Destring=Extrude - Getting nowhere fast!

Locked
sinned6915
Posts: 55
Joined: 28 Nov 2017, 10:06

Destring=Extrude - Getting nowhere fast!

Post by sinned6915 »

So I am trying to configure a new printer - gantry stylem, 900mm Bowden, I run through the wizards for Printer and Material and then start the Temp Tuning Wizard to get that worked out.

Somehow, this combination results in a perfect combination of settings that results in NO extrusion. The combination of default Destring and preload for PLA results in the retaction equaling almost EXACTLY the extrusion. Watch the tell tale on the extruder motor.

https://www.youtube.com/watch?v=ZUGQbGuOkHA

There are links in the description for watching the printing and generation of the test cube so that the settings might be visible. (I redid everything afte the first frustation attempt and decided to screengrab for godo measure).

If I set Destring to 0, the behavior seems to go away, at least I got a passable print- but think that the DeString is valuable feature and I dont want to loose it.

Here is a link to the gcode file- https://drive.google.com/open?id=1qOfg3 ... NoKUeIYmVf

How can I make this behave better?

Do I want to disable Destring?

I dont want to go back to the S3D or Cura dungeon...
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: Destring=Extrude - Getting nowhere fast!

Post by pjr »

Theoretically, when Preload is tuned correctly, destring value should be zero.

In practice, that is not usually the case, with small destrings being required.

Having looked at your G-code file, there is certainly sufficient positive filament feed to create an extrusion; nothing like as seen on your video. However I did only check a few preload/destrings within the G-Code as I have no idea which part of the model was being printed during your video.

Your G-code file does show that for every 1mm "suck" there is a 1mm "prime" and that each Preload "advance" at the start of the path, there is a Preload "retard" at the end of the path.

I note your G-code has some weird stuff at the start of many layers; looks like you have "Prefix" code in the "Warm same extruder" code area.

I also ran your G-code through Craftware viewer; it does appear from that to be OK.

Peter
sinned6915
Posts: 55
Joined: 28 Nov 2017, 10:06

Re: Destring=Extrude - Getting nowhere fast!

Post by sinned6915 »

Peter-

Thanks for taking the time to look at this.

The video was taken in the first layer of the 25mm KISS calibration cube. I would estimate that it was in the first half of printing.

Warm Same Extruder- I will redo the defaults. I had my initial wipe/prime code in there. Honestly, I do not understand the distinctions of those snippets. Is there a discussion somewhere about what should go where? The obvious ones are like Post or N-Layers. but the pre, warm same, cool same mystify me.

Destring- I had issues trying to do the PreloadVE tuning, gave up and set everything back to where I started. I need to revisit that. Setting it to 0 i get a descent print, but I am more confused now than before. Do we Preload tune with a Destring of 0? I might have missed that.

Is the filament retract amount automatic then? How do we control it? In S3D, my retract is set to 4.5mm and it works out pretty well.

Best regards,

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

Re: Destring=Extrude - Getting nowhere fast!

Post by pjr »

OK, so I checked the first layer of the G-code file.

There is a Destring Prime at the start of the layer and a Destring Suck at the end. No other destrings during the layer.

As for Preload, there is a Preload Advance at the start of the perimeter, a Retard at the end of that perimeter, an Advance at the start of the solid infill and a Retard at the end. Apart from the destring retards, all extrusions are positive.

Basically, the first layer is exactly as expected, so I really don't know where to go with this one.

Regarding the G-Code sections. I insert identifiable comments into each, slice a simple object and have a look at the G-code file, searching for the comments. I can't remember everything at the moment, so I don't want to document something that is wrong...

Destring during a preload tuning wizard print should not have any effect on the end result as sucks and primes will cancel each other out.

Destring in KISS is set as a single value on the Material tab - just below the PreloadVE setting; yours is currently set to 1mm.

Peter
sinned6915
Posts: 55
Joined: 28 Nov 2017, 10:06

Re: Destring=Extrude - Getting nowhere fast!

Post by sinned6915 »

PJR-

Yeah, its certainly a curious thing. I don't know what to do with it either. I am hitting the ceiling of my current knowledge base- time to move it!

I would really like to see how you dissect and analyse a gcode file. Would you consider a youtube live type session in doing something like that? Or one on your workflow in a small print? I think I would learn a lot watching someone else do it.

I am doing the PreloadVE tuning again, but this time with my other printer. I updated my previous thread.

Hopefully I can get that worked out on the Di3 then try it on the EV2.

Best regards,

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

Re: Destring=Extrude - Getting nowhere fast!

Post by pjr »

Dennis

Reading a G-code file is really simple; you just need a decent text editor (I use Notepad++) and off you go.

With KISS, the first 450 or so lines are comments (comments are text following a semi-colon) which detail the slicing parameters (Printer, Material, Style, Support) so that the G-code file can be read by KISS and the parameters read in - basically the KISS Project file.

Then there'a a section for the Prefix G-code and following that, the G-code to print the model.

At every layer change, there's a line like the following (from your G-code file):

Code: Select all

; BEGIN_LAYER_OBJECT z=0.200 z_thickness=0.200
At the start of the first layer, there is the "Selecting and warming Extruder" section and then the G-code for that layer.

Then there's the perimeter, starting with the comment line:

Code: Select all

; 'Perimeter Path', 0.7 [feed mm/s], 20.0 [head mm/s]
Then there's the perimeter print; the first few lines here extrude additional filament due to Preload - they build up the nozzle pressure with short X/Y movements while "extruding" much more than would normally be the case.

At the end of the perimeter, extrusion is reduced and then negated in the same manner as at the start, but this time to reduce nozzle pressure.

The the final section of the first layer is the infill "Solid Path", again with Preload Advance at the start and Retard at the end, followed by a destring suck at the end of the layer with Z-lift in preparation for moving to the start of the next layer.

Have a look at your G-Code file; it really will not take you very long to see what's going on.

Peter

BTW, this is the G-code wiki I use: http://reprap.org/wiki/G-code Be careful, there a whole load of commands there which are not relevant to either 3D printing or your printer's firmware. There really are just a few G-code commands you need to know about...
sinned6915
Posts: 55
Joined: 28 Nov 2017, 10:06

Re: Destring=Extrude - Getting nowhere fast!

Post by sinned6915 »

Peter-

THANK YOU!

I will dissect this file as you directed and see what I can figure out.

I too use Notepad++. I think I ran over a bunch of the things you pointed out trying to parse through the sections.

I posted followup to the PreloadVE thread I started before.

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

Re: Destring=Extrude - Getting nowhere fast!

Post by pjr »

Dennis

Pleasure to help. If you have any more queries, please ask.

Peter
Locked