By the way I couldn't look at the zip file as I think you deleted it already - jeez give me 24 hours next time?
By the way I couldn't look at the zip file as I think you deleted it already - jeez give me 24 hours next time?
Soooory. The I put the file back.
Nico I haven't seen this issue. It's kind of weird. What's the initial nozzle temp? I'm wondering if it is taking a very long time to heat the bed and the nozzle is hot for too long?
My computer with the CAD software and SD card reader is on a different floor so my workflow usually involves turning on the printer, setting build plate to 65C, then removing SD card and walking upstairs. By the time I get back to the printer the build plate is almost up to temp. So maybe it has to do with not heating up the nozzle above 180C for very long?
I didn't measure the nozzle temp before a print. But alone from looking at the filament that starts to ooze only after the bed is hot enough, I'd say it is not a problem.
My workflow is similar. The only difference is the printer beeing next door. I take the sd card with me, turn on the printer, select the print and wait until the initial extrusion so I can grab the filament strand. Then I leave.
Btw, soooory about the file. I put it back (link).
I have a problem printing.
As I suspect it might be because of under extrusion i have been reading this tread with interest.
I have run the cylinder test today and made small change on the way starting of at
a best of 5 cubic mm at 230 and finally reaching 10 cubic mm by move the filament spool, losen the skrews, lubricating the filament with WD40 (by just spraying a little in my hand and pulling the filament trough so its just a thin coat), rising temperature to 240 and running motor at 1500mA.
I'm not quite there as some succeeding at 230 deg.
Retching 8mm was done without putting more power on the motor.
To reach 9mm i had to run it at 1350mA and to reach the final 10mm i moved it up to the 1500mA.
Several had problem running at 1500mA as it grinded the filament.
So, there is quite some resistance in my system but not more than It can manage to get the filament trough at 240 deg.
Maybe if I had run at 230 deg then maybe I had gotten the grinding problem as well.
Anyway.
A couple of notes.
Even if the spool is placed in god position there is quite some force needed to get it of the spool.
Think of the filament as a plastic hook gripping the spool.
I think a lot more force is needed to straighten out the filament in order to get it of the spool than what force is needed to get the spool to turn. If you understand what i mean. I made a test trying to preheat the filament in order to get the tension of it but as some one spot get softer than the rest it folds at that point. The other thought i had was to have the filament enter the bowen with it's bend matching the tube for least friction.
I think I have reached the limit for my machine for now. If there is more to get out of it i think it has to do with getting the filament out through the print head and what is a realistic pressure to get it out. No point in something else breaking using brute force. I will follow this tread, it would be interesting to see if a solution will be found.
For my own part I worry mostly about surface quality and not so much about speed.
But even so,I feel like I now has to print silly slow.
At least compared to when I it was brand new.
I'm however not sure if the problem is under extrusion or not as it seems so random.
I used to be able to print with standard 0.1 mm settings with nice surface.
Now i have to go down to 25% speed or print 0.02 layers.
This is what i get now if I print really slooow.
And this is what i get printing normal speed (50mm). Is this under extrusion or is it the printhead moving randomly for some reason?
And an image of the current spool solution.
It stands on two rolls using printed roller bearings.
Not super smooth movement so i might build a better one with real ball bearings but a lot better than the original.
If it matters considering the stiffness of the PLA.
I think the slightly blobby finish may be down to a lack of cooling time between layers. How big are the parts that you're printing? Have you set a minimum layer time in Cura?
What happens to the finish if you print the same piece at the same (faster) speed, but print two pieces at once?
Can't verify it now but i have all the settings at standard settings exept for shell that is set to 0,8. If 5 seconds is default i have it.
Will check after work.
By printing two at the same time, are you interested in what happens with retraction or just that a layer takes more time?
Exept for the rough surface the one printed att normal speed is more matte and the slow
print is more glossy.
The part is abot 50mm.
Longer cooling time.
Ok.
Printing two pieces at the same time at 50mm/sec but with same result as before.
So longer cooling time does not matter.
Thought that in case this is under extruding then using the the tricks i used for the cylinder test to get less under extrusion and finally reach 10 cubic mm per second should help here to.
So I made a test.
This time printing a single part again.
From down and up.
First part is speed set to 50%, the slow way of printing i used before to get something to compare to.
Then i increased speed to 100% and turned the heat up from 220 to 230.
