Closed TOP surfaces

[esteban-pacheco 33]

Thanks Gr5 for your extense answer:

I have an Olson Block now (but i had also this issues since I've got the machine in the very first week i've noticed this - then i had the  normal nozzle)

1) my  heater block i not touching the fan shroud, theres 1 mm gap in between (i have steel gauges here)

2) isolator is 1 week old

3) it doesn't matter if the spool in new or not, i have the same issues.

4) Never Tried That

5) ill try that

5a) ill try that

5b) the gap between the PTFE and the steel isolator is 1.4 mm

6) Nozzle are new (olsonblock)

7) I was using the roberts feeder, it was (in my opinion) not so reliable... now I'm using the standard one, so, no matter if  tight or loose

8) which nuts?

9) filament from Colorfabb (i think is a good manufacturer).

Edited July 9, 2015 by Guest

[neotko 1,417]

Ok just another wild idea. Check the extruder and make sure it's thight and clean it with a wire brush. I mean this metal part.

[gr5 2,167]

8) which nuts?

9) filament from Colorfabb (i think is a good manufacturer).

 

Colorfabb is very good manufacturer.

If you have problem #8, which is rare, you can tell because the filament spirals through the feeder and there is a spiral pattern on the filament in the bowden tube. also if you look at the knurled feeder shaft it is tilted. To actually see the nut in question you have to remove the feeder motor and look underneath it. It is similar to having a pebble under an iron causing it to tilt a bit. I doubt this is your problem but check everything.

[gr5 2,167]

So back in January Ultimaker shipped me a UM2go and a UM extended to test out. At that point I had 4 ultimakers. The extended (only) had this problem - top surfaces would not fill. It was frustrating and took me 2 days to get it working nicely. I have since shipped the printer back because they wanted it back.

My notes are old but I printed a small test cube many times with .8mm wall and the walls were not making good contact such that I could put my fingernail between the 2 shells and bend the outer wall outward. It looked like your photos above. I had forgetten about this until today.

Remember - this was a completely brand new printer. Anyway I fixed it by doing a few things. I ended up printing a bit hotter, slower, moving filament to floor and increasing flow to 110%. My notes say my final print was 35mm/sec 220C with 50C bed (although I don't think the bed temp matters at all). I did .2mm layers (should be able to print twice as fast with .1mm layers).

I think putting the filament on the floor was important but I don't know.

Certainly the most important step was increasing the flow. I was never able to get it to look right without doing that. I have 4 Ultimakers now (but that one is gone) and I *never* have to increase the flow. Maybe I should have checked steps/mm. I probably did. I don't remember.

The first step to solving your problem is creating a very small test case that only takes a few mintues to print. If it takes more than 30 minutes to print to see the problem then you will take days and days (weeks?) to fix this. Not one day like myself. I was able to change settings and print again and again and mark each print with a permanent marker and keep good notes. I was so sure increasing the temperature was all I needed to do but that only helped so much. No matter how slow I printed it seemed to make no difference beyond a certain point.

Now I wish I still had that printer so I could do more tests.

[esteban-pacheco 33]

well, Gr5, i also came to the conclusion that increasing the flow to 110% is a part of the solution (despite what i've readed here in forum - that turning the flow up is like complete heresy).

maybe i can turn dow the filament thickness (say 2.84) on the machine.. that will generate a bigger flow without increasing the flow

whats that Steps / mm?

[gr5 2,167]

lowering thickness to 2.84 is no different than changing flow. They are just 2 ways to do the same thing. I would only modify filament thickness if the filament thickness actually changes.

steps/mm? Well all 4 steppers (x,y,z,e) have a parameter called steps/mm. You are supposed to set that to the correct calibrated value for your machine. It should be the same value for ALL ultimakers. It's possible to set that value purely through gcodes. For example if you connect to your printer through a USB and launch pronterface, before you do anything pronterface tells you a ton of things about your machine including PID settings and steps/mm for all axes. Acceleration settings, and much more. You might be able to change this from the control panel as well. I guess it's possible someone messed with this value by accident. Or through gcodes. I mean simply printing a gcode file could change the value but not one created by cura - it would be a custom gcode file meant to mess with your calibration.

