(AD5X) Poop-free operation

[ninjamida]

PLEASE NOTE: The code in this post is an OLD VERSION. I have left the original contents of this post up to keep the thread easy to follow, but please see this reply for an updated version: #175 (comment)

I used zMod + GuppyScreen, together with some customized macros, to make the AD5X print almost poop-free. I did find taking a poop then striking it was the best way to get the nozzle clean at the start of the print; and it still poops when changing filament outside of a print or before a print. However, no pooping for mid-print color changes.

As an added bonus, with the right slicer settings, it will obey the flush volumes set in the slicer!

NOTE: I have NOT tested pausing / resuming or resuming-on-error / resuming-on-power-loss with this. Use at your own risk! I do not advise integrating this into the mod (even as an option) until it's more tested; it will also likely need changes for integration, so that it's less hardcoded and more respectful of user settings (I just coded it to work for my settings - basically "I made this for me; if it helps anyone else, that's a bonus"). I will likely be using this as my standard setup going forward, so I should be able to provide feedback over time as to how reliable it is.

Also - I don’t know if zMod itself or my changes are responsible, but color changes involving TPU are not reliable with this setup. PLA has no problems, PETG should be supervised but generally works out fine.

Video: https://youtube.com/shorts/MxzDBuDXGA4?feature=share

Add this file in the same folder as your user.cfg, and include it by adding [include user-nopoop.cfg] to user.cfg.
user-nopoop.cfg.txt

Slicer settings: For Orca, go to Printer Settings -> Multimaterial -> enable Purge in prime tower. Also, make sure to set your flush volumes to reasonable values - so far I have found that using the autocalculated values, with a multiplier of 0.6, works well (tested with eSUN PLA+).

Filament.json: Set before cut lift to 20 and after cut to 5. Nozzle cleaning can be whatever works for you, I use 100.

Just in case it doesn't work / has issues without this, here's my user.cfg as well:
user.cfg.txt

And my SAVE_ZMOD_DATA settings:

info You are working WITHOUT native screen with GuppyScreen // DISPLAY_OFF GUPPY=1
21:48:01  info Parameters used at print start and bed leveling [START_PRINT]
21:48:01  // Don't build bed mesh on each print. // SAVE_ZMOD_DATA PRINT_LEVELING=0
21:48:01  // Don't use KAMP instead of BED_MESH_CALIBRATE // SAVE_ZMOD_DATA USE_KAMP=0
21:48:01  // Don't test bed mesh before printing // SAVE_ZMOD_DATA MESH_TEST=0
21:48:01  // Play MIDI on print start:  // SAVE_ZMOD_DATA MIDI_START=
21:48:01  // Play MIDI on print pause:  // SAVE_ZMOD_DATA MIDI_PAUSE=
21:48:01  // Check file MD5 checksum // SAVE_ZMOD_DATA FORCE_MD5=1
21:48:01  // SKEW correction disabled // SAVE_ZMOD_DATA DISABLE_SKEW=1
21:48:01  // Load Z-OFFSET from global settings // SAVE_ZMOD_DATA LOAD_ZOFFSET=1
21:48:01  // Enable nozzle line purge // SAVE_ZMOD_DATA DISABLE_PRIMING=0
21:48:01  // Purge algorithm: NV_PURGE_LINE // SAVE_ZMOD_DATA CLEAR=NV_PURGE_LINE
21:48:01  info Parameters used at print end and cancellation [END_PRINT]
21:48:01  // Play MIDI at print end:  // SAVE_ZMOD_DATA MIDI_END=
21:48:01  // Auto close dialogs on native screen after 20 sec (fast) // SAVE_ZMOD_DATA CLOSE_DIALOGS=2
21:48:01  // Auto disable motors after print/cancel in 25 sec // SAVE_ZMOD_DATA STOP_MOTOR=1
21:48:01  // Don't reboot printer automatically // SAVE_ZMOD_DATA AUTO_REBOOT=0
21:48:01  info System-wide parameters
21:48:01  // Save data for print recovery // SAVE_ZMOD_DATA SAVE_RESTORE=1
21:48:01  // Don't hide color selection when printing starts // SAVE_ZMOD_DATA SILENT=0
21:48:01  // Insert rod automatically // SAVE_ZMOD_DATA AUTOINSERT=1
21:48:01  // Don't Use trash when printing // SAVE_ZMOD_DATA USE_TRASH_ON_PRINT=0
21:48:01  // Use SCV from config for shaper graphs // SAVE_ZMOD_DATA FIX_SCV=1
21:48:01  // Fix E0011 error // SAVE_ZMOD_DATA FIX_E0011=1
21:48:01  // Fix E0017 error // SAVE_ZMOD_DATA FIX_E0017=1
21:48:01  // LED brightness at power on: 0 // SAVE_ZMOD_DATA LED=0
21:48:02  // Play MIDI at power on:  // SAVE_ZMOD_DATA MIDI_ON=
21:48:02  // Block Chinese clouds // SAVE_ZMOD_DATA CHINA_CLOUD=0
21:48:02  // Use Fluidd/Mainsail button settings from ZMOD // SAVE_ZMOD_DATA SAVE_MOONRAKER=0
21:48:02  // Turn off native screen after (sec): 20 // SAVE_ZMOD_DATA DISPLAY_OFF_TIMEOUT=20
21:48:02  // Using native Klipper 12
21:48:02  // Klipper process priority (1-min, 40-max): 20
21:48:02  // Weight control is not configured. // To configure: NOZZLE_CONTROL WEIGHT=1500
21:48:02  // Available languages: en, ru, de, fr, it, es, ja, ko, zh // LANG LANG=en

