1

Topic: But math?

Hello all,

Just FYI, I got my Solidoodle 3 about a year ago, but due to life and stuff didn't play with it much.  I have now gotten back into it and have been making stuff the past 2 weeks.  I played around and figured out the following thus far:

1. I got a glass sheet and use that for the bed.  It is much flatter than the metal, and I was able to level it and get good contact over the entire bed.  It also allows for the beds to be easily removed and put in the fridge (CTE diference makes the print just pop off)

2. I figured out how to get a really good first layer bond.  Hair spray, proper bed heating, leveled bed, and correct Z adjustment to encourage squishing works wonders.

3. I measured my filament width and feed rate with calipers and got that plugged in.  It turns out that my Solidoodle 3 out of the box was using 25% more plastic than it should have (I.E it pulled in 125 mm instead of 100mm when told to pull 100 mm).  So I set my feed rate to 80% to get it to pull evenly.

4. I meassured the width of my filament comming out of the nozzle (.38 mm) and set my printer accordingly

5.  I made a single thick box with the perimeter set to (Using advanced tab in slic3r) .38 mm and it came out to .41 mm width.  I added a .93 multiplier to extrusion rate and it now comes out at .38 mm.

The problem is that my solid infills now look 'under filled'  The strands don't touch each other, and they often wont touch the perimiter.  I didn't have this problem until after I messed with my extruder calibration (which effectively cut the flow rate to .8*.93).  So it seems I either have to chose between properly filled infills or having overfilled walls.  I made a simple sample cube with a 7.6 mm side and counted the strands, and there were 20 across.   So it isn't the slic3r missing strands.  It is just under extruding those.

So, is there some way to fix that?  I tried the following:

1. Yes, I checked belt tension and my pot voltage.  Those are fine.  I even used a digital scale to make sure tension on the belts was equal.

2.  I did the realignment thing to get good circles already.

3.  I check the extruder and made sure 100mm pulls exactly 100 mm.

4.  I tried fiddling with the extrusion width in the advanced tab.  That increased the number of layers, but also decreased the flow rate so the total infill percentage didn't change.

5.  I checked the extrusion width (Of a single perimeter at least) and it was exactly what it should be.  It just that the infill seems to be under filling.


So, anybody have these problems, and what did you do about it?

2

Re: But math?

For nozzle width, it needs to be .35, if your printer is a year old.  You can't go by the width that comes out of the nozzle into air because the expands as it exits (called die swell).  Since Slic3r thinks the nozzle size is .38 when it is .35, it may be spacing the threads a little too far apart.  Also the minimum extrusion width you can use is equal to the nozzle width.  .38 is pretty close to that, so .42 might give you more consistent results. 

Make sure you are measuring the filament as well, and don't assume it is actually 1.75.

3

Re: But math?

I checked the filament with the calipers.  It measures between 1.73 and 1.76.

As far as the extrusion width and nozzle width:

I have no idea what the nozzle width actually does.  I tried changing that number anywhere from .25 mm to .5 mm to see what would happen and as far as I can tell nothing changes (with the exception of the spacing of the filaments on the first layer after a raft).  But even that change doesn't affect the fill percentage because the strands will just get thinner/wider to compensate.

I also tried changing the extrusion width from .2 to .5 mm.  At .2, it just turned wispy and useless.  At .5 it is nice a nice and strong filament, but there are obvious gaps between the individual strands.

I tried the .35 and .42 settings you suggested and the problem still remains.

The only thing I can think of is that my filament has absorbed some water over the last year and is actually measuring thicker than it is when you heat it/drive the water out.

But the inevitable problem is the mismatch between the single wall thickness, and the proper infill of solid layers.  If I adjust filament diameter, extrusion multiplier, etc. in order to get a good infill, then it will inevitably lead to thicker than requested perimeters.

Also, other things to add.  I have ruled out the following:

It isn't print speed.  I changed perimeter and solid infill to both be low (30 mm/s) and the problem remains

It isn't nozzle temp.  I tried from 190 to 220 without any effect (Using ABS)

4

Re: But math?

the nozzle width dosent "do" anything its a variable that gets factored in with the other variables like your filament width. they are all part of some algorithm the firmware uses to tell the printerhead where to move and how much filament to push through the nozzle. if its not correct you will see it in your prints.

when you get all the settings right,adjusted, calibrated, bed leveled. your prints will start looking better

5

Re: But math?

Perhaps I was unclear.

When I say that nozzle width doesn't do anything, I mean it seems to be a variable which no equations use.  I can put whatever I want in there, but when I run slic3r, it doesn't make any difference.  The Gcode is the exact same, and when printed, to two models look the exact same.  That variable doesn't seem to affect slicing, or printing in any way.

Now obviously if I had a different nozzle (I.E. I put in a .4 mm replacement instead of a .35 mm) then it would have an impact on print quality.  But as long as I don't change the physical nozzle, just the number in slic3r, it seems to have no affect.

6

Re: But math?

