Poor print quality with RRF3 - especially 3.2.2.

JoergS5

This post is deleted!

deckingman

@JoergS5 I refer to my reply to another user above, timed at 17:41

hackinistrator

honestly ,i think you're so obsessed with the firmware (even if its for obvious reasons) that you no longer check the basics .
there are so many things wrong with your "shyte" prints that no firmware bug can generate .

Phaedrux

@hackinistrator Easy now.

deckingman

@hackinistrator said in Poor print quality with RRF3 - especially 3.2.2.:

honestly ,i think you're so obsessed with the firmware (even if its for obvious reasons) that you no longer check the basics .
there are so many things wrong with your "shyte" prints that no firmware bug can generate .

I'm trying my best not to let the experiences of the last 19 months (which have been due to numerous firmware issues) cloud my judgement.

You will note that I haven't directly stated that I think the issues are due to firmware. I have been very careful in my choice of words, using terms like "this is coincidental with changing to firmware xyz". Other people in this thread have stated that it looks like firmware issues - not I. I've also been careful neither to agree nor disagree with such comments.

How do you explain the print to print variability when nothing changes between prints - not even the filament? It's fine saying there are lots of things wrong with the "shite" prints, but how do you explain that repeating the exact same print, with the exact same settings even down to using the same filament from the same reel, gives "acceptable" (not perfect) results?

It's impossible to assess the impact of any changes unless one can establish a reliable and repeatable baseline. Until I can get consistent print to print repeatability, I can never be sure that the results of any change I make would be due to that change or this mysterious "something else" (whatever something else might be).

One could argue that you are the one obsessed with it not being a firmware problem. But lets' not go down that road.

I remain open to any constructive suggestions.

jens55

@akstrfn said in Poor print quality with RRF3 - especially 3.2.2.:

I always find it hilarious when frequent forum members ask you to check some basic thing knowing well how crazy you are with your machine.

It might be annoying but there has been situations for me, and more than once, where I overlooked the basics and was thankful to have been reminded of them.

JoergS5

This post is deleted!

hackinistrator

@deckingman said in Poor print quality with RRF3 - especially 3.2.2.:

How do you explain the print to print variability when nothing changes between prints - not even the filament?

firmware also didn't change between those prints .

i'm not saying the firmware is perfect . I'm just starting with duet3 and i can say that so far with all the various (minor) issues i had , the diagnostics (m122) was pretty much useless in my case .
i also didn't like the default homing files , which mark axis as homed even if i dont have a sensor attached to the board (no sensor = NO = HOMED...) . this could potentially cause damage as i'm using large high current motors .

back to topic ,
your print has no layer shifts , so i guess the problem is extrusion related .

if i were you , the first thing i would change is the filament . if you didn't print for 19 months , i guess that pla is more then 19 months old . pla degrades over time and its the easiest thing to change .that curling is good indication of degraded filament (if temps were ok).

next i would test extrusion / retraction (including firmware )
for example extrude 11mm and retract 1mm 10 times , measure if total is 100 . do this with various feed rates until you reach approx 1.5x your max printing feed rate.
if you think its something related to multi axis/ stepper movement then you can add movements during this test .

your test print is pretty large , i would use something smaller for tests , even benchy would be better .

fcwilt

@hackinistrator said in Poor print quality with RRF3 - especially 3.2.2.:

i also didn't like the default homing files , which mark axis as homed even if i dont have a sensor attached to the board (no sensor = NO = HOMED...) . this could potentially cause damage as i'm using large high current motors .

It can be setup either way - to stop the motors when homing is attempted or to allow them to run to the end of the axis.

You can of course use stall detection which doesn't rely on endstop sensors.

Frederick

JoergS5

This post is deleted!

jens55

Yeah putting another unicorn printer in close proximity might get you into a heap of trouble with piles of little unicorn printers popping up everywhere - not a pretty sight!

(Couldn't help myself)

JoergS5

This post is deleted!

dc42

@deckingman, please try a print using 3.3beta1. After homing and before printing, run M122 on all 4 boards to clear the stats. Then start the print. If/when it's looking bad, pause the print, run M122 on all the boards again, and post the results here. What I am particularly looking for is whether you get send timeouts on the main board and/or nonzero 'oos' counts on any of the expansion boards.

deckingman

@JoergS5 said in Poor print quality with RRF3 - especially 3.2.2.:

@hackinistrator I second that: I hesitated to propose to check something simple, because @deckingman is such an experienced user.

But if you look at the forum, there are often cases where, after a long search for complicated reasons, it was something simple like

dirty pulley or damaged belt

off-center pulley

old filament, being too wet

bent shaft of a stepper

damaged nozzle

The easiest would be to have a clone printer and compare the results. Unfortunately it's not so easy with a unicorn printer! But somehow it's necessary to find a decision between hardware and firmware problem.

Oh, please. A dirty pulley which magically cleans itself? (And the printer is in a dustproof booth). An off centre pulley which magically makes itself centred. The filament is a brand new reel but it's hard to imagine some metres of that reel could be wet, yet for the next print, it will be dry. A bent stepper shaft which straightens itself for one print and bends again for the next? A damaged nozzle which undamaged itself for the next print then damages itself again?

deckingman

@dc42 Thanks. Will do.

deckingman

One other thing that I hadn't mentioned, the printer doesn't have an RPi fitted - so this is all stand alone and anything to do with SBCs can be discounted. For what it's worth, I don't have a PanelDue either - just DWC on an aged laptop.

JoergS5

This post is deleted!

droftarts

@deckingman Can you take a picture of the hot end setup at the moment? I don't think I've seen a picture of how your machine is currently configured. Might give some insight into any potential physical (rather than firmware) issues.

I assume you're using one of the six extruders to feed the hot end, but the others are still mounted. Another test to rule out extruder issues would be to swap to using a different extruder (if you haven't done this already).

Recently I had the grub screw loosen, that holds the hobbed drivegear on the motor shaft on a BMG extruder, just enough for it to wiggle a bit on the flat but not completely rotate free, and took me a while to figure out what was causing extrusion problems that are not wildly different from yours ('random' over and under extrusion, particularly at the end of lines; basically retraction wasn't doing anything, so filament oozed on moves). This was caused by swapping extruders around, as I've recently got an Orbiter v1.5, and not fully tightening the grub screw on the shaft when assembling for the 100th time. Might also explain why pressure advance stopped working for you.

Just to be clear, this is in no way meant to be condescending, just an observation.

Ian

deckingman

@droftarts As it happens, I had considered using one of the other 5 extruders that are mounted but otherwise unused with this particular hot end. IMO, it seems to be more extrusion related than carriage motion related. Odd that the problems can appear to be present for the whole of one print and then completely absent for the whole of another print - but not beyond the realms of possibility. I'm planning on swapping to 3.3. beta as per Davids suggestion, but if I get a bad print and nothing shows up on the diagnostics, I'll try another extruder.

The best way to describe the current hot end arrangement is to point you to a recent video I've posted in another thread https://forum.duet3d.com/topic/21919/curious-audible-pitch-change-with-varying-move-length

That link will take you to a video showing it "in action" but if you want some pics, let me know.

Edit, 'er on doors has got car problems that I need to sort out, so I might not get much done today.

droftarts

@deckingman Thanks, I was just watching the video!

Ian