I may or may not try to fix issues with setups other than my own if issues are reported; I probably won't actively test other configurations to look for issues.

[ninjamida]

Video of a full-length print with this setup: https://www.youtube.com/watch?v=uGp1FpAGvqE

[ninjamida]

Edited first post to add a note about filament.json settings.

[ninjamida]

I finally tested this beyond just PLA. PETG is a bit iffier with it, lots of nasty looking blobs on the wipe tower where it changed (the actual print is fine though, but the tower looks like it could fail at any time). Might be fixable just by reducing the Z distance from the tower when it parks. Might also just be a result of that I’m mixing materials in this print. I’ll test more and find out. I also want to rewrite it to firstly, respect existing settings more, and secondly, allow customising some of the parameters of this code itself. I definitely can see this being a candidate for future inclusion in the mod, but it will need more work first.

[ninjamida]

Major issues with TPU, it won’t load at all (doesn’t even get into a print, just trying to load through Color macro). Not sure if that’s related to my customisation or an issue with zMod itself, need to investigate further.

Edit: Or maybe just some random oddity, as it’s working fine now.

[ninjamida]

I just ran a print involving TPU again, intending to watch closely and see if I could figure out what needed work and...

...it worked perfectly fine? (And that's despite being mixed-material, PETG and TPU, not just "two colors of TPU".) I haven't checked on the quality of it yet (it literally just finished, and I've been watching it via camera) but all the material switches succeeded.

I did reduce the load speed for TPU prior to this print, I wonder if that's all that was needed to make it work better...

[Absolutely-D]

Hi, I've been running this config for a few days now, and it's definitely looking promising. I've not done particularly large or complex multicolour prints, but I ran a 6 hour print just fine yesterday. I did have to up the flushing volumes into the prime tower slightly to stop colour bleed though (using about 90% ATM with petg)
I did initially have a few issues with filament stalls when changing, but I upped the previous filament retraction length to 120 in filament.JSON which seems to have made it ok at least in my limited testing.
Not tried multi material yet either, that's next on the list

[ninjamida]

Just wondering, what style of 4-to-1 adapter does your AD5X have?

It seems there's two different designs - a narrower and taller one that attaches with a screw, and a wider but shorter one that just clicks into place. Mine has the latter kind. From photos, it looks like the former kind might need a longer retraction distance, and that might be why the shorter distance worked for me but not for you, if you have that one.

