Inequal Wall Thickness When Testing Flow

Turbo

Doing some double walled tests to determine flow after calibrating steps/mm, and each wall is showing a different measurement.
line width is set to 0.4, so the expected outcome is 0.8mm wall thickness with two perims, however all 4 faces measure different. Measurements vary anywhere between 0.75mm up to 1.1mm. This is a repeatable issue and have recorded it on 7 different boxes, with no changes with different slicers. No retraction or pressure advance is enabled. What could be my issue?

0.4mm line width set to 80mm/s locked
2 walls
this is tuning for abs on my corexy system

3DPMicro

my first guess is if its consistent and the g-code says its extruding the same on each wall at each layer then the nozzle is dorked up. I calibrate extrusion width with a single wall @ 1.125 x the nozzle diameter = extrusion width. How long are the walls?

Turbo

@3dpmicro 100mm square

Turbo

doing a single walled print yields similar results, with two being undersized, and two being oversized when compared to the line width.

CCS86

@turbo

What are the printed box measured dimensions?

What is the distribution of the error? (1.1mm near corners and 0.75 mm midspan? Random?)

alankilian

Please post your GCODE.

Turbo

@ccs86 doing a standard 30mm box, measurements range fron 29.8-29.9mm, with the lowest measurements near the center of each edge. There is no pressure advance enabled as to keep it out of the equation.

engikeneer

@turbo FYI, in PrusaSlicer (and likely in most other slicers), two 0.4mm line widths don't make a 0.8mm thick wall:
https://help.prusa3d.com/en/article/layers-and-perimeters_1748/

Is it repeatably the x-walls or the y-walls? Maybe some artefacts on the walls? It could be that you have some small backlash/flex in one direction of your gantry which means the layers don't always sit on top of each other?

Might be unrelated I've struggled with some inconsistencies on my printer because my part cooling fan blows from the rear side (and the duct is not very good), so the back of the prints get over-cooled and the front barely gets any. If you're doing fast single-walled prints, there's not always enough time for each of the layers to cool before the next one comes.

Do you get the same results if you turn the box 45deg?

CCS86

@turbo said in Inequal Wall Thickness When Testing Flow:

@ccs86 doing a standard 30mm box, measurements range fron 29.8-29.9mm, with the lowest measurements near the center of each edge. There is no pressure advance enabled as to keep it out of the equation.

Bowden printer?

You will get MUCH better help if you make an effort to provide all relevant info: printer specifics, firmware config file, gcode, photos, etc.

Turbo

@ccs86 it is direct drive. It is a Voron 2.4 printer. Gcode looks fine and doesnt have any anomalies. firmware is klipper so i dont know what good itll do being posted here.

Turbo

@engikeneer ill try a 45 deg rotated cube to see what happens

alankilian

@turbo said in Inequal Wall Thickness When Testing Flow:

@ccs86 it is direct drive. It is a Voron 2.4 printer. Gcode looks fine and doesnt have any anomalies. firmware is klipper so i dont know what good itll do being posted here.

Many people have come here asking for help and they are 100% sure something isn't wrong... Until it turns out that's what was wrong.

That's why people ask for things like GCODE, config.g homedelta.g, photos of the wiring etc when people ask for help.

Phaedrux

backlash in the mechanics? Tension on the filament?

CCS86

@turbo said in Inequal Wall Thickness When Testing Flow:

@ccs86 it is direct drive. It is a Voron 2.4 printer. Gcode looks fine and doesnt have any anomalies. firmware is klipper so i dont know what good itll do being posted here.

Seems like the Voron discord what be a better diagnostic source for a Voron running Klipper.

PaulHew

@Turbo Being an owner or sufferer of a Voron 2.4 with the After Burner, Klipper+Pi, it IS a well known and documented issue with extrusion widths. Random Wall extrsuions
There is a Prusa issue number reporting the same issue, that no one has got to the bottom of.
Voron shutdown discussions on it and deleted the post and went as far as banning people from the Voron discord for discussing it.

I did upgrade from Dual X MGN9's to single MGN12 and noticed a huge difference in wall quality. (I think the MGN12 is an official Voron Mod now)

I am going to presume you have done the basics. I was a bugger for skipping parts and have learnt why they are important.
This is the Overview which I use.
PiD - Auto PiD tune hotend (HEat Soak first)
Estep - Mark 120mm from extruder, extruder 100mm, measure difference
Temps - temp tower print (Print 4 towers at a time)
Flow - Use ss cali tool for rough setting, fine tune with bearing fit (Top part of Voron cube)
Pa - Not necessary as needs to be re calibrated after Shaper (allows you to use small test prints for calibration)
Squares - bed mesh calibration (manual edit)
Shaper - standard Input Shaper procedure
Accels - Accel tower print, highest acceleration without loss of detail after shaper
Pa - standard PA procedure (Either Marlins or Klippers)
Top flow - Large flat surface - tune top surface flow multiplier
Overlap Percentage - varying overlap value for top layer
Retraction - retraction print test
Part Cooling - vary values to get the best of layer adhesion and overhang quality

Work arounds are using one of the extruders by Annex. Or use one of the fan and duct mods in VoronUsers.
Use a decent hotend, Dragon or Mosquito.

