I use thermal grease, door with a fan (of course 
) and 6mm nozzle switch retraction and it have not failed on me since I installed the fan.
Also maybe counterintuitive but going faster is better. More flow cools parts.
1 Like
Im down at 1mm and 30mm/sec and not having issues. the defaults are WAY to high.
I changed the fan for another axial with a little more airflow and have not encountered an issue on a 8h print with a lot of tool change and retractions
1 Like
Iāve been holding off using my new Dual Head, based upon the issues everyone has been having. Iām now ready to give it a go, but I wondered if there was a way to ignore the first-time calibration routine, in order to update the Dual Head to the latest firmware?
Seemed prudent to get the latest firmware on board before calibrating.
Is this possible?
Just accept the defaults. When it asks you to compare the lines.
you only need to run them once, and you dont need to print the test squares.
then manually do them with cura or prusa.
OK, thanks!
just a little update.
gone through 1kg of carbon fibre filled filament (PA and PETG CF) on left 0.4mm nozzle and about half a kg of pva and pla through the right nozzle and not a single clog since I changed the retract at extruder change set to 2mm.
without the door mod or additional thermal grease.
the hotend that I modded with grease is still sitting on my shelf.
All printed with closed door and 36Ā°C inside the enclosure.
Also got almost no oozing of the inactive nozzle once the filament was properly dry.
When changing filament I was cautious to set the unloading temp as low as possible to ensure no molten filament is dragged up into the metal tube of the hotend.
canāt believe that there is so much variance in the DEX that some work and some donāt?
2 Likes
Hi @elopan, could you please recap all your most relevant printing settings?
Thanks!
I basically took the Luban material profiles as a baseline and set the retract at extruder change for both materials that are in use to 2mm.
As I said, I based my settings on the experiences shared by various dualextruder users from the ultimaker raise3D etc. forums.
I also increased the retract speed at extruderchange a little bit to about 35mm/s, depending on how stiff the filament is, you can also go faster. For PA I set it to 65mm/s, but I dont think that would even be necessary, as it is pulled up only once.
Standbytemp as Luban suggests, about 150Ā°C for most materials.
(I read that higher would be better, as the cooldown and heat up of the filament causes shrinking and expanding of the filament, which can cause oozing, but I didnt touch that and kept it at the temp that Luban had as default.)
Didnāt touch the rest of the settings regarding extruder change.
Setting the retract distance higher didnāt help with oozing of the inactive nozzle. You canāt pull out the molten filament from the nozzle tip and if you retract more, you only create an air bubble that expands as it heats up and pushes more material out. So 2mm is about best as I found out, at 3mm there was already significantly more oozing.
If oozing causes problems, the fan speed setting of the active nozzle can be relevant as it cools down the inactive nozzle faster.
With Nylon I had to set it to 0 to get good layer adhesion. Set at 10% already improved the oozing of the inactive nozzle, but at the cost of worse layeradhesion of the nylon. Just to keep that in mind.
Dry filament is a must. I am printing directly out of a Sunlu filament dryer. The Snapmaker Breakaway PLA was oozing like hell at the beginning and completely stopped once it was dry. Same with the Nylon.
Snapmaker PLA that was supplied with the Artisan worked from the start without oozing, same with extrudr xPETG CF, no oozing at all.
Can upload some of the profiles later, I am at work atm.
1 Like
@elopan Thanks for your support!
1 Like
Hi folks, wanted to give a bit of an update on my case.
TL;DR: Problem only occurs for me on dual color prints, and I am in the process of testing a new door design provided by Snapmaker support. Obviously they are designing a repair kit for affected users.
Long story:
I contacted support, and as always when I talk to them, they are very helpful! They sent me a mail very similar to that of @jlropes - but they also attached an alternative door STL file for me to try out - unfortunately they did not give me green light yet to share it here, will do as soon as they allow me to. Hereās the Mailtext:
At present, it is suspected that it has something to do with the heat dissipation of the module, which is also coupled with the working environment of the Dual Extrusion Module.
1ć If the left hot end is changed to the right extruder, is the printing on the right head normal? Use this to determine if the problem is in the hot end or in the extruder
2ćPlease observe whether the two fans facing the door are turned on.
3ćIf you are using an enclosure, please open the enclosure door when printing PLA material and print.
4ćIf you print with the module-door open, whether it will also be clogged ļ¼Once you print success, could print attached file? Replacing the metal door plate. If there is a visible improvement, we will follow up with a complete kit for you.
So to do AI 1, I sliced a new model (I did not want to re-print the old one after I finally had a successful print) and printed it only on the right nozzle (without changing the hotends), to make sure the error was there still. To my utter surprise, after 5+ hours I had a good finished print there! So I infered that a single nozzle does not put enough heat into the whole system to cause filament problems. So I made the same model artificially dual-color, and I tried to roughly stay in the same ratio between color 1 (left nozzle, only a little) and color 2 (right nozzle, a lot to print) as with the model that failed previously. This time I also exchanged the hotends. Immediate āsuccessā after less than 20 minutes - right nozzle fails to extrude. Post mortem shows filament broadening over more than 30 mm. So in my case it seems to be really related to the right side, not to the hotend module. AI 2 I can say that both fans are always spinning.
AI 3 & 4 I will skip and directly try the door by Snapmaker, as it looks like a good design to me. I printed it, and will do the 2 color print again - but only tomorrow, too late now. Will keep you posted!
A few pictures:
The new door design - printed in clear PETG (did not come out too well, but I did not care, is functional enoughā¦). You can see how they channel air directly to the lower part of the heat sinks.