[Absolutely-D]

Funny you should ask actually - I have both. The no screw variant came with my printer, but I had seen the different styles around and asked ff support. They sent me one with the screw out. I haven't tried it with shorter retraction settings since switching it out though, warrants some testing!
I have since taken the old one apart, because I'd toyed with the idea of trying e.g. a bambu 4 to 1 adapter in its place, but need that metal end piece to do so - so may get on that at some point too

[Absolutely-D]

Oh actually re reading your reply properly, I had the same as you initially, without the screw - which was when I was having the issues. Strange

[ninjamida]

That’s great to hear! I have made minor tweaks since my original post, I’ll try and remember to share those later today when I get a chance.

[Absolutely-D]

Ah nice, it'll be interesting to see what you've changed

[ninjamida]

Sorry - haven't forgotten about this, just found a minor issue that I want to look at before sharing it again. (and I'll probably integrate your suggestion too re: increasing the retract length, it's only going to add half a second or so to each color change so not a big deal and if it helps...)

EDIT: That's all done and sorted now, but I also want to make it so that if the user sets use_trash_on_print = 1, then it will have exactly the same behavior as "stock zMod" - and my custom behavior only kicks in as basically a replacement behavior for use_trash_on_print = 0.

I've written the code for this, but I'm in the middle of a print right now so haven't been able to test it yet.

[ninjamida]

I can also say I’ve used pausing and resuming a few times now and it seems to work well. Haven’t tried resume after power loss yet.

[vgilmore]

Not sure where I went wrong but when it does the filament change I get a jam message on the screen. I here the IFS feeding the filament and then it starts clicking and that is when I get the error. I am just testing with the 20x20 calibration cube and the prime tower is really large. I am using Orca Slicer.

[ninjamida]

If you look at your filament tubes above the print head - does it look like the one it's switching from has been properly pulled back?

Someone else suggested upping the retraction length for the previous filament (the "nozzle_cleaning_length" parameter in filament.json) to 120 and it solved their issues.

[ninjamida]

Okay... time to release an update!

This now respects the use_trash_on_print setting. If it's on, you get almost standard zMod behavior, with two minor tweaks (see readme). If you turn it off, it replaces zMod's no-trash-use with my setup. Be sure to see the notes in the readme about what your slicer settings should be when using this.

I have included both English and Russian versions here. I don't speak Russian, so the Russian version uses either the original zMod Russian texts, or the output from Google Translate - I've done this in case ghzserg wants to integrate this into zMod, as I recall him saying he prefers submissions to be for the Russian version.

I think at this point, this can be considered not so experimental, at least for PLA - and to be honest, I'd have a fair bit of confidence in it for PETG too at this point. (TPU is a bit more of a wildcard, and it remains untested with anything else - although I'll likely do some testing with ABS in the coming days.) I would say it could be considered for inclusion in zMod now, though with the caveat that it probably shouldn't be the default option just yet. More testing is needed, but at this point that's more the "what's the optimal settings" kind of testing (in particular around flushing volumes, as Orca's defaults don't work well - some are way too low, others are ridiculously excessive).

ETA: Also, in the original post I said that pause / resume was untested, I can now say it works fine.

Download: AD5X Poopless.zip

EDIT: Testing updates:

• Works very well with ABS.
• Works fine with a 0.25mm nozzle (PLA).

[Absolutely-D]

awesome, i'll do some testing meself - looks super promising tho

[ninjamida]

So, I took a brave new step with this modification last night, and left a print running overnight with it.

I woke up at some point during the night, checked the status on my phone while I was up, and noticed it had paused with an error during a filament switch. I forget the exact message but something about a timeout waiting for a response from the IFS. I wasn't going to get up and properly investigate, so I just hit resume, didn't even watch what happened next, just hoped for the best. Woke up this morning to a finished and completely fine print; so not really sure what happened there. It's possible this was an actual issue with my IFS - at the time this happened, it was switching to channel 4, and even before I installed zMod, occasionally I've had issues where channel 4 thinks it's empty when it's not (only channel 4, never had it happen on any of the others).

