Gaps on the first 3 layers

[thorleif-jacobsen 0]

The warped thing looks really good. But there is on each layer a "play" (2, 1, 2, 1) its just covered by the layers.. So the problem is there. I tried to tilt the whole extrudion head from left to right. It does move a little when i wiggle it left to right. The bearings is the one "moving" looks like the bearings on the left and right side allows me a little play moving the heated top some about 1 mm or so.. I see that the inner ring of the bearing also is moving a little upwards as i force it and wiggle it.

3 questions:

1. Can that be the case? bad bearings? It's under a week old. Its been like this from the start. (Its only the left to right bearing. So that wont explain why the backlash goes both ways(?)

2. Printing a big thing with 20% infill makes a grid inside it. If i make a 10x10cm box with 20% infill it looks perfect until it tries to fill "in air".. Ofc the PLA starts falling down and making a ugly sight.. That will always be an uneven surface? How can i fix that?

3. This is not a question but a big thank you for all the help and tips. Alcohol i dont have in my house I'll try to buy tomorrow. Currently I'm pringint a gridded one (no bottom layer). Dont need the bottom layer for this

Its at 10% after 40 minutes. Lot shorter time than 3% at 6 hours haha.

 

[illuminarti 18]

Are you talking about printing solid layers over infill... e.g., the top of a cube? That's going to work better if you print slower, and make sure your fan is on. It does take several layers to fill it in smoothly. So set the top thickness to be 6 to 10 times your layer height.

(See: http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/?p=17300 for an example of the effect of cooling on top surface quality)

 

[thorleif-jacobsen 0]

Yeaah, i had 0.4 on bottom and top layer. I had removed the "bottom" layer from expert settings and i got this:

http://link.jepb.no/QHVx

So what you say, 0.4 should actually be more like 2mm thickness.. since i do 0.2layer height that is around 10 layers extra? But how can it make it "smooth" and level when there are things falling?

Also after tightning the spring to 11.5 the plastic wheel inside pressing on the PLA is making a "unoiled door hinge sound". Is that normal? Sounds like it's gonna break the thing..

Edit: Fan is at 100% from layer 2 and up anyways

EDIT 2: The spring was tighten to 10.5 instead.. 11.5 silent <3 little squeeking but not so irritating.

 

[illuminarti 18]

No, it shouldn't be squeaking - make sure that the wheel rotates freely when the extruder is open.

With a 0.2mm layer height, you might get away with fewer layers. Try a setting of 1mm, and judge it from there. You'll definitely need at least 5 layers, I think - maybe as many as ten. It depends on the speed, and temperature, and the infill percentage - how far apart the lines of infill are. Covering over the square pattern infill in the new Cura seems to be harder than I had before using hex infill patterns. The first layers will be pretty poor, but each layer closes the gaps up a bit, so that after a few layers you get a smooth surface.

 

[illuminarti 18]

Btw - I asked before but I think it got lost in the chatter. What is your steps-per-e value? Also, when you print things of a known size, do the x and y dimensions come out correct? (E.g., if you print a 100mm square - does it come out 100mm on a side?)

 

[gr5 2,172]

What is your steps-per-e value?

 

Look in preferences to get the answer.

 

(Its only the left to right bearing. So that wont explain why the backlash goes both ways(?)

 

That's probably your problem. See second photo here which shows pattern if only X bearing slipping:

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

 

Fan is at 100% from layer 2 and up any

 

Make sure it isn't connected backwards, make sure you feel air coming out.

 

PLA starts falling down and making a ugly sight.. That will always be an uneven surface? How can i fix that?

 

Each layer will get better. That's why you need 4 or more "top" layers. Not just 2 layers.

 

[thorleif-jacobsen 0]

Steps Pr E i dont know i calibrated them a little. I printet a 2x2x2 cube. The measurement is:

Z = 22mm (Assuming Up and down)

X = 19.7mm (Assuming left to right)

Y =19.7mm (Assuming front to back)

I saw how the layer thing worked, also setting the speed down to 20mm it helped a lot only on the two first layers. Perfect

So i will add a 3-5 layer on each "importent" project.

Getting there!!

The images you sent link to the backlash seems more like a Y backslash since what i got is 2 perfect layers. Then a spacing, then 2 perfect layers.

On this print you see the bottom is kinda seperated from the "walls" i kinda managed to rip off the walls and the internal stood there. Its not "hard" to split the wall and the content. Is this normal?

bottom

https://www.dropbox.com/s/lhnl6mpsdtg9tue/Photo%2019.07.13%2016%2027%2041.jpg

top

https://www.dropbox.com/s/imbjbv1pkpkd3ae/Photo%2019.07.13%2016%2026%2054.jpg

as you see here i slowed it down to 25% on the tp layer then i speeded it up to 100% to finish it off just to see the difference..

 

[gr5 2,172]

The images you sent link to the backlash seems more like a Y backslash since what i got is 2 perfect layers. Then a spacing, then 2 perfect layers.

 

My images assume a tall rectangle. You are printing a square. It matters if the infill stops at a vertical or horizontal wall. With a square it stops at a vertical wall on one side and a horizontal on the other. My images didn't consider this and if they did you get a different but similar pattern.

