bottom (rather than building on-top of it all the way up to the antenna). This design isn’t printable without lots of compromise as-is, and that’s not a limitation of the Mini, it’s a limitation of FDM printing.

[Graham]

@mrblac

Also what is the lowest heat temperature in PLA the mini can handle printing at?

Perhaps the best thing you could do is print a cylinder and modify the Gcode to change the hotend temperature every X millimeters. You could test a range of temps, but definitely stay within the recommended temps printed on your filament reel / as recommended by the manufacturer.

The smoothest finish would perhaps be the sweet point, however, that would be relative to the cooling time between each layer application. So while it would be good for the cylinder, it might not be as relevant when compared to printing something much larger / smaller. Trial and error is the only way to really find the perfect temperature, and we find it rarely matters if the model is designed with heat management in mind.

[mrblac]

Are you kidding me?? It prints with supports because I have that option picked. Plus I’m printing it solid now too, so there’s no problems on small wall thickness as well.

Like serious this is a basic, basic little 3D print for any printer. If Lulzbot printers are too shit to handle basic concepts then I’ll go get a 200$ printer off eBay and you can have the mini back. Did you actually try and print this yourself???
Surely with all your expert knowledge you say you have, you’d be able to prove for real one way or the other what the printer can do… Maybe I got a faulty mini and you need to do it on your printer to check if that is the case too…???

[Graham]

[quote=“mrblac, post:11, topic:263, full:true”]
It prints with supports because I have that option picked.[/quote]
Nice, I understand and was more-so leaning to how difficult those supports were going to be to remove from areas such as this (which are less than 1mm thick, it’s going to be challenging!)

I’m guessing no supports were built up to the top overhang (the actual antenna on the mast, at the top of the print); that would be due to the awkward placement of the ‘floating’ object below it (which could be moved easily enough in Cura).

Plus I’m printing it solid now too, so there’s no problems on small wall thickness as well.

That’s not entirely true. Towards the top of the print, the hotend will spend a lot of time going around in circles which can/will affect the quality of the following layers as the plastic hasn’t cooled enough.

The show-stopping issue is @17.5 mm height, this object is 3.62 mm wide, at it’s widest point. The remainder of this object is <1.5mm wide. The nozzle size on the Mini is 0.5mm, which leaves you with a max 7 lines of resolution at the widest part of this object (which is fine). Other parts of this print (the mast, etc) will be printed with more limitations, many areas are less than 1 nozzle width. This doesn’t seem like a normal object to 3D print, from my experience, it is going to be arduous and involve lots of tinkering with heat/material limitations which will get it over the line, but it’ll be a slog!

If you feel your printer isn’t working properly, then get the best quality print you can, send us the gcode, along with the details of material you are using. We’ll load up the gcode and same material. This will produce the exact same object, and isolate if it’s hardware or software.

With that said, you don’t have to engage us as the best way is to simply go back and print a known good “printable” object such as the Lulzbot Roktopus.

[Graham]

A quick note about designing models that are 3D printable, consider this recent print we did on the Lulzbot Mini (via Thingyverse user Geoffro). An incredibly detailed model, in particular consider the efforts Geoffro has applied to supports to make this print possible.

“Enable Supports - Everywhere” in Cura would not have made this print possible. Cura (any slicer) would have struggled to resolve support paths for those obstructed overhangs.

Below is our actual print of this model, the final print quality was sensational. Geoffro’s approach is a great demonstration of designing around the limitations of FDM printing.

[mrblac]

Ok thanks for all this, I did however lay it on its side to print so it had support the whole way along the mast area. Which again with printing solid has given me the best results by far.
Id say it’s about 85% good now. So I will go over the rest of your notes on the weekend and tinker a bit more with it and see if I can squeeze out an extra 10% to make it a good satisfactory print.
I’ll upload you my Gcode on Sunday night and by all means I’m happy if you can get it better than what I can. But I think I won’t be far off by then.
Thanks for the support and info.

Cheers
Darren