Gaps on the first 3 layers

[thorleif-jacobsen 0]

Hi in the middle of this image you see the third layer started.. All other is two layers. What might be the problem to thoose enormous gaps?

Also see the inner ring, it dosent even connect the bottom layer to the inner ring :/

 

[illuminarti 18]

I take it this is supposed to be printing solid? What speed and temperature are you printing at?

The inside circle barely meets the infill, which might suggest backlash issues - but then it looks like the outer wall meets the infill ok, so maybe it's not backlash, and just that the inner wall is too thin and has maybe pulled away from the intended path a little bit. If in general you're seeing that the infill isn't touching the walls, then I'd look at the belt tension - especially the short belts - loosed the motor screws, and press the motors down firmly.

It looks like your bed height is a little off, so it's starting to print a little bit too high, causing a rather weak first layer. Adjust the 4 leveling screws to make sure that the head touching the platform at z=0. (Heat the nozzle. Auto home the bed, disable steppers, and then move the head by hand to near each screw in turn. Slide a sheet of paper under the nozzle. Tighten the screw until the piece of paper slides freely without catching on the nozzle; loosen it until you just feel the head catching on the nozzle (make sure its the metal of the head that is catching, not any PLA that has leaked out); and then add an extra 1/8 turn looser on the screw. Repeat in each corner, going round until the screws don't need adjusting any more).

You also look to be under-extruding quite a lot. Measure your filament, and use that value in the slicer. Make sure you don't have any flow adjustment dialed in to the slicer or ulticontroller. But I'd be inclined to check belt tension and first layer height first. Then maybe post a picture of just the first layer, still on the bed, so we can get a better look at that to diagnose.

 

[thorleif-jacobsen 0]

At the looks at it it seems like it is taking 2 lines right next to eachother. Then a little gap. then two lines. Then gap.. then when done it goes the other direction and does the same.

 

 

I'm printing a layer now to show. Will post it asap

 

[illuminarti 18]

Aha! This is something we've only recently been diagnosing on the forum, but current thought is that two-lines-then-a-gap infill patterns is due to backlash, after all. The head doesn't move far enough as it switches from one line to the next - so it prints the two more or less on top of one another... then the next line moves it far enough for the head to finally move, and so you get a gap.

So look at the tension of your belts - especially the short belts.

 

[thorleif-jacobsen 0]

I've moved all the short belts down "hard" pressed hard on the stepper to move it down and screwd it back on. (Before that print i believe.)

Now i leveled my bed and got this result:

then comming to level 2 i got this:

At the printer i have "fR" at 100% without even knowing what that is i believe its 75mm from the setting .. so 100% = 75mm, 50% = half of 75mm e.t.c.

 

[gr5 2,172]

Here is what illuminarti is talking about:

http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/?p=14474

You have a problem with backlash (above link) but I think there are other problems.

I think you have some serious underextrusion as well in some areas (or maybe the bed wasn't level enough?).

Underextrusion in your case can be caused by many things but most likely your extruder isn't quite assembled right. Please post a photo of your extruder. There is a slot in it - the part that goes into that vertical slot is supposed to slide all the way to the bottom of the slot which compresses the spring properly. With it compressed properly the spring should be 11 to 11.5mm long.

Another way to test your extruder is that it can pull 22 pounds. So if there is no load (filament not reaching the print head yet) you can pull easily 10 pounds force and the extruder should still pull filament no problem.

 

[thorleif-jacobsen 0]

I think the extruder is perfectly. I ran a calibration on it, and when i ask for 100mm it presses out exactly 100mm (just in the air), i dont know if this is a way to see it works good or not?

The spring is not 11.5. The measurements from the black edge to the back of the hex screw head is 15.5mm should i tighten this up a looot ?

 

[illuminarti 18]

Layer 1 looks pretty good now... although the first layer tends to squish out a bit, and that can mask under-extrusion.

The second layer again looks like it's under-extruding. As George said, make sure the filament is properly locked in the extruder (and the bolt on the lever is all the way to the bottom of the slot). Try tightening the screw to shorten the spring a little bit - give it a few turns, and see if the print gets better.

Extruding in air is one thing - once you're printing, and pushing filament through the nozzle, then you get back pressure, and the extruder can start to slip. You're probably printing the first layer slower than the second and subsequent ones, so that might explain some of why it prints better now - slower = less pressure to overcome in the head.

Yes, the 'fr' on the Ulticontroller is the linear head speed, as a percentage of whatever you set in Cura.

 

[gr5 2,172]

Default settings in Cura have the first layer much slower which means there is much less plastic going through the nozzle which means much less pressure/force on the filament. 2nd layer needs more force. Again, try the 10 pound test. Tighten the spring one full turn. Send us a picture of your extruder it might be built wrong and....

