jltx
Posts: 439
Joined: Jan 30, 2017 3:28 am
Contact: Send private message

Still zero successful prints

Oct 27, 2018 1:52 pm

I can get a few layers and color changes but eventually it always skips, stalls, jams, or catches fire. Is there a troubleshooting guide (not Prusa’s which is useless)?
You want parts? I got MK3 parts https://www.thingiverse.com/jltx/designs

paul.m27
Posts: 436
Joined: Sep 23, 2017 5:47 pm
Answers: 2
Contact: Send private message

Re: Still zero successful prints

Oct 27, 2018 3:28 pm

What filament are you using? What slic3r settings for temp and advanced settings (load/unload/ram)?

For me, the keys to success are:
- debug one thing at a time
- patience

Pick your most common fail and describe it here. We'll work on that.

General steps:
=> go through this guide carefully: troubleshooting.

Align everything multiple times. No room for slop. Consider putting on a fresh nozzle (completely anecdotal, but I had better luck once I replaced the nozzle I've been using for 9 months).

Look in from the right side of the extruder (with the door open). You'll see one of the bondtech gears and the little white PTFE tube below it. Make very sure they are aligned, that the top of the PTFE tube looks a bit like a funnel, that it has clean/well-formed edges, and it looks like it will catch filament coming in from above and guide it into the hot end. For me, this is one of the primary fail points. If it doesn't look good, open up your extruder and replace it (the kit came with a spare).

Make sure you have relatively low friction on your filament path.

Then try printing with two filaments, a part that has multiple colors on the first layer, preferably. Don't load them into filament 1 and filament 2 slots, put them in 1 and 3. Start printing. At *each* filament swap, after the selector moves over to the new filament, push the just unloaded filament out of the MMU2 so you can see the tip. Take a picture! Do this for the first 10 or so unloads of each filament.

If you have pretty well formed tips, like this:
IMG_2199.JPG
IMG_2199.JPG (190.66 KiB) Viewed 2416 times
and you are still having jams or other issues, come back here and explain the specifics. Your tip looks good and the issue is a mechanical one in the MMU2 or printer, not too hard to troubleshoot.

If your tips look more like this (or worse, or tangled):
IMG_2196.JPG
IMG_2196.JPG (312.49 KiB) Viewed 2416 times
then expect jams/issues. That tip is disaster waiting to happen. It can (and will):
  • cause unload issues because it keeps the finda triggered even after the MMU2 gears have unloaded the filament as far as they can
  • cause load issues because the tip gets caught and prevents the filament moving forward (at selector, bondtech, ptfe entrances)
  • cause skipped layers because the tip catches in the extruder. The printer thinks filament is loaded, but the bondtech is just grinding away not moving the filament
  • cause random issues as the strings get sheared off in the selector and start to build up over time (I've blown out dozens of little strings after some fails)
To control those strings, no perfect method, but some ideas:
  • try different filament. Prusament is supposed to be a good answer.
  • lower your temps a bit
  • try to increase your ram time (filament->advanced in slic3r)
  • try to change your unload speed
The important part here is to not just change a bunch of setting, watch another print fail, get frustrated and give up. (easy to do, I've been there). Instead, change one setting or group of settings, try a print, and look at your tips! Take pictures! Compare the pictures with the previous settings. See if the strings are longer/shorter/thicker/thinner. Learn how that change helped/hurt the tips. If it helped, try a bit further in the same direction. If it hurt, try the opposite direction. Then layer other changes on until you get clean tips.

I'm still working through this, but I do have some pretty successful prints. Best print so far, 12 hours, 0.15, 4 filament, Inland PLA:
56212103399__7BD76157-A763-467C-A4EF-F4822EFFEDD6.JPG
56212103399__7BD76157-A763-467C-A4EF-F4822EFFEDD6.JPG (288.75 KiB) Viewed 2416 times

jltx
Posts: 439
Joined: Jan 30, 2017 3:28 am
Contact: Send private message

Re: Still zero successful prints

Oct 27, 2018 3:49 pm

Thanks for the link to Peter's troubleshooting as I was looking for that. Unfortunately that is too basic for me. I'm well past that. Your idea to not use adjacent slots is very good (should have thought of that myself). The tips are for sure (most of) my problem. I am using Prusa PLA for initial testing and only 2 colors at a time. I tried filamentum but that jammed immediately. I agree only change one variable at a time.

The main problem I have is the debug iteration is way too long and most of the path is hidden (clear bowden would have been better). For example, I want to change ramming settings to see if it helps. But I have to re-slice, calibrate loading, wait for bed to heat, other fiddling, print, fail, tear down machine to clear filament, beat my kids, start over. Is there a faster method? If the ramming was on the menu you could test right at the hot printer, over and over, with one filament, until the tip looks good.
You want parts? I got MK3 parts https://www.thingiverse.com/jltx/designs

paul.m27
Posts: 436
Joined: Sep 23, 2017 5:47 pm
Answers: 2
Contact: Send private message