But that's a fairly good sign. Even if this issue is due to zMod or my modifications, it's a "print pauses until you intervene" error, not a "print gets messed up" (or worse, "printer gets messed up") issue.

[ninjamida]

Confirmed: Works just as well with ABS as it does with PLA, as I expected.

[ghzserg]

https://github.com/ghzserg/zmod/wiki/AD5X_en

Thank you very much for the work done

[ninjamida]

Also seems to work well with 0.25 nozzle based on initial testing (one test using PLA).

[geleia665br]

Hello, sorry for the translator's English, I'm Brazilian.

I've been in the 3D printing world for 10 years, I've built machines, I've been using Klipper for many years, and I bought my first Flashforge, the AD5X.

I went to do purge tests and calibration and discovered that there's no point in resetting the purge volumes on the slicer because it's configured directly in the firmware, and this disappointed me.

I've always been used to having full control over my printers.

With this mod, will I be able to disable this purge, right?

Is it configured in a macro in the original firmware, or does this mod already disable it?

[ninjamida]

Adding to my reply below (I always forget to reply to the specific comment on GitHub, my bad), if you wanted to take this concept even further, you could likely make the T0 T1 etc commands outright do nothing, and instead write custom GCode in the slicer that mimics the behavior of them. In such a case, it would probably be useful to write some macros for the IFS stuff.

I might look into that in the future, but for now, I'm quite happy with what this has achieved.

[ghzserg]

I plan to merge your changes into the main branch for the mode without use_trash_on_print. But I don't have the opportunity to test it, so I'm waiting for more people to test it and report back that everything is fine.

[ninjamida]

Do you only have the 5M? If so, it’s very impressive you’ve done this much work for the 5X without having one - aside from my own changes I’ve shared, I’m using your setup and it works very very well. (I should note I’m not one of those people with dozens of printers - I have exactly two; my AD5X and a very old Da Vinci I don’t use often. So this is on my main printer, not just a spare.)

[geleia665br]

Hello again, it worked here, but I'm having some problems.

I don't know if it's because of Zmod or Poop Free.

Sometimes it gives an error like the nozzle is clogged.

Then I click OK on the screen and tell it to continue printing, and it goes normally.

In other tests, I was experiencing underextrusion on the last layer, but apparently it wasn't clogging because I manually told it to extrude, and it did normally.

I really liked that you put the settings in the JSON format, so I have more control, adjusting how much to retract before cutting the filament, etc.

I'm using ABS.

[geleia665br]

Coming back to give some feedback.

Today I went back to testing, I replaced the initial gcode that was in Orca with the one from Flash Orca, the problem that was occurring stopped, and on the last layer it is coming out, it is bad but without underextrusion, now it is just a flow adjustment.

I will continue testing and calibrate the flow, then I will do the test with color printing and come back with the news, thank you.

[geleia665br]

I'm still having problems, sometimes it starts printing before the hotend temperature reaches the setting in the slicer, it's probably something in the start gcode, could you send me the gcodes that I should configure in the orca slicer?

[ninjamida]

I don’t believe it’s any different from the normal zMod start / end code, but I’ll double check tonight.

[ninjamida]

Sorry, I completely forgot about this until I came here for an unrelated reason just now. (My memory sucks. If I say I'll reply to something then I don't, 99% chance I just forgot.)

Here's my start and end gcode:

Start

START_PRINT EXTRUDER_TEMP=[nozzle_temperature_initial_layer] BED_TEMP=[bed_temperature_initial_layer_single]
SET_PRINT_STATS_INFO TOTAL_LAYER=[total_layer_count]

End

END_PRINT

After layer change

SET_PRINT_STATS_INFO CURRENT_LAYER={layer_num + 1}

Pause

M25

All other machine G-codes, and filament start/end G-codes, are blank or only contain comments.

In case it matters, I am using regular Orca (with profiles derived from its ones for the AD5X, not generic ones), not Orca-Flashforge.