If you felt the x axis print head was loose then that's probably the problem.

 

(Its only the left to right bearing. So that wont explain why the backlash goes both ways(?)

 

It's complicated to explain. My diagrams help but again my diagrams only show one case, they don't show infill for a square so I still think this is the problem.

 

Its not "hard" to split the wall and the content. Is this normal?

 

Not really, no. Again this is your looseness in the X axis. You will need to fix this.

Are you sure it's a bearing? Is it loose in the wood? Or defective? You can get a new one from ultimaker if you email support.

 

[thorleif-jacobsen 0]

Hi i saw a error with the bearings, i lifted the left side and the axis iron rod which goes into that bearing is "moving" inside the bearing. This happends on FRONT and LEFT. RIGHT and BACK are steady as rocks. Cant even make it jiggle a bit.. So looks like its the bearings on both FRONT and LEFT. Can that explain it? As it hits the front left wall in speed it'll have a little free play going back that way. That might create the space and on the way back its staight as an arrow.

I dont know if support will send me a new bearing since i already got a missing part sendt. Its proabably expensive for them to send new parts. :( I'll ask anyway right now. ASk them to check this thread

Edit: I did the test by putting my finger under the top plate of the extruder on each sides and lift. Both front and left lifted a little. (moving the extruder hot end to 1mm from origin) the other side no movement)

 

[gr5 2,172]

Bearings like this are very cheap. I would just order a few on the internet. You only need 3 dimensions: inner diameter, outer diameter, width. In mm.

 

[thorleif-jacobsen 0]

Anyone got a ebay link? They seem a little special :-/ Looks like its both bearings who are off then since its only two for both axis..

And i dont really want to dismantle the whole thing right now :-P Finally printing a little "ok"

http://wiki.ultimaker.com/Bearing_Pack

 

 

[thorleif-jacobsen 0]

Also, i got it calibrated on x and y using theese e-steps on M92

X79.57578142754516 Y79.57578142754516

Giving me a 20.03 mm measurement now

How can i calibrate the Y axis? I'ts a little to high. 2.3mm right now.

 

[gr5 2,172]

Anyone got a ebay link? They seem a little special :-/ Looks like its both bearings who are off then since its only two for both axis..

 

Oh. Those bearings. I forgot what they looked like. Those seem a little more specialized. No idea, sorry.

 

[illuminarti 18]

Be careful calibrating the x, y and z... you are going to get some amount of shrinkage with your prints, and that is probably going to be proportional to the size of the object to some extend, but also depend on the geometry. Calibrating to particular models may cause problems with other models, so you should test your settings on a variety of sizes and shapes.

Bear in mind also that the x, y and especially Z settings are a function of the mechanics of the printer. While the steps-per-e setting for the extruder has a little more room for variability, due to filament softness, spring pressure etc, the X & Y have less room for variability, and are mostly a function of the pulley diameters.

The Z stage steps are pretty much a fixed quantity though, because the motor directly drives the z-screw. There are 200 full steps in a full rotation of the motor, and it is set to 8x microstepping, so that's 1600 (micro-)steps in a full rotation. The pitch of the z-screw is 3mm, so the steps per z should be 1600/3 = 533.3333.

When checking your z-height calibration, bear in mind that if you set a first layer height that is larger than the normal layer, then the overall height of your object is going to end up taller by the amount of that extra layer height. (The first layer is sliced as if it was the normal height, and then just printed thicker than usual).

 

[thorleif-jacobsen 0]

So basiccly you can never get a 100% accurate print on any size on this?

 

[illuminarti 18]

Well, I generally find on objects on the order of 100mm, my dimensions are within a tenth of a mm or so... which is good enough for my purposes. But it's the nature of thermoplastics in general that they are going to shrink as they cool, so you need to make allowances for that. Even injection molding production has to take that into account.

With FFF printers like the Ultimaker, the biggest dimensional challenges seem to be with holes in things. Because of both shrinkage, and the fact that circular paths are approximated by multi-side polygons (that are typically going to be smaller than the equivalent circle), holes are particularly difficult to get right, and typically need to be designed larger than the intended size.

 

[gr5 2,172]

So basiccly you can never get a 100% accurate print on any size on this?

 

You can get damn close. You can get to about .1mm if you adjust your CAD parts. Plastic shrinks. You have the same problem with other manufacturing techniques - if you have critical dimensions you sometimes have to do a test print first to figure out how far off your critical dimensions are. Typically my only critical dimension might be along one axis or might just be the size of holes. With experience and consistency of PLA and Cura settings you can predict how much to make the holes larger ahead of time and don't need to print a test print.

My rule of thumb is always make holes .5mm larger and if anything else is critical, print about 2mm height, pause the printer, measure it on the bed and if it's too big or small, abort the print and adjust the CAD model. Z dimensions tend to be the most accurate and for large parts X and Y tends to be extremely accurate.

 

[thorleif-jacobsen 0]

Ok, so i should be pretty happy with my setings now. Just waiting for a reply regarding the bearings It cant be anything else than that which is making the backlash. I can see it shake every time it goes on the front or left direction a little