illuminarti
-
Posts
2,785 -
Joined
-
Last visited
-
Days Won
1
Content Type
Forums
Events
3D Prints
Posts posted by illuminarti
-
-
I've often wondered if the 'side facing the fan does better' story is really true? Any part of the print should get pretty much equal instantaneous cooling as it gets extruded, since the fan is right there by the nozzle - although admittedly a bead being extruded while the head is moving in positive X has the least cooling active fan time (since it soon disappears under the fan), while a bead that is printed in negative y direction enjoys the longest potential cooling as the fan remains close and moves along the freshly printed bead as it goes (since most of the fan opening is usually 'behind' the head in the y direction).
But, in terms of position on the bed, the parts of the print on the left (towards the Y axis) typically gets the least cooling overall, since it is most often under/behind the fan opening when the head is printing the parts of the print on the right. And that means it gets pretty much no fan assisted cooling. Whereas the parts of the print on the right are pretty much always getting some fan-assisted air movement over them, even if the fan has moved away somewhat. They're never in the shadow of the fan itself.
Certainly what I have seen regarding warping is that if I'm going to get curling up off the bed, then assuming similar geometry across the bed, I'm most likely to see warping on the right side of the print, since it gets more cooling overall from the fan. (I can get rid of the warping in those circumstances by turning the fan off completely).
I haven't studied it in detail, but I suspect the truth is more complex than 'the side facing the fan prints best'. At the very least I suspect it depends on the direction that the print head is traveling.
Anyone have any thoughts or examples to share?
-
I suspect that it's a problem in the slicing itself - your settings look sane, and the minimum layer time of 7s should be giving long enough for it to cool between layers. I think that probably the head is backtracking to start the new layer, and its causing problems. This can happen with shapes like this, where the print segments are long, but the corners precess around the shape.
Older versions of Cura had problems with this and would keep trying to move back to the vertex nearest the origin to start the next layer.. that may be what is happening here too. When the head moves back, it drags back over the print, oozing, and making a mess.
Can you post the gcode here as an attachment, or email it to me at gcode@fbrc8.com, and I'll take a look to see what's happening?
-
Certainly looks like you're getting some nice results!
-
Brim may be overkill - it's normally set to a lot of loops, and its attached to the print. I usually find that a skirt of 3 or 4 passes is sufficient, and you don't have to worry about separating that from the print afterwards.
The other thing that will probably help is to increase the amount of priming extrusion in your starting gcode. In cura, there's a line in the starting gcode that looks something like:
G1 F200 E3 ;extrude 3mm of feed stock
I prefer to change this to:
G1 F90 E6 ;extrude 6mm of feed stock
This extrudes twice as much plastic, and at a slower pace that won't over-tax the nozzle. You could also add a delay command after that, to give yourself a chance to clean up the nozzle for easily...
G4 P2000 ; pause 2000ms
-
Yes, but I think maybe you're missing the point about wall thickness settings... your STL has a 1mm wall defined in the file. But what is your 'skin thickness' set to? If that is also 1mm, while your nozzle is set to 0.35mm, then I think that Cura is going to probably change your nozzle thickness on the fly to 0.5mm (at least that's how it worked in older versions - see this write up for details. Note that the bugs described there are now fixed in the latest Cura, as far as I know, but I think it probably still works the same way in terms of nozzle/skin thickness interaction).
Because you can't fit two 1mm thick walls inside a 1mm space, I guess it's just doing a single pass in that area. Maybe with a bit of under-extrusion... but Cura probably doesn't think there's a gap in there that needs to be filled. Traveling along that thin wall may indeed be a problem though.
George is right that almost certainly you don't really want a 60% overlap - if its needed, then you're probably masking problems elsewhere (most likely backlash). If so, it would be better to solve the real problem - conventional wisdom says a 15% overlap is ideal.
And also, consider whether you really want a 'skin thickness' setting of 1mm if that's what you indeed have. I think it's going to cause problems in a model that only has 1mm as the physical thickness of some walls.