[esteban-pacheco 33]

steps/mm?  Well all 4 steppers (x,y,z,e) have a parameter called steps/mm.  You are supposed to set that to the correct calibrated value for your machine.  It should be the same value for ALL ultimakers.  It's possible to set that value purely through gcodes.  For example if you connect to your printer through a USB and launch pronterface, before you do anything pronterface tells you a ton of things about your machine including PID settings and steps/mm for all axes.  Acceleration settings, and much more.  You might be able to change this from the control panel as well.  I guess it's possible someone messed with this value by accident.  Or through gcodes.  I mean simply printing a gcode file could change the value but not one created by cura - it would be a custom gcode file meant to mess with your calibration.

 

WELL, NOW THATS INTERESTING! Thanks GR5 I'm gonna check the Steps/mm settings of my printer. Where can i Change that? Control Panel? you mean the ulticontroller? maybe under movement right?

what are the normal Range or values?

Ps, i checked the screw under the feeder motor, and all is ok... no tilting....

Edited July 10, 2015 by Guest

[danilius 118]

I'm kicking myself for having forgotten that one!

I regularly change the flow % during a print. It all depends on how lazy I am to measure the filament. I also change the diameter of the filament to suit if I am a little less lazy :-)

Sometimes I have a no-name brand that I have no idea what it can do, and I print a test part and fiddle with the flow % to get nice fill.

So, play away with the flow % or the filament diameter. There is nothing special about them.

A good tip as well is to sit by your print and watch what happens. So, print a small test cube and watch what happens.

I would also suggest leaving the steps/mm alone for now. First check that adjusting flow % or reducing filament diameter solves your problem. Messing with the config could result in more pain that you want!

[gr5 2,167]

@esteban-pacheco - I can't tell you how to modify steps/mm unless you tell me what printer you have.  It doesn't say in the first post.  It doesn't say in your profile settings.  And I don't want to go back through 6 pages to find out. Please update your profile settings.

Edited July 10, 2015 by Guest

[esteban-pacheco 33]

@esteban-pacheco - I can't tell you how to modify steps/mm unless you tell me what printer you have.  It doesn't say in the first post.  It doesn't say in your profile settings.  And I don't want to go back through 6 pages to find out.  Please update your profile settings.

 

O i never changed that... Profile is already updated!

Ultimaker 2

[gr5 2,167]

On UMO go to CONTROL MOTION TUNE and E steps/mm should be about 836. Ideally you want to put filament in the extruder, mark it, tell it to extrude 100mm of filament then measure how far it moved. If it moved only 90mm then increase steps/mm by 10%. Don't mess with these values normally. Particularly don't touch the x,y,z steps/mm.

If you change this value in the TUNE menu and then turn off your printer you will lose your setting. You have to save it to eeprom.

[gr5 2,167]

If you have a UMO then it changes many of my answers above! In particular you should check the spring on the feeder - you want it compressed until it is about 10-11mm when it is clamped on the filament.

[gr5 2,167]

On UM2 to update steps/mm you will need to get pronterface firmware here:

http://koti.kapsi.fi/~kliment/printrun/

Connect a USB cable from computer to printer.  Get pronterface to connect.  Once it connects it will tell you the current values for many settings including PID settings and the steps/mm for all 4 axes.

You can change the steps/mm using gcodes but first let us know what value you have currently.  Test by removing filament (just pull it half way out with head at 180C but power off - you shouldn't have to pull very hard).  With filament half way out, mark the filament with a pen.  Then using pronterface tell extruder to move 100mm.  Then measure movement with a ruler.  If it moves 5% less than 100mm then increase steps/mm by 5% and so on.

gcodes explained here:

http://reprap.org/wiki/G-code

M500 stores current values to the eeprom so you don't lose when you power off

M501 displays current settings (also pronterface does this automatically when it connects)

To set the steps/mm for E axis (extruder) do this:

M92 E836

Where 836 in above example is steps/mm.  After changing this value test again and then save with M500 command.