[SunX99]

Adding my own experience here as I'm not getting the expected results.
AD5X 1.1.7 ZMOD installed 1.6.1
Settings as follows:
New Text Document.txt
With the most important part to this:
SAVE_ZMOD_DATA USE_TRASH_ON_PRINT=0
My understanding is this should not then use the 'trash/purge' area. (during color changes, but maybe still during priming?)

I printed the simple Orca Tolerance Test with the following changes:

• Separate the tolerance hex from the rest of the mold
• Color it separately (white in this case), then the base (black)
• Disable purge towers
• Print sequence - by object

The slicer (FF-Orca) results were listed as 2 filament changes. White object printed first, black object second.

Expected behavior-
Without the purge tower, and with the aforementioned USE_TRASH_ON_PRINT=0, I expected the printer to level the bed, maybe prime-purge?, lay down the test line and start with the white hex, complete the white hex, move (somewhere?) and do cut, retract, filament switch, engage. Resulting in a white to black color bleed in the base object. This was indented as part of the test.

What I got-

1. Start job
2. Go to approx start line location at front of bed
3. Kiss the plate way too close and lay down nothing for 80cm then stop
4. Almost abort the job(?)
5. Lower bed, goto trash, and start pooping 3x
6. Then remember you were supposed to do a bed level check. do that.
7. (maybe purge again- I don't recall)
8. Lay down your priming line like you're supposed to
9. Start the white hex
10. Finish white hex
11. Go to trash and swap filaments, but still purge out and poop
12. Now go start the black base object with a clean

So my confusion is:

1. why is it still doing the poop thing?! I thought the TRASH=0 was enough?
2. why are you kissing the bed and almost aborting? I've seen that behavior 2x now since installing zmod

I'll attach my gcode, and plate export for funizes if someone wants to try and ID what I'm doing wrong.
OrcaToleranceTest_Multicolor.zip

[SunX99]

Adding snapshots from the sensors. You can see around 15:04 where the lite purple line for the nozzle sensitivity spikes. That's when it buried it's head in the plate. Later it goes and does the regular bed leveling measurements like it's supposed to.
I know we didn't completely figure out what those other sensors are doing but we can see the dark purple of the filament cuts, and the orange of (what I'm guessing) filament pressure?

[Absolutely-D]

You talking about where it sorta "scrapes" the nozzle right at the front center of the build plate before levelling? I think that's intended, sort of as an attempt to wipe the nozzle so it can level cleanly (though clearly, fairly aggressively)

[ninjamida]

My changes achieve what you’re after. They’re not integrated into zMod itself at this stage, just linked to from the docs; you need to install them manually.

[SunX99]

Thanks ninjamida- I saw the work you had done and saw you were still doing some testing with it. I was under the impression I shouldn’t have received any poop purges with the flag set to 0. I’ll be doing some more testing later in the weekend. 

…

My changes achieve what you’re after. They’re not integrated into zMod itself at this stage, just linked to from the docs; you need to install them manually.

[ninjamida]

I’d say it’s pretty reliable at least for PLA and ABS at this point. So far it’s been good for PETG and TPU too for me, but I feel like those could use more testing. Whereas PLA and ABS I have total confidence in, and I feel comfortable to leave PLA or ABS prints unsupervised with this mod at this point.

[ghzserg]

I see from your screenshots that you don't have the latest updates installed.
I think you should install the latest updates.
And also install the modification from ninjamida

[ghzserg]

ghzserg/zmod_ff5x#6
The idea came up to implement plugins in zmod.
So that users could activate your work with a single command.
And if you needed to change the data in the plugin, you could always do so.

[ghzserg]

I created a plugin based on your work.
https://github.com/ghzserg/nopoop

If you plan to further develop the functionality,
I suggest cloning the repository above or creating a copy of the data at
https://github.com/ninjamida/nopoop
And I will include this plugin in zmod.

