Hitman

Show off your prints!!!
jsc
Posts: 1864
Joined: Thu Apr 10, 2014 4:00 am

Re: Hitman

Post by jsc » Fri Feb 19, 2016 12:38 am

Your e-steps calibration should have the same overall result as adjusting the extrusion calibration, so it's not "wrong". In fact, I remember some discussion earlier about whether the default e-steps was theoretically incorrect, maybe from Tim? But you seem to have increased esteps/mm, which means (I think) you effectively increased the extrusion multiplier. Which is odd, because most of us have found an extrusion multiplier around .9 to be correct.

charles.yates
Posts: 59
Joined: Sat Dec 19, 2015 9:12 pm

Re: Hitman

Post by charles.yates » Fri Feb 19, 2016 2:06 pm

jsc wrote:If you're just doing a simple plane cut, you can do it without any additional software. Just duplicate the object, rotate one 180 around x or y, and drop both below the plane of the bed to the same cut line. A little simple arithmetic may be involved.
Is that a Simplify3D thing? Not sure if you can do that in repetiereHost.

FWIW, I used openjscad to load the STL, then renamed the generated function from main to charizard and after a bit of messing about, ended up with a new main which read:

Code: Select all

function main( )
{
    return intersection(
        charizard( ).rotateX( 0.178944 ).rotateY( 0.087082 ).rotateZ( -3.6 ).rotateZ( 0 ),
        cube( 150 ).translate( [ 0, -75, 0 ] ) );
}
I got the X and Y values by using 'Lay Flat' in the rotation option of repetierHost, and the Z by trial and error. The last Z was changed to 180 to get the other side of the model. The size of the cube was chosen to match the maximum extremity of the model itself. I repaired the generated models in netfabb.
jsc wrote:Your e-steps calibration should have the same overall result as adjusting the extrusion calibration, so it's not "wrong". In fact, I remember some discussion earlier about whether the default e-steps was theoretically incorrect, maybe from Tim? But you seem to have increased esteps/mm, which means (I think) you effectively increased the extrusion multiplier. Which is odd, because most of us have found an extrusion multiplier around .9 to be correct.
Yeah - that bothers me too. The only thing I've printed since the initial test was the Test your 3D printer v2 thing and that came out great, but the M4 hole is too small to accept the M4 nut, so I think I will need to tweak this some more.

I tried a multiplier of 0.9 with these settings, and that failed to even lay a 0.2mm first layer - I had to up it to 0.98 before I was getting a good result (haven't checked how this impacts the thin wall test or the test above yet). Something else for the weekend :).

jsc
Posts: 1864
Joined: Thu Apr 10, 2014 4:00 am

Re: Hitman

Post by jsc » Fri Feb 19, 2016 8:34 pm

Your first layer depends crucially on the z end stop setting, so you can't use first layer adhesion as a test of your extrusion. You need to dial in extrusion, then adjust the z height either mechanically or using a z offset. The calibration cube can help you do that (there is a link to how to do that from the post I linked to).

charles.yates
Posts: 59
Joined: Sat Dec 19, 2015 9:12 pm

Re: Hitman

Post by charles.yates » Fri Feb 19, 2016 11:27 pm

jsc wrote:Your first layer depends crucially on the z end stop setting, so you can't use first layer adhesion as a test of your extrusion. You need to dial in extrusion, then adjust the z height either mechanically or using a z offset. The calibration cube can help you do that (there is a link to how to do that from the post I linked to).
Hmm - been through all that :) - all fine until I lowered the extrusion multiplier. You sure they're not connected? Seems to me that a really low extrusion will simply not provide enough filament and you'll see that on the first layer as I was.

User avatar
ednisley
Posts: 1188
Joined: Fri Apr 11, 2014 5:34 pm
Location: Halfway up the Hudson
Contact:

Re: Hitman

Post by ednisley » Sat Feb 20, 2016 12:57 am

charles.yates wrote:all fine until I lowered the extrusion multiplier.
Judging from what you've written so far, you've been turning too many knobs at once. If you follow the step-by-step calibration process in Jules' guide, you will get a better outcome with less drama.

By and large,what you need to do: reset the step/mm value to whatever it was, measure the actual filament diameter and tell the slicer, set the Extrusion Multiplier to 1.00, reset the Z offset in the slicer, find a physical Z setting that lets you print thinwall calibration squares, measure the walls to set the Extrusion Multiplier so the slicer agrees with reality, then measure the height to fine-tune the slicer's Z offset to get the first layer spot on.

At that point, you have a level platform with a good Z=0 setting, the extruder will produce threads with the width that the slicer expects, and first-layer adhesion will be working.

