Looks as if it's not just the very top layer but also the second last layer. Is that correct?
Have you yet found out if it's not enough filament or rather to much volume for the correct amount of filament, i.e. a larger z step or similar than there should be?
I guess the spacing of the lines is correct, isn't it?
The space between the lines increase with z-height. The first two layers are tight, then the lines start to drift apart. I specified the correct filament diameter but other there's nothing else to fiddle with. Well there is, but I don't like playing with the flow parameter.
I agree, leave the flow at 100%
Would you say the gaps grow linearly and it gets worse and worse and worse... (would it completely stop if you print higher objects?)?
Or does it get worse for the first - uhm - five or so layers and does it stabilize after that height?
Not sure. I'll have to babysit the print to see. But fromy my observations so far, the first two layers are perfect, then the strands start to drift apart and on the sixt layer (1.2 top/bottom thickness set in Cura) they are about 0.2mm apart and stay like this.
I was too quick to judge this effect. Apparently if the flat area is big enough, the problem will start to sho from the second layer on.
And I thought I'd be safe if I would print curved objects, but that's not the case :sad:
I have not followed everything, but your second printer has printed sometime normal? Or you have this problem right from the beginning?
Markus
I fear your alarm bells were not ringing for nothing. :(
It looks rather like a classic underextrusion problem than like a problem with not-correct z-shifting as it occurs between layer 2 and 6. It's exactly what I know from my UM1 when using settings which result in an underextrusion. It's not happening on the first layer as due to thermal expansion of the hotend the gap is actually a bit smaller than the amount of filament is calculated for.
Before you despair of having an underextrusion issue again also with the second machine, make sure you checked all the simple reasons like a filament diameter entered thicker as it is (btw: how do you determine the diameter?). 
On the UM1 there is a simple test to check if the feeder is behaving as it should. It's the test where you dismount the Bowden at the feeder's end and measure how much filament it is transporting actually. Maybe it's worth doing this test also on an UM2? If this test is ok, then you might have to check how well the filament slides inside the Bowden tube (maybe you have to dismount the feeder for that test?).
I don't know if it is an option for you to exchange the feeder. From what I read so far in the forum, Robert's design works quite flawlessly. 
@Markus: Yes, this issue appeared over time.
@Dim3nsioneer: But if it were underextrusion wouldn't I also have it on the outer shell? Like this here:
I measure the filament diameter in different spots (unspooling a few metres) and using the average. I also did the extruder calibration once (with the first UM2) and it was pretty good.
By the way, the cylinder test was printed with colorFabb at 220°. New material, new headaches
To my personal experience underextrusion is first seen in the non-solid-infill, then in the solid infill and finally in the shells.
Some time ago I determined the filament diameter exactly the same way as you do. Then I had some underextrusion effects and began to take the smallest diameter I measure in ten measurements. So far this works very well without having overextrusion. Might be worth trying for you?
I hope you get better results with the Colorfabb PLA/PHA than me... :???: I decided to stay with the pure stuff (only PLA)...
Good idea about the measurement method - I'll try it. But I bet this ain't the issue here. Even with odd filament diameter settings, these lines are too far apart to be "just" underextrusion (although I'd be happy if it were like this).
Aha, I knew something was different with colorFabb filament (didn't buy it myself, just swapped filament with another user here on the forums). I smells different and is a bit softer. Not a bad thing and the color is killer!
:rolleyes: Hm, just without thinking wrote:
A problem with the X / Y positioning, it may actually not be, because the strands are too narrow in both directions.
Have you ever tried the machine to give a much lower material diameter than actually measured?
So you could, the material at the end but still use.
Markus
Hey Markus,
I don't really understand what you mean about lowering the material diameter. What good could it do?
Hi Nico
Just another attempt:
Cheating the machine, and write in the machine settings a less material-diameter. Perhaps you set 0:10 to 0:25 mm less than average real measured.
Let's see what happens.
Bye for now
Markus
That's actually the very same as increasing flow above 100%...
I discovered this morning I accidentally had my nozzle size set to 0.6, I'm guessing I thought i as setting the layer height to 0.06... I'm a bit daft.
But, anyway, thought I'd mention it in case you've done something similar.
Sadly not, but thanks for the input.
When I first experimented much with my alternative stepper, I knew nothing of xx -e steps per 1mm filament. I have therefore tried directly on the machine to adjust the flow of materials to obtain similar results as with the standard engine. Some time later informed illuminarti me the following: My alternative Stepper makes compared to the original only half the number of steps. Consequently, the material flow 50 percent should be suitable in order to compensate for this fact. Well, with 65 percent of material flow I got similar results.
But with increasing printing speed I always had to further increase the flow of material manually, otherwise I got different results with insufficient amount of material. That's what I can see very clearly with the increasing speeds during the reprinting of the test cylinder.
It will probably be so, but I find it very hard to believe that a percentage increase in the flow of material is similar to a corresponding change in material diameter.
Maybe someone could tell me the thing once a little closer.
Markus
Marlin takes the volumetric information from the UltiGCode and divides it by the cross-section of the filament which is derived from the filament diameter. It then computes the necessary amount of steps for the extrusion length.
So the diameter is just a scaling factor as the flow percentage.
Hi guys,
new UM2 same problem :( :( :(
Those filament strands for flat areas gradually drift apart with increasing z-height:
W T F ?
send me your gcode, I will run it too.
Can you compare four layers on which the speed is the same? It's not the case for 1-4... so the effect here might be... uhmmm.... underextrusion...
ok... this is now a maybe completely stupid question, but have you ever tried to run one of your UM2s somewhere else than its usual location?
@Sander: Sure, here it is:
@Dim3nsioneer: We do not say the u-word around here... or else you can hear me scream right up to Switzerland
Except for the first layer, the layer 2-4 should be the same.
Regarding the printer location, I did run the UM2 in the living room too. Not now, but with the former ones and I don't remember the issue to have changed then.
However, when I first encountered this issue a few months ago, I remember that it started gradually. In the beginning it was not there. With this UM2, it was right from the begining.
Except for the first layer, the layer 2-4 should be the same.
Are you sure? I thought to remember the speed increasing gradually for the first about five layers...
And I appologize for the U-word... :oops:
ArunC posted a topic in UltiMaker Cura,
.thumb.jpeg.0b7a05eafc09add17b8338efde5852e9.jpeg)
Dustin posted a topic in Firmware,
Recommended Posts
Top Posters In This Topic
135
46
41
31
Popular Days
Nov 20
35
Nov 18
21
Oct 2
16
Oct 12
15
Top Posters In This Topic
Nicolinux 135 posts
gr5 46 posts
visu-al 41 posts
mevander 31 posts
Popular Days
Nov 20 2014
35 posts
Nov 18 2014
21 posts
Oct 2 2014
16 posts
Oct 12 2015
15 posts
Posted Images
Nicolinux 288
Top
Link to post
Share on other sites