If you don't plan to further develop this plugin, I will include it in the mod at this address: https://github.com/ghzserg/nopoop

[ninjamida]

While I do have further ideas I'd "like" to add, realistically, at this stage it's doing everything I need it to, so I don't feel particular urgency to do much more work on it.

[ninjamida]

I'll take a look at this. I feel there are a few improvements I could make to it.

How easily can Bitbucket be supported (if it isn't already)? I usually host my code there.

[ghzserg]

I'd recommend using GitHub, as Bitbucket occasionally blocks IP addresses from Russia. It would be annoying if updates don't work for some users.

[ghzserg]

I've integrated your module into the plugin system.
https://github.com/ghzserg/nopoop/
I've also moved the functionality to the main zmod branch.
Currently, there are two overridden macros left in the module.

• _SBROS_TRASH_DAVIM
• _INSERT_PRUTOK_IFS
  I'd appreciate it if you could take a look.
  PS: The changes were made because your code didn't account for the different coordinates of the thread cutting location.

[Absolutely-D]

I don't know if what you've said here indicates that the plugin wouldn't work in its current state? but following 1.6.2 and the plugins update, with only nopoop enabled, my IFS appears not to change color while printing. I see
Unknown command:"T0"
Unknown command:"_NOPOOP"

at print start, and every time i'd expect a color change

I have since tried a factory firmware restore, and re-installing zmod via usb, then re-enabling the plugin, but it's still the same sadly. Am I doing something wrong?

Edit
Actually please ignore me, I was testing bambufy and still had the filament change gcode set in Orca - sorry

[N3rd7]

Hi Ninjamida and ghzserg, thank you so much for the development of the mod and pluggin, I can't wait to try! This post helped answering some of the questions I had (I'll remove the questions I added in the issue page), except for native display: in the zmod installation instructions, it is said to disable native screen before setting use_trash_on_print=0. Once this is done, can I turn the native display back on? Or will this setting only work if the AD5X is working screenless? Apologies if it's a dumb question...

[N3rd7]

Thank you! You say in your first post that you run it with guppy screen, right? I've used zmod and forge-x on my ad5m and never adventured to guppy screen, so I'll give that a try on the ad5x, just because this waste is so annoying! And I also want to print where colours just bleed into each other (kind of my own rainbow filament, just with different spools!)

[ninjamida]

Yeah, I do, although honestly - I barely actually use the screen anymore. My printer is in an enclosure (the kind that the entire damn printer goes inside of, not one that attaches to the printer and only encloses the build area) and I have a tablet next to it that's permanently open on Fluidd, so I just use that.

You can easily revert back to native screen if you don't like it, at any rate.

Side note, sounds like this may be of interest to you: https://www.printables.com/model/1384164-parametric-printable-custom-filaments-color-blend

[N3rd7]

Awesome, thank you for sharing this link, and developing the no poop, and answering questions, and and and!!! By the way, I see through this post that you improved the plugin overtime. Is the plugin posted by ghzserg your latest version? I wouldn't be sure where to find yours (or is the link in your first post updated to always send to your most recent version?)

[ghzserg]

The nopop plugin is currently supported in the zmod repository, but I'd be happy if @ninjamida hosted it and improved it. Currently, only two macros remain redefined. Everything else is integrated into the mod.

[N3rd7]

Setup complete! First test worked AMAZING! Thank you both for your help.

[ghzserg]

The plugin code is fully integrated into the main ZMOD code for the SAVE_ZMOD_DATA mode, use_trash_on_print=0
Installing the plugin downloads the modified filament.json and enables the SAVE_ZMOD_DATA parameter use_trash_on_print=0
Thanks @ninjamida

[ninjamida]

No worries!

If that's all the plugin does, and the main behavior is just part of the mod now, then maybe it's best to modify it so that it doesn't need the changed filament.json.

If I remember correctly, what matters is that filament_unload_before_cutting + filament_unload_after_cutting + nozzle_cleaning_length should add up to 145 total, as well as the slight filament_unload_after_drop. But also - I think most of this is true even without the mod, and the only thing I've really done with my filament.json is reduce waste a bit by pushing the unload before cutting a bit higher (and fixing the speed for TPU, as it was too fast before and causing errors). The rest is really just my setup / preferences more than anything else, I think.