You have another problem for sure now - backlash. You need to tighten your belts. Like illuminarti said, it's most likely the short belts which are very easy to tighten - just loosen the 4 bolts, push down on motor hard and retighten.

 

[thorleif-jacobsen 0]

Some images, i tighten the screw 3 turns. and tried to manually turn the gear and it was more "pull" on it now. The steppermotor belts are as tight as it can be. I have the screws on the lowest part at this moment. I'll have to check a smaller 2 layer print later. Is there any test "thingiverse" links for not so big print to test layer 1,2,3,4 ?

 

[gr5 2,172]

You can tell it to print faster on the first layer if you want. But then you might not come out as good - it's usually best to print slow on the first layer. The reason it takes a few layers to mess up is the printing speed is faster and faster as you go up layers until it hits nominal speed around layer 3 or 4.

The ultimaker robot is a good model for this test. You can find it in the cura folders:

...\Cura_13.06.4\Cura\resources\example\UltimakerRobot_support.stl

 

it was more "pull" on it now.

 

Did you try the 22 pound test? How much more pull? Another test is to look at the filament after it has gone through the feeder. The filament should have a clear pattern on it (difficult to see with black filament). Like this:

http://i.imgur.com/2MeDA.jpg

You probably just need to tighten the screw some more until the spring is compressed to 11.5mm.

 

[thorleif-jacobsen 0]

I did compress the screw to 11.5mm now (the spring). I did a 6 hour print now, that round circle there.. it started lifting the print up on each sides. So only the middle circle was touching? Why is that?

It pulls a lot, i dont have exactly 22 pound to test it but it cant be that far off, it drags my finger when i hold it and i dont feel like holding it harder.

This is what i print:

http://link.jepb.no/QGTR

this is what i saw at home:

http://link.jepb.no/QGib

and

http://link.jepb.no/QHtR

This is the settings:

http://link.jepb.no/QHmd

:-/ boring 6 hour waste of print And the motors (smaller belts) cannot be tighten anymore downwards since the screws are at the bottom now just so you know it. Still got the 2 lines, gap, 2 lines gap problem. :-(

 

[illuminarti 18]

The first two links didn't work for me.

The sides are lifting up due to 'warping' - cooling stresses in the plastic causing the edges to pull off the tape.

The best way to prevent that is to wipe the blue tape down with isopropyl alcohol just before printing. That should keep the print stuck so well that you will probably tear the tape removing the print. Then you can soak the bottom of the print in more alcohol for a few mins, and the tape will slide right off.

Do the short belts seem to be tight now? You should get a note when you pluck them, and there should seem to be very taut.

What about your long belts? Are those reasonably tight?

 

[gr5 2,172]

The ultimaker can be frustrating. There are lots of things for you to learn still but you got past at least one key hurdles (extrusion working now).

 

6 hours is a bit long for that print. PLA is very strong. I would lower your infill to 20%. That will speed things up drastically.

 

Also I'm not sure if you really need .1mm layers. How perfect does it need to look?

 

Let's talk about printing speed for a minute:

 

Right now you have 50mm/sec 210C .1mm layer.

 

The constraints are volume of plastic versus temperature.

At 210C I can print about 100mm/sec .2mm layer (or 200mm/sec .1mm layer).

At 240C about 150mm/sec .2mm layer.

 

These are pushing it. To be safe you should be able to print at 75% of these values if your feeder is working well now. So you can either up your speed to 100 or 150mm/sec (which will speed things up) or you can increase your layer thickness (preferred!) to .2mm and your speed up to maybe 75mm/sec.

 

If you have an ulticontroller like I do consider settings your speed always to 100mm/sec. Then you can use the ulticontroller to set the "feed rate" percent. If you set percent to 75 then you get 75mm/sec (100mm/sec X 75% =75mm/sec) and so on. You can dial around different speeds and see how it comes out. I recommend printing a test cube say 2cmX2cmX2cm, trying one speed/temp combination and then printing faster and faster until you get underextrusion (holes in the sides).

 

These two changes (switching to .2mm and 80% infill to 15% infill) should speed up your print by at least 4X making it not much more than an hour to print.

 

[gr5 2,172]

As far as lifting/warping - this is a problem on large prints like this. Easily fixed. Here's my spiel about this:

Curling parts

1) First wipe your blue tape with isopropyl alcohol (also called rubbing alcohol). You can buy this at any store that sells bandages such as a supermarket or drugstore. This removes the wax from the blue side of the tape and increases stickyness. Also if the blue tape itself lifts off the bed it helps to get larger pieces of tape available at any store that sells paint and paint brushes. Ask for "blue painters tape". This one thing will help you more than any other thing below.

2) Add a brim. Cura has a brim setting on the adheasion section. Do about 10 passes. This will help keep the corners of your part down.