-
I think that this was an inevitable direction for the company - especially once they started taking on investors a year or two back. I think I'd probably have done the same thing, if I'd been in their position. They've obviously been trying to make themselves seem more corporate and 'professional' although it seems to have come at the expense of engagement and drive - personally I thought their presence at Maker Faire was utterly lackluster this year. Compared to the buzz around Ultimaker, and some of the other printers, a couple of corporate-looking guys standing behind a table with a Replicator 2 on it didn't really do much for me.
Thingiverse may be the more interesting part of the equation for Stratasys. There's nothing technically innovative about the printers really, compared to the technologies that Stratasys already has. Makerbot probably has more awareness built up than most other brands among the general public, but its still virtually zero; there's not a lot of advantage in that. But Thingiverse would give Stratasys a way to control and monetize models for printing - they probably see it as a fledgling 'iTunes Store for 3D printing'. At the very least, they'd probably prefer that even if for now it remains free, they want to be the ones controlling it, rather than a competitor.
I imagine that we'll start to see a lot more DRM-style controls getting introduced as designers try to monetize their models, and printer manufacturers try to find a way to earn on-going revenue once they have sold the printer (and yes, I expect more printers will start switching to proprietary filament cartridges so they can own that revenue stream as well).
-
It looks like the exit area is quite small compared to the intake area of the fan. I remember seeing discussions in the past that these sort of designs didn't work well because the standard 40mm square CPU cooler fans aren't capable of sustaining the necessary pressure to handle much reduction in output area. Does it work ok, in practice for you?
Do you have any results showing the improved print quality that these designs makes possible?
-
If you've already checked that the belts aren't rubbing, then the most likely cause is that one of the X-Pulleys is loose.
Check that they are all properly tight, including the one on the motor shaft itself. I gather that UM now includes replacement grub screw for the pulleys, in case the ones included in the pulleys by default don't fix tightly enough.
I would try making a mark on each pulley and the shaft it is attached to with a sharpie, so you can see if there is any movement during the print. And I'd definitely avoid glue. Not sure how well that's going to attach anyway, but I think it's going ot cause more problems than it solves.
-
As an example, here's a print I did yesterday at 0.06mm layers. Generally I print at 0.2 - 0.1mm, but recently I've been experimenting with thinner layers.
This was from a 3D scan I did, sliced with Cura 13.06.2, and then printed pretty quickly - 95mm/s, 2 loops (of my 0.65mm nozzle), no infill, no support, 0.6mm top, but with the solid base not printed. I was pleased to see that the overhangs mostly printed fine (just a few droops to cut off under the chin), and even the top of the head closed up perfectly. Total print time was about 2 hours.
For figurative pieces like this, printing without infill is definitely the way to minimize print times, and not printing the base saves a lot of time too, where you can get away with it - and it helps reduce warping stresses by avoiding the solid mass of plastic.
-
I generally consider extrusion volume in terms of bead width x layer height x linear speed. But that should be identical to the requested filament speed x filament diameter. My testing has shown that in practice the delivered speed is rather lower than requested speed due to head pressure causing filament slippage in the extruder. I find I get 5% - 15% underextrusion as you request higher speeds, until the filament strips totally.
I havent quantified the difference between melt time and nozzle pressure in any scientific way, but it's pretty easy to see it in practice. If you lower the bed, and heat the head to say 230ºC and let it sit for a few minutes, and then begin to turn the extruder gear by hand. You'll see that at first the (thoroughly heated) plastic runs out quite easily and then as fresh unheated plastic moves into the melt chamber, the rate of extrusion slows dramatically, and the pressure goes up, and the texture of the plastic extrusion changes to become much thicker and viscous. I take it that this is because the plastic isn't actually able to reach the target temperature in time before it is extruded.
-
No, the bed leveling and height really only affect the first layer or two - if those print fine, then the rest should also. If you were having problems later in the print it's probably due to issues with your filament, or perhaps a pressure build up, or some other sort of nozzle blockage or feeder problems leading to under extrusion.