Last part i increased to 240 deg.
When testing the cylinder 240 deg took me up to 8 cubic mm per second.
In this print I only used 0,4 x 1,0 x 50 = 2 cubic mm per second.
So, I'm quite sure that it's not under extrusion.
My first thought was that head was loose and wobbling or something but it feels stable.
But what is it, and if it is not under extrusion should i post in another tread?
Yes, why don't you start another thread. I suspect it's periodic over extrusion, due to excessive detail in the print path that causes head slow downs when printing fast, and those tend to blob a bit. Recent Cura's have had a problem that caused too much detail to get included in the path.
What version of Cura did you slice with?
I used latest version. 14.03.
I have noticed Cura getting a lot slower in it's handling in last version so I expected there was quite some changes in it. I will try some previous versions.
Do i have to change the firmware as well?
No, there's no particular tie between the firmware and Cura versions.
Hi,
just wanted to remind you that filament plays a major role regarding underextrusion. I printed the famous owl models with exactly the same settings - and look at the differences. The dark one looks awful while the pearl white owl is fantastic. There is underextrusion on the bronze owl. I have barely used bronze filament so the curvature on the spool should be less extreme.
This bronze filament sucks. It is ugly and prints badly. I don't like it and now I have two spools of it :/
Print settings:
OK.
As illuminarti suggested my problem could be software related so i run a full speed test with Cura 14.01 and Cura 13.12 in order to compare it with my failed print made in Cura 14.03.
And he was right.
The 14.01 at bottom and 13.12 at top looks fine except for some small problems on them.
OK so, I suppose this should be reported as a software error in Cura.
But I also notice another thing.
If i not had printed both those out in order to compare them I would not have thought that the small bumps on the surface would had been anything else than a random air bubble or whatever. But, they are in exactly the same spots on both prints so I suppose that they actually are software related as well.
So some things we think depends on filament etc can actually have a totally different source.
I have also noticed that speed can affect bumps (photo of pumpkin in post #12):
http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/?p=24010
I don't know what causes this but I think it is because when you speed up and slow down, the extruder has a lag. A delay. With the lower speeds, the head moves at a more constant velocity (less slowing down) so you get more even flow from the extruder.
This problem doesn't seem to occur on simple parts like a sphere or a cube, but on something more complicated like a pumpking or a human body it does happen.
Wallan you seem to be doing some great experiments. I would love to see what happens if you take the gcode from either of the most recent photos and load it into repetier host. It is a VERY easy program to use and it is free. Once you run it just drag and drop your gcode file onto it and see if the bumps are visible.
http://www.repetier.com/download/
This is what it looks like in Repetier.
Note: I dont know why Repetier shows infill because i had not any infill in the print.
ALSO!!
I will reprint the models tomorrow as the 13.12 not looked the same as the 14.01 in Repetier.
I think I might have printed one of the files twice (or they come out looking the same anyway).
So the comparison is not guarantied valid for a 13.12 vs 14.01 comparison.
However, as an indication of that something we might think is something random not really is.
For that it is still valid.
So, at the moment, as I dont really know witch file is witch I can really only draw one conclusion.
Both of the files of witch at least one was used for the latest print had quite a smooth surface wieved in Repetier (except for the spike in the 13.12 version)
Compared to the 14.03 witch with some zoom in shows a jaggy pattern
As I said I have to redo the test.
I will try to do that tomorrow.
I have posted a new topic in the Cura forum about the 14.03 issues.
As the problem not seem to be about under extrusion it does not belong in this tread.
After printing several jobs overnight, we got confident enough to print the stone for 'Project Egg'. The longest and highest print we made so far.
The set-up:
The settings:
After watching and grooming* the first few layers, the job was left unattended and under-extrusion ocurred.
While approaching the point where the object curved inwards, more severe under-extrusion happend. We decided to print slower and adjusted the printing speed to 35 mm/s, the first few layers went well and we went to bed.
The result next morning:
For a few centimeters, during the outward curve, the surface is consistent...after that, under-extrusion, lightening up like a city-night-view in de shade. Worse: the object is weakened along some of those lines, not quite a building block...
We are still in the learning curve, I already discarded another job for the same stone because of this issue.