For instance, running Klipper on my Duet boards on a RailCore, was abysmal, Bad levelling and my print quality degraded.

Bit the bullet and bought a Mini 5+ and a toolboard for the RailCore and have not looked back.

My friend is a Voron PIF provider and helped me tweak my settings in SuperSlicer too, and I am very pleased with the output now on the RailCore

He also documented a complete checklist on tuning which I use. (See above)

IIRC, he suffered from bad cooling, so runs different fan and ducts.

I have not used the 2.4 in months as it is not reliable enough.
I need to finish the Voron V0.1 with Duet Mini, (85% there), then strip the 2.4 down and build it with a Annex Beta release Gantry for Voron 2.4's.
I will be using a Sherpa Extruder with it and microswitch bed levelling, no more inductive probe issues.
Also thinking of upgrading the Pi and Skr to be able to run RRF.

As you mentioned, if it is thicker at the corners and thinner in the middle, I would look at making sure your hotend PID tuning is good. Heat soak first.
I also presume you are using 'Docs' SS profile. He runs a HighFlow Dragon IIRC.
Heat soaking a voron as with most enclosed printers for printing ABS is also important.
My RailCore is soaked at 100 bed and 120 nozzle for an hour minimum from Cold. Cooling fan is also on to move the air around.
The Voron I used to soak for 1.5 hours

I appreciate this a Duet forum, but I hate to see people struggle

HTH
Paul.

Turbo

@paulhew so ive gone through that checklist, and i keep getting stuck on flow, cause my parts are coming out undersized. i did a 45 deg cube, and it showed perfect on one edge, and undersized by around .15-.2mm across a 20mm cube. Logic says that theres some friction on that drive right? well i cant find any, no racking, equal tension, nothing i can find is wrong. heat soak is done for about an hour as well, until i see the hotend thermistor hit 50c (kinda as a chamber temp gauge. i know its on the low end but it worked fine for my other printer).. thoughts?
I havent gotten much valuable info from the discord since things just keep moving and getting buried unanswered, so i came back to where i know i could get some good answers, as i have 2 other printers running duet. Didnt know they were banning people for talking about an issue. its not issue no. 6 is it?
This thing is causing me more headaches than any other printer. while its a great design, i just cant keep smacking my head against the wall like this.

Turbo

@ccs86 i dont believe it to be a software issue though. im trying to think of any mechanical issues that could be causing it, besides friction. Because ive quadruple checked that theres no racking or bad movement on all axis

Turbo

@phaedrux im 99% sure its mechanical, but idk where to look. ive already made sure theres no racking/friction issues. as for tension on the filament, im hitting the expected volumetric flow of the mosquito hotend without skipping (0.5a on motor), so i dont think its an issue?

PaulHew

@turbo Morning. When Mak refers to flow, step 4 that's Extrusion Multiplier in filament section. I am sure you worked that out.

Make sure you are running enough amps on the extruder.
(I hate the Klipper calc for amps)

The prusa issue number was 602.

My only suggestion now is to dump the AB, use something like the Sherpa or go Bowden which I know has its own issues, but might be good for a test.
3.1 mobius iirc. the 2.1 extruder. Try the M4?

If you are running dual X, disconnect the bottom rail. Go MGN12
Try a different fan mod. My friend runs the 'BadNoob version 25'

I remember looking at this.
https://github.com/VoronDesign/VoronUsers/tree/master/printer_mods/Badnoob/Rail_Alignment

Tempted to build a Tiny-M at some point.
The V0.1 is so I can take places and print stuff for the Just-In-Case!

I will dm you about the Beta I am going to do to the 2.4.

P.

Turbo

@paulhew Well aware that extrusion mult = flow. I tried an lgx extruder cause why not, but got some z banding issues outta nowhere and the for some reason wasn't able to reach as high of a volumetric flow rate (despite current adjustments and temp increases), so i went back to clockwork. As for the x axis. its already on the single MGN12.
Already went through the rail alignment process, and considered the improved duct, but im getting fantastic overhangs with the current setup so i dont want to change it too much (always on 25% 65% for bridging. Maybe thats too muchand is making the abs shrink too quickly?).

My v0 is also for on the go, and rarely gets turned on while im home. The pi has its own network setup so i dont need wifi to use it.