In this case, it was probably due to trying to print too fast - that is, too high of a volume of plastic per second. The most you can reliably print is around 8 to 10 mm³/s with a stock nozzle. With a bead width of 0.4mm, and a layer height of 0.2mm and a speed of 100mm per second, you're printing 0.4 x 0.2 x 100 = 8mm³/s. If you increase the layer height to 0.3mm, that increases to 12mm³/s - and that causes very high pressures in the nozzle, and can lead to stripping of the filament in the extruder drive, causing under extrusion and even blockages.
-
What did you like about the skin option? As implemented in the old version of Cura, it just printed the same perimeter twice each time.
-
I presume that the STL file you are printing actually has the walls modeled as thin structures - it's not a solid block, that you're only printing the outside edge of? If so, how thick are the physical walls in the model? The 'wall thickness' in Cura is now renamed skin thickness, which better indicates what it does - it's the thickness of solid plastic you want behind each finished surface. Typically you want this to be one or two times your nozzle size, to get either one or two passes of the head. This is nothing to do with the thickness of the walls in your STL file, unless you have a solid object that you want to just print a thin wall around (in which case you set the wall thickness here, and then set 0% infill). If, as I suspect is the case here, you have 1mm thick walls in your STL file, and have the 'skin thickness' set to 1mm, you're actually asking for something impossible - 1mm of plastic loops behind the outer wall, and a separate 1mm behind the inner wall, which would overlap. I expect that Cura is able to handle it, but it's worth bearing in mind.
The newer version has a totally different slicer engine, so it handles small gaps differently.To be honest, the gcode pictures you posted for 13.03 and 13.06 look to me like they were the wrong way round - usually 13.06 zigzags to fill in small gaps that earlier cura's left empty. Maybe your excessive wall thickness is confusing it, though - it is still in beta?
What I think is happening may just be that it's printing each layer a lot quicker, since it isn't doing the zigzag infill, and so the plastic isn't having time to cool between layers. Try setting a longer minimum layer time, so that the plastic can cool between layers.
-
Also, not to hijack the thread, but I'd be interested to hear your experiences (and anyone else's) experience with Jeremie's blocks, as I'm considering installing those when I take everything apart to install the new pulleys.
-
Looks really awesome, Lars.
One thought... by minimizing the melt zone, what does that do to your ability to print quickly - i.e., at high volume per second. In my testing with a 0.65mm nozzle, the speed at which I could print seemed to me limited by the ability to melt the plastic before it got to the hot end.
For instance, trying to print at 30mm³/sec means advancing the raw filament into and through the print head at roughly 5mm per second. I seemed to be running into a problem that the standard head couldn't reliably melt the filament in the time it took to go from entering the hot zone to exiting the nozzle, due to the time it takes to conduct the heat into the filament.
Minimizing the hot zone would presumably give even less time between when you start heating the filament, and when it gets extruded. As such, I suspect that the peak throughput of the nozzle may be reduced somewhat. How does it behave in practice?
-
Happens for me to.
I think that maybe the 13.06.3 install disk actually has 13.06.2 on it... but I'm not exactly sure what's going on. I expect Daid will have it sorted out soon.
-
I'm testing 13.06.3 (which seems to be (or think it is) 13.06.2 for mac, in fact). I've been printing some 3D scans of people, fast, 0.06mm layers, with a couple of loops, and no infill. In general it's coming out amazingly well.
However, I notice that it's not doing retraction for the interior moves, and there doesn't seem to be a way to enable that as there was in earlier versions. I'm printing with no bottom, so the inside space is visible, and it's making for quite a messy interior to the print, with lots of stringing.
I'd kinda like the option to turn on retraction, or alternatively, perhaps retraction should be enabled automatically for moves for prints with 0 infill, and an open top and/or bottom?
-
Can you post some photos of the extruder drive mechanism, showing it open and closed? Also, what speed settings are you using?
Perhaps you could attach a gcode file as well, so we can see what settings you are using?
-
No it doesn't affect acceleration - it's just a multiplier for the target speed of all moves. (Printing and non-printing).