It's really the macros, and the slicer settings, that are more important here, I think.

If I remember correctly - I can get back to the unmodified filament.json just by deleting my existing copy, right?

I wonder if there's a way that this can all be handed over to the slicer to control, similar to how most non-Flashforge multicolor printers work? (So the slicer can tell the printer "poop this much, then the rest goes into tower etc".) That's getting a bit beyond my capability though.

[ninjamida]

I remember now, the other thing I thought about doing was coding it so that when reloading the new filament, it calculates the distance it loads it such that the new filament load distance is calculated based on the old filament's unload amount. But I never got around to doing this, and instead just used hardcoded amounts.

So, I guess one (not both) of the following would be ideal:

1. When use_trash_on_print == 0, alter the reload distance based on the previous filament's unload_before_cutting. If it's less than 20, then increase how far the new filament is inserted; if it's more than 20, reduce how far the new filament is inserted.
2. When use_trash_on_print == 0, ignore filament.json's values for how far to unload the filament, and just use 20 before cut, 5 after cut, 3 after drop, and 120 nozzle clean length regardless of filament.json. (Actually, there's probably no need to ignore nozzle_clean_length; just ignore the others.)

I think if one of the above changes is made, then the custom filament.json will no longer be necessary. At that point, there's no need for a plugin at all - just change USE_TRASH_ON_PRINT to enable or disable it. Idea #1 would be better, but I don't know how easy it would be to do; I couldn't see an obvious way to get the previous filament's settings.

[N3rd7]

The no poop no purge works perfectly, but I do get a LOT of oozing with PETG. Oozing when loading new filament, or at color change during print. It continues oozing after nozzle wipe so there is a 1 inch dangle that falls on the bed or gets stuck on the print. I tried adjusting the unload length after cutting and after loading in filament.json, but it made no difference. The printer actually doesn't seem to retract any filament at those times (maybe I misunderstand what the unload parameters in the filament.json are, but I thought it was a slight retraction of the filament by rewinding the spool just a bit?)
Any suggestion to minimize oozing?

[ninjamida]

PETG is quite ooze-prone in general. You can reduce it by ensuring it's dry and using a lower printing temperature (make sure PETG's temperature configured in filament.json matches your print temperature, too, otherwise your printer will waste time heating up / cooling down when changing), but this mod doesn't work as well with PETG as it does with PLA or ABS.

Retraction doesn't always guarantee you won't get ooze, especially with filaments like PETG that tend to flow very freely when melted. It helps reduce it, but the main purpose of the retractions during a color change are to get the old filament out of the way so the new filament can be loaded.

Perhaps a useful addon to this mod would be to allow configuring it per-filament (ie: the user could set it so that the printer will still poop - though maybe poop less - for PETG, but will not poop for PLA or ABS). Alternatively, you could possibly just add turning USE_TRASH_ON_PRINT on or off to your start gcode based on the filament type.

[N3rd7]

I didn't even think of adjusting the temperature! I have been using high speed petg which has a lower print temp. I'll try reducing the temperature to 245 in filament.json and see how that goes. I'll update soon!

[cookwjc]

Thank you all for your work on this. I've implemented as described in documentation. There are a few behaviors I've observed that I haven't found documented and wanted to ask if my interpretation is correct or if I have a different issue that I haven't been able to unwind. If any of these items are true, it's probably worthwhile capturing some of this in FAQs with a relevant animal.

Observations in my set up:

1. Moving to GuppyScreen with DISPLAY_OFF disables the ability for Orca-FlashForge to communicate with the printer (LAN Only.)
   1a) Independent of whether we should want to continue using that version of Orca, I was still using it because it was otherwise functional before disabling the native screen.
   1b) I also recognize guidance exists to adjust the protocol in Orca to "Octo/Klipper" and that protocol control is not available in Orca-FF even though the 3mf delivery is still functional.
2. Because of the limitations above it seems, based on my experience, that guidance could be given to preemptively specify that the Orca-FF slicer cannot work with the DISPLAY_OFF command, which is a prerequisite for using nopoop.
3. Moving away from Orca-FF requires migration from printing the 3MF file (apparently sent by Orca-FF) to using raw GCODE uploaded through the slicer or web UI.