Hello Trevor,

I got the Solidoodle 2 days ago and I still can't print. I notice the feedrate is not ok either (125mm instead of 100mm). Where can I change it to 80? I change the feedrate under speed multiply and did not work.

Any help will be apreciated!!

7

Re: But math?

The easiest way I found is to open your reptier host (I assume your using that one), then click config -> firmware eeprom settings (You will need to connect to your printer in order to do this.)

From there, change the steps per mm for your E axis to whatever it is currently * whatever your calibration multiplier came back as.  For me, that means changing it to 110 to account for the 25% over extrusion.  I also had to change my Y axis from 88 to 89 steps in order to get my prints to come out square, but I don't know if that was just me.

I still have issues with getting a full infill without over extruding, but I have just resolved to live with a slightly underfilled infill, and a slightly overfill perimeter for now.  I am currently looking at other slicers besides Slic3r to see if I can get better results.  I am also looking for something that does overhangs/supports better as Slic3r just makes supports that can't be easily removed from curved undersides.

8

Re: But math?

Thanks Trevor. I was able to calibrate my extruder. I made a cube using 0.3mm resolution and I notice some warping at the bottom of the printed cube (see attachment). Anyone knows how to fix this?

Second, I tried to print using 0.1mm and the print failed. I'm having trouble with bridging. Please see attachment. What do you guys think? To hot the temperature?
I'm running at 207C in the extruder (aprox 227C) and 105 in the bed.

Thanks for the help!
Jose

Post's attachments

0.1mm Part.jpeg 345.29 kb, 31 downloads since 2014-03-29 

Cube 1 - 0.3mm Resolution.jpeg 508.9 kb, 25 downloads since 2014-03-29 

You don't have the permssions to download the attachments of this post.

9

Re: But math?

Ahh, yes the peeling.

The lifting you saw on your black cube is due to a problem where the layers are deposited warm, and then cool.  As they cool, it shrinks.  This causes the top of the part to want to shrink relative to the bottom part, which makes it want to curl upwards.  There are a couple things you can do about this:

1. Better bed adhesion.  If it sticks to the bed, it wont peel up.  Try turning your Z stop screw a bit counter clockwise to make the layer squish more.  Or you could also increase the first layer extrusion width on your advanced settings in slicer to glob down a thicker base.  You could also try some adhesives.  I am using hair spray to excellent (almost too good) results.  Although, if you are going to use adhesives, you might want to get some glass plates for ease of cleaning.  Just search for glass plates or glass bed on here.

2.  Lower your print head temp.  Lower temp = less shrinking, and also better bridging.  I use 205 C for my first layer, and then 190C for the rest.  High temp gets good sticking, but low temp gives better results.

3.  Go slow.  Give your previous layer time to solidify before you lay down the next.  The filament -> cooling tab in slic3r is good for this.

As for your bridging:

1. Lower print temp

2. A fan is a good idea if your going to be doing alot of bridging (I am installing one this weekend)

3. Faster bridging speed

4. Mess with your bridging flow rate.


And just beware that calibration can take time.

10

Re: But math?

I will try your settings today
Thank You!

11

Re: But math?

How do you setup the first layer 205? and then 190?

12

Re: But math?

Hi Trevor,

Just read your experience which mirrors mine.  What settings did you finally use to get the extruder to solid fill and not just deposit thin strands on every layer above the first?
Much appreciated
Peter

13

Re: But math?

fispe wrote:

Hi Trevor,

Just read your experience which mirrors mine.  What settings did you finally use to get the extruder to solid fill and not just deposit thin strands on every layer above the first?
Much appreciated
Peter

Sounds like your extruder is not quite calibrated yet and you're underextruding. I'd recommend the following:

(1) Start with the extruder steps-per-mm calibration here: http://solidoodletips.wordpress.com/201 … libration/

(2) Follow on to do a wall-thickness calibration, instructions here. Make sure your target wall thickness and the wall thickness setting in slic3r (under the advanced section) are the same (if your machine is fairly new note that you probably want to target 0.48mm instead of 0.42mm to account for the slightly larger bore nozzle in newer machines): http://solidoodletips.wordpress.com/201 … flow-rate/

(3) Readjust your z-stop screw and first layer wall thickness setting (again in slic3r advanced section) as required to get the first layer to adhere well again.

SD3. Mk2b + glass, heated enclosure, GT2 belts, direct drive y shaft, linear bearings, bowden-feed E3D v5 w/ 0.9° stepper
Smoothieboard via Octoprint on RPi

14

Re: But math?

mendez wrote:

How do you setup the first layer 205? and then 190?

In Slic3r.

It's in the Filament Settings.

15

Re: But math?

HMM! Did all the calibrations a while ago. Just learned I may have caused my on problem,  I upgrades from the original RepetierHost 0_85 to 0_95 as well as Slic3r  0-9-10 to 1-1-1 and it was then that things got scrambles. I did a clean reinstall of the latest versions and now get a "G-Code not found" message after slicing (see attached screen sot). An thoughts on this?

Post's attachments

G-Code not found.doc 42 kb, 3 downloads since 2014-04-15 

You don't have the permssions to download the attachments of this post.