-
No, there's a particular gcode that cause the 'heat and wait' behavior. If that isn't there, it won't heat (or wait). I strongly recommend that you take a look at the gcode that your slicer is generating, and compare it to a reference list like this one, in order to get a feel for how it works. Its really very simple. The only thing to remember is that speeds are in mm per minute (i.e., 60mm/s = F3600), and once a speed is set using an F parameter on a move command, it stays set unless changed, even if it isn't given on subsequent lines.
That said, here's the code I'm using these days:
; Move to home and a relative height of 40mm
G28 X0 Y0 F3600 ; home x and y
G91 ; relative position
G1 Z40 F600 ; move up 4cm
M84 ; release steppers
It's also included as an attachment, below, that you can copy onto your SD card.
-
Set origin was removed in more recent versions, because it was not desperately useful in most cases. You use prepare -> auto home to find the physical endstops and set the 0,0,0 point.
Moving the z-axis using the Ulticontroller doesn't always seem reliable in my experience. It can sometimes lead to missed steps (the motor makes a chirp sound, instead of moving completely). Usually this isn't critical, if you just want to get the head out of the way.
Another alternative that I find useful is to hand-craft a gcode file to move the head to a standard position- e.g., home x and y, and move the head up 20mm from whereever it currently is. I save that in a 'Utility' folder on my SD card, and then, for instance, if I stop a print due to a problem, I can quickly run the gcode from that folder in order to get the head out of the way.
Finally, I strongly recommend against using the Ulticontroller menu to advance the extruder. It pushes plastic through the nozzle far too quickly, and can cause head blockages. Instead, just turn the big gear by hand (you may need to select Prepare -> Disable Steppers first, or wait a minute for the motors to turn off, otherwise the gear can be hard/impossible to turn by hand). If you turn it by hand, you get a better idea of the rate of feed that the head can handle - you'll start to feel resistance if you turn too fast, and can adjust.
-
I don't know much about netfabb, but you will want to make sure that you have retraction enabled; I seem to recall that for a long time netfabb didn't do retraction and/or didn't do it right. Are you fully up-to-date with the latest version?
-
@12much - weird, that's what I'm using too, and I'm not seeing problems. Java shouldn't affect it - it's a javascript program, which is totally unrelated to Java. (But, wearing my day-job hat as an Internet Security consultant, I definitely recommend keeping up with Java updates, yes, as there have been a bunch of Java-related security exploits lately).
-
Yes, the sound should be a continuous, smooth note as the bed rises. Any noise you're hearing could either be directly mechanical (something catching, for instance) or a momentary change in the stepper sound perhaps due to steps being missed. However, aditionally, using the Ulticontroller to move the bed up and down is problematic. The code that converts knob rotation into movement isn't great, and I think it can artificially introduce problems. So, I'd be inclined to use something like pronterface to drive the z-axis for testing - that also lets you control the speed of the movement.
I'd be inclined to home the platform, then move it down, slowly, by 1 cm. Put some blue tape on the back of the printer, and mark the position of the bed on that. Then slowly move the bed down 15cm (hopefully hearing a smooth sound), and mark that position. Measure the distance, and make sure it moved exactly the right distance. Then move the bed up and down the 15cm amount, at various speeds, and see if seems to move smoothly, and end up in the right place. If the movement distance is the right amount, and the bed ends up in the right place, then any noises are mechanical, affecting the smoothness of moving, but not causing skipped steps. If the bed doesn't end up in the right place, then you're skipping steps.
Complete Hot-end and Print Head Redesign
in Third party products & modifications
Posted
Yes, but why not aim to print fast, and create strong, reliable, good looking parts?
That was why I asked you a long time ago, when you first started your hot-end redesign project, what the real-world problems were that you were trying to overcome. It remains to be seen how things turn out in practice as you test more, but it may well be that by minimizing the heated area in the head, you have perhaps solved the (for me) almost-non-existent clogging problem, at the expense of even lower sustainable print speeds.