Edited July 10, 2015 by Guest

[gr5 2,167]

I would be ABSOLUTELY AMAZED if your steps/mm for the E axis is the wrong value. If that were the case I would have to tell many people about this.

[gr5 2,167]

steps/mm for UM2 family extruder axis is supposed to be 282.

[gr5 2,167]

Oh one more thing - before you can move the extruder you need to enable "cold extrudes" because the E axis won't move unless the nozzle is > 170C.

From pronterface you can enable cold extrudes with the gcode:

M302

To disable cold extrudes just turn your printer off and on again.

[esteban-pacheco 33]

Printer is now online.

echo:Marlin 1.0.0

echo: Last Updated: Apr 16 2015 12:22:44 | Author: Version:_15.04

Compiled: Apr 16 2015

echo: Free Memory: 1961  PlannerBufferBytes: 1232

echo:Stored settings retrieved

echo:Steps per unit:

echo:  M92 X80.00 Y80.00 Z200.00 E282.00

echo:Maximum feedrates (mm/s):

echo:  M203 X300.00 Y300.00 Z40.00 E45.00

echo:Maximum Acceleration (mm/s2):

echo:  M201 X9000 Y9000 Z100 E10000

echo:Acceleration: S=acceleration, T=retract acceleration

echo:  M204 S3000.00 T3000.00

echo:Advanced variables: S=Min feedrate (mm/s), T=Min travel feedrate (mm/s), B=minimum segment time (ms), X=maximum XY jerk (mm/s),  Z=maximum Z jerk (mm/s),  E=maximum E jerk (mm/s)

echo:  M205 S0.00 T0.00 B20000 X20.00 Z0.40 E5.00

echo:Home offset (mm):

echo:  M206 X0.00 Y0.00 Z-19.15

echo:PID settings:

echo:   M301 P10.00 I2.50 D100.00

echo:SD card ok

[gr5 2,167]

So it is in fact 282 steps/mm. That's the nominal value - the value in 99% of UM2s out there. Did you test how far filament moves when it's not in the head and you move it 100mm?

[esteban-pacheco 33]

Gr5, yes i tested it, it moved 100mm

(i took the bowden out then maker the filament with a pen exactly where the bowdentube ends, the in pronterface used the gcode 302 and move it 100 mm) all was perfect...

so now that we now that the e axis value is ok, and we did the 100mm move test...

why I'm having this problems with the top layers?

Edited July 11, 2015 by Guest

[gr5 2,167]

I don't know. For now you might want to just increase flow to 110%, increase temp to 225C.

[esteban-pacheco 33]

well, that was was the solution I've found, is not counterproductive?

[gr5 2,167]

As long as it keeps working I think you will print at 110% for thousands of prints with this machine - but if you ever get another UM2 I think you will find it prints fine at 100%. It's a mystery and it bugs me. I'm very curious as to the cause. When someone figures this out I will post here but it may be months or years before someone figures this out. You aren't the only one with this problem but not enough people have the problem that someone smart figured it out.

[francfalco 37]

Finally got around to testing the .stl.

I cropped into one corner to save time and tested 'quick print' Fast (10% infill)

Then tested 20 & 25% infill and tweaked the infill speed to match layer speed (all printed at 50mm/s at 210C PLA from Faberdashery) [/media][/media]

25% infill was best, but even 10% was acceptable but did suffer in a couple of corners to not quite cover the surface.

So model is good and Cura does a pretty good job on my UM2, so your covering/under-extruding issue needs more investigation, but it can be achieved?

Edited July 11, 2015 by Guest

[esteban-pacheco 33]

can you send me the Gcode Franc Falco?

[gr5 2,167]

francfalco is smart - he made the part smaller for testing purposes. I would go even smaller. A 10mm by 10mm by 5mm tall part.