Yet another tuning wizard thread.

ruffle
Posts: 12
Joined: 29 May 2018, 03:40

Yet another tuning wizard thread.

Post by ruffle »

I'm trying to get KISSlicer (1.6.3 on Debian Linux) tuned for my Lulzbot TAZ6 with an Aerostruder head (E3D Titan Aero with V6 hot end).

I've printed and used Michael Hackney's extrusion measuring tool, tweaked my e-steps and it's now repeatable to +/- 0.1mm (99.9 to 100.1mm extruded for 100mm commanded). I did cut down the 180mm/m extrusion rate in the middle of his gcode to 120mm/m as I'm using ~3mm filament and at 180mm/m it was going a bit crinkly and curly.

Doing the temperature test using Mattforge PLA I don't see any appreciable difference over a 180-220C range so I've stuck with the manufacturers suggested 195C.
IMG_0078.JPG
The results of the flow rate test is puzzling me though. I've seen lots of references on this forum to the E3D V6 flowing 10 mm^3/s but my results suggest something more like a touch under 4mm^3/s (the test in the image below goes from 10 at Z0 to 1 at the top (left).
IMG_0084.JPG
I tried again with a 4(at Z0) to 1 range:
IMG_0086.JPG
Which confirms the flow rate at ~3.5mm^3/s.... huh? :?:

My nozzle is 0.5mm purchased directly from E3D and the Titan Aero is the Lulzbot Aerostruder version. No cloned parts. The filament is 2.80 diameter (checked with micrometer in several places and averaged; not a lot of variance).

I did change the filament width to 2.80 in the Matl tab and the extrusion width in the Wizard Temperature style to 0.5mm to match my nozzle and to check I mic'd the result.... which shows a wall thickness of 0.556 mm; not what I expected.
IMG_0087.JPG
So I did a couple more experiments:

Code: Select all

Extrusion width with 0.5mm nozzle (Nozdia1) and 0.9 Nozlen1 (from E3D Drawing).

Diam     Setting         Flow          Actual

2.85         0.5            1             0.558
2.80         0.5            1             0.566
2.80         0.5            0.88          0.5021
2.80         0.6            0.88          0.5954
Reducing the filament diameter from the theoretical 2.85 to the measured 2.80 increased the wall thickness as one might expect.

However, I had to reduce the flow to 88% to get the 0.5mm wall thickness I'd set in KISSlicer. FWIW, I also tried setting 0.6 extrusion width and at 88% flow it was pretty close.

So.... can anyone suggest why my E3D V6 nozzle/Titan Extruder doesn't extrude at anywhere near the expected 10mm^3/s...

...and why the actual wall thickness is some 12% thicker than programmed? I really wasn't expecting to have to reduce the flow to 88%.

TIA.
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: Yet another tuning wizard thread.

Post by pjr »

According to E3D, the maximum melt rate of the V6 with 1.75mm filament producing good prints is 11.6mm^3/s at "notmal" temps with a 30W heater cartridge.

So, if your melt is lower, then you have an issue somewhere... I think I would be tempted to increase the temp to 210 or even 215 degrees and print the flow tower again.

Peter
ruffle
Posts: 12
Joined: 29 May 2018, 03:40

Re: Yet another tuning wizard thread.

Post by ruffle »

Thanks for the reply.

I tried another flow test (10 at Z0 to 1 at the top) at 215C and it doesn't really look any different to the one at 195C :-(
20180530_111735.jpg
So I tried some Colorfab Ngen at 230C (filament diameter checked with mic and set back to 2.85)
20180530_124627.jpg
which looks very similar to my PLA test... maybe even a touch worse... and the wall thickness is 0.602 - some 20% thicker than expected rather than the ~12% I was getting with the PLA.

I take it I am meant to be looking at the seam for this test? The help says "..where the surface quality looks good" but doesn't mention the seam specifically.

Should I be expecting the wall thickness to be 0.5mm as programmed (I'm using an 0.5mm nozzle)?

I'm really confused now and running out of things to try. Suggestions very welcome.
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: Yet another tuning wizard thread.

Post by pjr »

Any chance you could upload the G-code (zipped) of that test cube?

Thanks

Peter
ruffle
Posts: 12
Joined: 29 May 2018, 03:40

Re: Yet another tuning wizard thread.

Post by ruffle »

Certainly.

ks_wizard(35) is the PLA at 2.8 mm diameter and 215C, ks_wizard(36) is the 2.85 and 230C for NGen.

I've also included the model but it's the standard wizard one.

Thanks for looking at this.
Attachments
tests.zip
STL and gcode files.
(153.24 KiB) Downloaded 93 times
ruffle
Posts: 12
Joined: 29 May 2018, 03:40

Re: Yet another tuning wizard thread.

Post by ruffle »

Hmmm.... just found this thread: viewtopic.php?f=23&t=2066 where Michael Hackney talks about the flow test and rips.

So I tried another test, this time with the flow range from 20 to 1 and I get rips as shown in that other thread. They stop at around 14.5 mm which translates to a best flow rate of ~9mm^3/s - much more like the number I was expecting.

Have I been looking at the wrong thing all along?

Should one go with a high flow rate that 'rips' and see where the ripping stops to get the actual max flow rate?
20180530_165210.jpg
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: Yet another tuning wizard thread.

Post by pjr »

Looking at your settings I would comment as follows:

Set the max flow rate at 10mm^3/s - your speed settings will not get near the 60mm/s max print speed setting.
I believe that both your preload and destring [mm] settings are too high.
velDstr looks high (for a geared extruder)
for PLA, you can reduce Min Layer [s] to 4 (or less)
Set "Cool" to 100
Z-lift looks very low.
Don't get too hung up with the wizards.

Peter
ruffle
Posts: 12
Joined: 29 May 2018, 03:40

Re: Yet another tuning wizard thread.

Post by ruffle »

Thanks for taking the time to look at this Peter.

It seems that I've completely misunderstood the use/purpose of the Tuning Wizards. I was under the impression that
one should start with the default settings (as generated by the Profile Wizards) and tune from there.

Most of the settings you commented on (preload, destring, velDstr, Min Layer, Z-Lift) were generated by the Profile Wizards and I agree, are probably not
optimal for my TAZ 6/Aerostruder.

I originally started down the path of using the Tuning Wizards as I'm suffering from blobbing on cylindrical wall's seams and the popup help bubble on Seam Gap suggests that one has to tune the preload for this to be effective.

I'll go back to my 'normal' settings and play more with the seam stuff to see if I can get rid of those pesky blobs.
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: Yet another tuning wizard thread.

Post by pjr »

No problems.

The tuning wizards are really only a guide; when it comes to printing real models, things can change. Yes, they do give a good indication of required settings and the "Preload Wizard" is the one I use the most. But you have to take other things into account.

Retraction (destring) only needs to remove the nozzle pressure and when preload is tuned, the destring should be very small. I think E3D recommend a maximum retraction of 0.8mm (for 1.75mm filament - sorry, I run "original" Prusa printers using 1.75mm so I don't know too much about 3mm settings).

Peter
Cthutq
Posts: 53
Joined: 12 Jul 2018, 13:45

Re: Yet another tuning wizard thread.

Post by Cthutq »

This is not a concrete, but a general question
I want to ask why Michael Hackney stopped doing a video tutorial on setting up Kissliser? Promised and does not. It was very convenient to study the program on his video and do not ask unnecessary questions on the forum. Is it possible to know whether the author plans to continue to shoot training videos?
Post Reply