3) Print first layer slower and hotter. I always set my temperature to 0 in cura and adjust it by hand using either cura print dialog or ulticontroller. For large parts I print the first layer at 240C and then lower the temperature after it is done with the first layer. Also squish that PLA into your bed nice and hard with the bed nice and high.

4) Don't infill the first layer. This might not be practical depending on what you are printing but cura allows you to not fill the bottom or top layer - it's a checkbox.

Of all the advice, isopropyl alcohol is probably all you need.

 

[gr5 2,172]

Everyone with an ultimaker has gone through your frustration but you are getting there. Don't give up.

 

[gr5 2,172]

Still got the 2 lines, gap, 2 lines gap problem.

 

Probably the long belts then. Here is a video that explains how to test and tighten the long belts. It's good because it uses pitch to test which is pretty accurate:

 

Having your motors all the way down the slots is probably too tight.

 

[illuminarti 18]

Not necessarily... When I installed Jeremie's short-belt tensioning brackets after I'd had my printer for just a few months last summer, I actually had to file the slots slightly longer to get sufficient tension on my short belts to eliminate backlash.

 

Having your motors all the way down the slots is probably too tight.

 

[thorleif-jacobsen 0]

i think i had the bed too high since it kinda "squeshed" if thats a word the pla flat then two "borders" on each side. I took it so i could put a paper under it and move it so i felt it "skratched" the paper (ordinary a4 paper).

Here is the sound from the belts: https://www.dropbox.com/s/0egwi0k93nkk4js/Video%2016.07.13%2020%2007%2015.mov?m

Here are some prints:

https://dl-web.dropbox.com/get/Camera%20Uploads/Photo%2016.07.13%2021%2007%2050.jpg?w=AACz2WjG2h4cuUaWj-rfot3CQZlp0nKdzYyYZyw9GfVhdQ

https://dl-web.dropbox.com/get/Camera%20Uploads/Photo%2016.07.13%2020%2035%2038.jpg?w=AABw0gfYFk1u_LG7Qrlg-906rKiVJQDgM3f5DK0r0Ar2vA

 

[illuminarti 18]

The belts look pretty tight. Don't think that's your problem at this point.

What is your steps-per-e value?

 

[illuminarti 18]

Oh, the last two drop box links in your previous post give me errors, btw.

 

[gr5 2,172]

>i think i had the bed too high since it kinda "squeshed" if thats a word the pla flat

Squishing flat is good. It's probably okay. If your first layer is .3mm then it should be almost square. If your first layer is .2 or .1mm then it should be squished. You can see the first layer height on the ulticontroller if you have one or look at the cura settings "first layer height".

 

Bed height is very critical. It's often the most common problem. You have to relevel every few hours or everytime the temperature changes and so on. After many months the bed doesn't need adjusting every few hours - but still - it's a common annoyance (always having to level the bed).

 

>Here are some prints:

Can't see the prints. Sorry.

 

>Here is the sound from the belts

Wow. Tight. Plenty tight I think.

 

If you still have backlash/play it can be from somewhere else maybe. It could be something loose in the print head. If you grab the nozzle and push it one way, then the other, does it move without the steppers moving? Turn the stepper motors on and push the head around. It shouldn't move. You can do this with UC or with Cura. With either, just move the X and Y axis and the steppers will stay on for a little while.

 

Maybe you don't have backlash (aka play) anymore?

 

[thorleif-jacobsen 0]

The prints which dident work:

 

1 layer just started the second layer. : http://link.jepb.no/QHmg

1 layer squished: http://link.jepb.no/QGzd

 

[gr5 2,172]

Squished looks good to me. But it's blurry - hard to say.

The first image - um - something looks wrong. It looks like both backlash and underextrusion. Still. It's hard to tell. Maybe you should stick with the squished print and do the second layer over *that*.

When you did you belt tension testing - I'm not sure if you did all 3 sections. Because there are 2 belts on each axis and because all the pulley's are locked the tension can be unequal in the 3 sections of the pulleys. Make sure they are all similar (don't have to be identical.

Also I'm wondering if you have a loose pulley at this point. That might explain the backlash. There are 6 pulleys per axis and if only one of the long belt pulleys is loose it's not obvious. Tighten the heck out of the 8 pulley set-screws on the long belts. Make sure things seem perpendicular (move the blocks around a bit until as centered as possible) before tightening. Consider marking the pulleys and the rods with a sharpie (permanent marker), then push the head around tiny amounts and see if one of the 8 pulleys is slipping.

 

[gr5 2,172]

But your warped print looked pretty good other than the warping. What is different? And did you put alcohol on the blue tape? The alcohol should help PLA stick to blue tape. Print a skirt or brim and try to lift with your fingernail. It should be very difficult. If it isn't being squished into the tape, stop the print, raise the bed and try again. Otherwise it will lift and warp.