Re: Still zero successful prints

Oct 27, 2018 6:47 pm

jltx wrote:
Oct 27, 2018 3:49 pm
The main problem I have is the debug iteration is way too long and most of the path is hidden (clear bowden would have been better). For example, I want to change ramming settings to see if it helps. But I have to re-slice, calibrate loading, wait for bed to heat, other fiddling, print, fail, tear down machine to clear filament, beat my kids, start over. Is there a faster method? If the ramming was on the menu you could test right at the hot printer, over and over, with one filament, until the tip looks good.
Agreed. I'm working on that. I see several stages:
  • a script that pulls the filament unload (ram) section of a gcode file out, and packs many load/unloads into a single gcode file. Ideally I'll unfold the ram into a straight line, and have hundreds of straight lines across the plate, each with a different setting. Ideally I can procedurally generate the ram settings, rather than just grepping them all from Slic3r gcode.
  • above script, combined with hacked MMU2 firmware to
    • move the selector away from the unloaded filament
    • use the mmu2 drive to stick the filament out to be visible
    • pull it back in
    • load the filament again
    • lather/rinse/repeat
  • much later: drive this all from octopi to trigger a photo of each filament unload. Then I can launch this, let it run for an hour, and have 100 different settings to examine the tips for!
I'm working on the script now. Firmware will be later. Pi much later.

User avatar
PJR
Forum moderator
Forum moderator
Posts: 5458
Joined: Jun 01, 2015 6:37 am
Answers: 27
Location: Kamnik, Slovenija
Contact: Send private message Website

Re: Still zero successful prints

Oct 27, 2018 7:00 pm

jltx wrote:
Oct 27, 2018 3:49 pm
Thanks for the link to Peter's troubleshooting as I was looking for that. Unfortunately that is too basic for me.
Sorry :(

Some PR support technicians have offered to help; I will give them a shout. It will be interesting to see how it works out.

Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage or loss.

ted.t5
Posts: 65
Joined: May 31, 2018 2:50 pm
Contact: Send private message

Re: Still zero successful prints

Oct 27, 2018 8:56 pm

paul.m27 wrote:
Oct 27, 2018 6:47 pm
  • a script that pulls the filament unload (ram) section of a gcode file out, and packs many load/unloads into a single gcode file. Ideally I'll unfold the ram into a straight line, and have hundreds of straight lines across the plate, each with a different setting. Ideally I can procedurally generate the ram settings, rather than just grepping them all from Slic3r gcode.

This would be really nice for us MMU2 noobs that are pulling our hair out.

jltx
Posts: 439
Joined: Jan 30, 2017 3:28 am
Contact: Send private message

Re: Still zero successful prints

Oct 28, 2018 12:38 am

PJR wrote:
Oct 27, 2018 7:00 pm
jltx wrote:
Oct 27, 2018 3:49 pm
Thanks for the link to Peter's troubleshooting as I was looking for that. Unfortunately that is too basic for me.
Sorry :(

Some PR support technicians have offered to help; I will give them a shout. It will be interesting to see how it works out.

Peter
No worries. Good stuff, I'm just beyond that now.
You want parts? I got MK3 parts https://www.thingiverse.com/jltx/designs

jltx
Posts: 439
Joined: Jan 30, 2017 3:28 am
Contact: Send private message

Re: Still zero successful prints

Oct 28, 2018 12:39 am

paul.m27 wrote:
Oct 27, 2018 6:47 pm
jltx wrote:
Oct 27, 2018 3:49 pm
The main problem I have is the debug iteration is way too long and most of the path is hidden (clear bowden would have been better). For example, I want to change ramming settings to see if it helps. But I have to re-slice, calibrate loading, wait for bed to heat, other fiddling, print, fail, tear down machine to clear filament, beat my kids, start over. Is there a faster method? If the ramming was on the menu you could test right at the hot printer, over and over, with one filament, until the tip looks good.
Agreed. I'm working on that. I see several stages:
  • a script that pulls the filament unload (ram) section of a gcode file out, and packs many load/unloads into a single gcode file. Ideally I'll unfold the ram into a straight line, and have hundreds of straight lines across the plate, each with a different setting. Ideally I can procedurally generate the ram settings, rather than just grepping them all from Slic3r gcode.
  • above script, combined with hacked MMU2 firmware to
    • move the selector away from the unloaded filament
    • use the mmu2 drive to stick the filament out to be visible
    • pull it back in
    • load the filament again
    • lather/rinse/repeat
  • much later: drive this all from octopi to trigger a photo of each filament unload. Then I can launch this, let it run for an hour, and have 100 different settings to examine the tips for!
I'm working on the script now. Firmware will be later. Pi much later.
Sounds great!
You want parts? I got MK3 parts https://www.thingiverse.com/jltx/designs

flaviu.i
Posts: 142
Joined: Feb 04, 2018 8:53 pm
Location: Bavaria
Contact: Send private message

Re: Still zero successful prints

Oct 28, 2018 9:02 am

I think a lot of people are now on the same success level as I am which is prints with around 100-150 color changes (around 5 hours print) works fine.

But going up to over 300 color changes (12+ hours of print) produces at least one mistake in the part.

From here on doing the next step in the right direction with the settings is very time and filament consuming :cry:

jltx
Posts: 439
Joined: Jan 30, 2017 3:28 am
Contact: Send private message

Re: Still zero successful prints

Oct 28, 2018 2:28 pm

I would like to point out that I believe my only problem is the stringy tips and that there is a razor blade built in to solve this but it is disabled. Putting all this debug burden on the users is poor choice. Just slice off those tips. What am I missing?
You want parts? I got MK3 parts https://www.thingiverse.com/jltx/designs

Return to “Assembly and first prints troubleshooting”

Who is online

Users browsing this forum: No registered users and 1 guest