Then run all the printer test pieces you want...

charles.yates
Posts: 59
Joined: Sat Dec 19, 2015 9:12 pm

Re: Hitman

Post by charles.yates » Sat Feb 20, 2016 1:08 am

ednisley wrote:
charles.yates wrote:all fine until I lowered the extrusion multiplier.
Judging from what you've written so far, you've been turning too many knobs at once. If you follow the step-by-step calibration process in Jules' guide, you will get a better outcome with less drama.

By and large,what you need to do: reset the step/mm value to whatever it was, measure the actual filament diameter and tell the slicer, set the Extrusion Multiplier to 1.00, reset the Z offset in the slicer, find a physical Z setting that lets you print thinwall calibration squares, measure the walls to set the Extrusion Multiplier so the slicer agrees with reality, then measure the height to fine-tune the slicer's Z offset to get the first layer spot on.

At that point, you have a level platform with a good Z=0 setting, the extruder will produce threads with the width that the slicer expects, and first-layer adhesion will be working.

Then run all the printer test pieces you want...
Well, in truth, I've kinda condensed things here so it looks that way I guess.

But for me, the point is that I'm getting significantly better prints right now without dicking about with the settings beyond what I have written above :). Keep in mind that I was getting good prints before (as evidenced by the opening posts and others). The discussion has been about improving certain aspects which, I feel, I have. But also, it's about viewing how I look at a print - suggestions like cutting a print down the middle were very welcome for example.

Spent the evening printing replacement parts for the M2 in the black PLA, and so far, 3 excellent parts produced with a fourth on its way, so I'm happy :).

User avatar
PcS
Posts: 667
Joined: Mon Mar 09, 2015 12:19 pm
Location: Michigan

Re: Hitman

Post by PcS » Fri Feb 26, 2016 2:06 pm

After reading this whole post my take is at a .98 multiplier I can say I would bet your z is still to wide. The higher the multiplier the thicker your first thread. And as you said you have had bed adhesion issues. Most can be attributed to not getting a good first layer. I used .89 for my print at 30 mms only retract when crossing open spaces .2 coast no wipe no z hop no cooling .2 layer 2 perimeters 4 bottom and top layers and 40% infill.This was my first attempt and sliced with s3d. I did Auto generate the supports and remove a few and add others.
Attachments
20160226_060329.jpg
20160226_060901.jpg

charles.yates
Posts: 59
Joined: Sat Dec 19, 2015 9:12 pm

Re: Hitman

Post by charles.yates » Sat Feb 27, 2016 12:18 pm

Many thanks for the feedback PcS. Nice print too :).

I really don't have bed adhesion issues that I'm aware of - the only time I had issues was when I reduced the extrusion multiplier. Or do you mean that the toppling of the supports is indicative of the multiplier being too high? The thin wall test is still checking out for me with my current set up (modified E, diameter 1.75mm, multiplier of 1, extrusion width 0.4).

It occurs to me that I may have the option to give s3d a trial run if someone could send over their gcode - would be interesting to compare with slic3r's output at least. (I realise that some stuff may have to be tweaked to run here, and I'm bypassing all the UI interaction which is, I'm sure, part of the appeal, but mostly I'm curious about how it compares at the point of printing). The gcode for the Charizard would also be useful, if it's not too much trouble.

FWIW, I did the two half print and that came out really well - I'll post some more pics down the line, but right now, I'm busy with another couple of projects which I'll post details of once I get the stability I'm looking for :).

User avatar
PcS
Posts: 667
Joined: Mon Mar 09, 2015 12:19 pm
Location: Michigan

Re: Hitman

Post by PcS » Sat Feb 27, 2016 12:42 pm

Yes your extrusion multiplier is really high for any layer height. If you set it at .90 and run a print it probably does not stick. The M2 comes set with spot on steps for the extruder and almost all users hover within a point or two of .90. I have in the past when I have set a too tight Z gap lowered it down to .80 to make up for it. In other words there is a direct correlation between your extrusion multiplier and thread thickness. You can use a caulk gun at .25 and keep it the same distance across an object but if you push too much you get fatter lines and it will also fill a larger z gap. I will attach Hitman as I ran him. I noticed Chazared does not sit flat when dropped on the build plate . At least at a .2 layer height. I noticed it never started to print his feet until layer number two. I will upload that file as well with that problem fixed.
Attachments
hitman.zip
(2.79 MiB) Downloaded 406 times
Charizard.zip
(1.48 MiB) Downloaded 394 times

charles.yates
Posts: 59
Joined: Sat Dec 19, 2015 9:12 pm

Re: Hitman

Post by charles.yates » Sat Feb 27, 2016 1:00 pm

Yeah - when I lower it, it wasn't sticking.

OK - I'll try with the default E steps again and try the 0.9 once I'm done with my current project (it's kinda finely tuned involving interlocking parts - feel that messing about with the parameters now could mess me up).

As for Charizard not being flat on the bed, yeah - I needed to use the Lay Flat option in repetierHost too.

Many thanks for the gcode stuff - will look into it shortly.

Post Reply