Tangentially, I have a bit more reading to do on this, but I've also observed sending prints via Orca result in the color selection being presented on the web page (presumably because Orca-FF handles that with custom code). As a result, I've seen an anomaly switching to the 4th color that results in the following error:

Filament change error: Error: CHANNEL 4 is out of range (max 3) Pausing print
Allowing it to resume continues printing with Color 3.

Finally, for others that may stumble upon this that have followed a similar path to me, I've previously disable the prime tower to save filament and have been reasonably successful since chamber pressure appears to get close to sufficient with the poops. Disabling the poops and not having the prime tower results in failed layer adhesion between changes. This appears to occur (in my case) as a result of insufficient filament load during the change after the filament cut and retraction. It appears that the first part of the next layer is 'printed' without filament since flow hasn't been reestablished. That can probably be tuned with retraction/load settings, but I also expect reenabling a small prime tower will resolve the issue as well.

[N3rd7]

This is a good call out to add a note about orca-flashforge, I had not realized it wouldn't work but it makes sense. I also use orca-FF sometimes, but this is only to use their AI model generator to be honest. The lack of some tools like "fix model" is a big no go for me.
For all model handling and prepping for print, I use the regular version orcaslicer, which is more updated and performant than the flashforge version. Have you tried it? If you are used to orcaFF, this will be a very easy transition for you. They also incorporate the printers and filament settings from orcaFF which kind of makes OrcaFF irrelevant.
In the regular orcaslicer, you can definitely send to print (rather than export g-code). Orcaslicer will also display the printer controls in the device tab when z-mod is installed (regardless of native screen on or off). Just adjust the printer settings in orcaslicer as per the zmod instructions.

[cookwjc]

Hi, yes. I've started transitioning over to native orca and that's been reasonably successful. The initial issues I was experiencing with the 'channel 4 out of range' error appears to have been a result of not having copied over my printer settings from orcaFF before trying to print (just an oversight as I was trying to understand the nopoop implementation.). Now that I've configured orca with equivalents from orcaFF, things appear to be working as expected. I'm still getting my head around the filament selection process tho. I'm not yet grasping the mapping of spool number to filament number yet, event after re-reading 3. How to Use the Color Selection Menu (Macro COLOR)

[ezeeri]

Hi. I have big problem with this configuration. I checked everythinh three times but still while printing in many colours in moment of changing filament my AD5X is not cutting filament before retraction. It finishes all the time with filament blocked in extruder. Can you help? It is very fast retraction but no move of carriage to the point that should cut the filament. I have 1.1.7 firmware

[ninjamida]

Hi. I have big problem with this configuration. I checked everythinh three times but still while printing in many colours in moment of changing filament my AD5X is not cutting filament before retraction. It finishes all the time with filament blocked in extruder. Can you help? It is very fast retraction but no move of carriage to the point that should cut the filament. I have 1.1.7 firmware

To confirm - your printer is not moving to the cutter at all (rather than trying but failing to cut)?

Have you installed any config files / etc from this discussion thread? If so, remove them - any functionality I discussed in this discussion, has long since been integrated into zMod itself. You just need to install the "nopoop" plugin and make the changes to your slicer settings.

[ezeeri]

Yes. I installed all of your configs and it looks like it was a problem (still testing). Now printer installed 1.2.3 firmware and it is not supported by ZMOD in this moment so I'm downgrading. It was not moving to cut place at all. just at moment for change filament it was doing very fast retraction (i think too fast) of few centimeters and after it filament was unable unload from extruder. Thank you

[ezeeri]

It is working. Thank you very much. Now i need to calibrate parameters, Its printing too fast and print quality needs to be mastered.