Remembering the tips on the fora: thank Illuminarti for the extrusion test
The cilinders on pictures 1 to 4 are unattended, the cilinder on picture 5 has been 'groomed' into submission, the cilinder on picture 6, broke loose from the glue on the glass plate during printing.
(Allthough of little meaning in the under-extrusion series, number 6 plays a part in another issue we are experiencing: inward curving of the profile of the first view layers, which might be related to the heater bed temperature.)
The way the filament is fed into the feeder mechanism influences the flow from the nozzle of our UM2: right now it seems that with the current set-up, (except for the first view layers) the chance of getting a good, unattended print of the stone would be bigger around 5 or 6 mm3, at 235 C, heater bed temperature of 40 C, layer hight of 0,25 mm. As we have little experience with an overnight PLA print with those temperatures, I'd might prefer to go lower and slower...
This has been a great lesson to get some insight in the settings and their workings, a few more cilinders will certainly be printed, the exam will be a decent stone.
* grooming: checking and adjusting the tension of the filament into the feeder manually.
Seems like a good test would be to run with the same print speed, temp, and model, with each iteration using a different STL tolerance, to see where transitions occur between faceted shape, good quality, and bad artifacts.
The default bed temp of 75C for PLA is definitely too hot. It doesn't allow the layers to cool and the part is printed over air and pulled inward giving you that inward curve of the parts.
I don't recommend going colder than 60C as the inward movement is small. You can reduce it by printing a little more slowly or by getting the fan to come on sooner.
For unattended prints I also recommend you put the filament on the floor to reduce the friction at the entrance to the feeder:
On the video there seems to be a clunk sound right at the point of blobbing - what is that? It doesn't seem pressure related or bed/head closeness - it's more like an un-needed de-retract that makes the feeder skip back, so there's subsequent under-extrusion. I'd say it's a slicing or firmware bug, rather than something mechanical. It just looks like it tries to massively over-extrude in place, right at the start of the print.
Hey Simon,
I have re-read parts of this thread since underextrusion strikes again and I am looking for clues. The sound accompanied by the blob comes from the extruder. It skips on the first extrusion. Also the bowden tube buckles at the head end which indicates that there is way to much pressure building up very quickly.
For unattended prints I also recommend you put the filament on the floor to reduce the friction at the entrance to the feeder:
That could be an option, we already have been looking in anticipation to some solutions on YouMagine
With new settings the stone printed with almost no under-extrusion, save for one line that under-extruded partially, nearing the end of completion.
There are some issues with the surface, but right now that's another story.
Made some more tests today with 235 degrees nozzle temperature compared to 210. This time the nozzle oozed PLA before the bar was full, I didn'nt experience that earlier with the extrusion tests of 235 degrees.
The thread that was laid down to start the job was a bit blotchy. I had trouble with one job to get a good bottom layer: nothing came out at all.
Followed it up with a 210 degrees job to see if that worked normally, which it did:
I'm not sure if I'll use a temperature of 235 degrees again for this very filament.
I *think* I get it.
The first tests that were done with 235 degrees nozzle temperature had a lower heater-bed temperature then the recent ones.
I followed the heating progress of the nozzle temperature and the heater-bed temperature on de oled-screen.
If the targeted heater-bed temperature was 75 degrees, then the nozzle had already reached it's 235 degrees, while the heater-bed still was somewhere around 40 degrees.
So the PLA started to ooze. If there was nothing to force the melted plastic down, it seemed to cook and expand in volume at the tip of the nozzle.
Recommended Posts
Top Posters In This Topic
196
87
78
42
Popular Days
Jan 27
48
Jan 15
41
Jan 18
33
Jan 17
26
Top Posters In This Topic
Nicolinux 196 posts
gr5 87 posts
illuminarti 78 posts
braddock 42 posts
Popular Days
Jan 27 2014
48 posts
Jan 15 2014
41 posts
Jan 18 2014
33 posts
Jan 17 2014
26 posts
gr5 2,157
Nico I haven't seen this issue. It's kind of weird. What's the initial nozzle temp? I'm wondering if it is taking a very long time to heat the bed and the nozzle is hot for too long?
My computer with the CAD software and SD card reader is on a different floor so my workflow usually involves turning on the printer, setting build plate to 65C, then removing SD card and walking upstairs. By the time I get back to the printer the build plate is almost up to temp. So maybe it has to do with not heating up the nozzle above 180C for very long?
Link to post
Share on other sites