The broadened filament:
4 Likes
@Hauke, Have you seen any success from this replacement door? I have been unable to complete a single print despite many settings changes. Every print for me ends with clogging, expanded filament, and grinding. Thanks!
Hi all,
so finally got around trying the new door design, and it was a full success! The model that failed after less than 20 mins before, now went through in 7 hours. The filament does not show any hint of broadening/heat creep, neither on left nor on right nozzle. Already informed support about the success, now waiting patiently for what they make from it - my assumption is that theyāll release a new front door and send this to those affected. I am curious how many this might be - would you mind answering this poll:
Side remark: Since the new firmware now issues a warning if your standard retraction exceeds 2mm, I lowered it to 2mm. I will not leave it there - my black filament oozed like hell! I personally believe that retraction settings that do not make molten filament get into the cold zone will be of no problem - from looking at the size of the heater block Iād assume that anything up to 10 mm would be safeā¦ but thatās currently just an opinion, will establish this over time by testing.
Oh, and btw., Support said as soon as I verified that the door is good, I can share it - find it here:
Dual extrusion front plate.STL (220.1 KB)
A big THANKS to Snapmaker support for taking care and providing this file!
6 Likes
Do we all have to send in help tickets for the kit they mentioned?
Howās it working a few days later? Iām in the process of moving and wonāt be able to print it for a month or so.
I had nothing to print since my last report (I generally do not print too often - only functional parts when I need them), so I cannot tell, but the test result was so obviously different I consider my problem āsolvedā.
Snapmaker support reached out to me and asked for a shipping address for the repair kit - so they seem to have it ready at hand. Will report as soon as I received it.
I guess soā¦ looking at the poll I linked in above and on a similar poll on Facebook I added to the thread, the problem to my surprise does not seem to be universal - there seem to be a lot of working units out in the wild, so they canāt just send out a repair to just everyone. I really wonder what exactly is different in those that fail. I ordered on day one and received a unit of the very first batch - maybe they already changed something right after that?
I like that Snapmaker sends me an āupdateā without much ado (thanks to support for that!), taking responsibility for their mistake, but I guess Iād like it even more if they would be a bit more open about what exactly went wrongā¦
1 Like
Good to know. I sent them a request. Guess weāll see what kind of troubleshooting they want me to do before they send me the fix.
| Hauke Regular
May 30 |
- | - |
Mads0100:
Howās it working a few days later?
I had nothing to print since my last report (I generally do not print too often - only functional parts when I need them), so I cannot tell, but the test result was so obviously different I consider my problem āsolvedā.
Snapmaker support reached out to me and asked for a shipping address for the repair kit - so they seem to have it ready at hand. Will report as soon as I received it.
Do we all have to send in help tickets for the kit they mentioned?
I guess soā¦ looking at the poll I linked in above and on a similar poll on Facebook I added to the thread, the problem to my surprise does not seem to be universal - there seem to be a lot of working units out in the wild, so they canāt just send out a repair to just everyone. I really wonder what exactly is different in those that fail. I ordered on day one and received a unit of the very first batch - maybe they already changed something right after that?
I like that Snapmaker sends me an āupdateā without much ado (thanks to support for that!), taking responsibility for their mistake, but I guess Iād like it even more if they would be a bit more open about what exactly went wrongā¦
1 Like
I think itās endemic. Youāre just not getting the results. It would be nice if the company just emailed everyone who bought it.
Itās not like our case is special. Iām surprised holes in the door make that big of a difference.
1 Like
Iām not. Snapmaker decided to design the airflow to suck air out of the housing instead of blowing cold air onto the heat sink. Which means that where the air that blows across the heat sinks come from is less predictable. It is not unusual, it is e.g. a typical way to vent rooms, but for the DX it might not have been the best choice. My assumption is that it already makes a difference how the cables in the housing are sitting, and that may vary with each DX head, which would explain the spread in experience. Admittedly, that is a rather weak assumption, but if Iām right, some people that at some point disassemble their DX e.g. to replace the filament detection switch, might be in for a nasty surprise because their previously nicely working DX suddenly develops heat creep. The holes in the front door open up a very distinct path for cool air to flow in, so I can immediatly see why it helps.
So you may ask why did they design it this way? Again, I can only make assumptions, but my guess is that they ālearnedā from the first version of the SX head, where they used air that was blown across the heat sink for print cooling, which was a failure. So they tried to avoid that they blow warm air on the heat sink or that warm air is channeld in an unpredictable way, and their idea was to suck it out and have ventilation slits at the top of the housing, which should be as cool as it gets. But since you cannot get the head airtight in other places, itās a bit of a gamble. I would have expected that a flawed design would not have made it past the prototype phase, but if the problem is not universal and not easy to reproduce, it might have evaded themā¦
Long story short: Such things happen, no bad feelings from my side, as long as they take responsibility and replace/repair it, which in my case they do.
2 Likes
I think I might have been too optimistic.
both nozzles clogged 2h apart during a long print now.
havent taken the hotend apart yet to see what the cause was.
I have to add that I printed another kg of nylon before that problem occured and did about 10 filament swaps in the last two days as I wanted to try a few material combinations (breakaway, standard pla, petg CF and PLA CF) but both clogging at about the same time now looks a bit strange to me. Will definitely print the new door now as soon as I cleaned the hotends.
Is snapmakers official response the .stl file for the new diy door or will there be an official solution with new metal door? Have they told you anything in that regard?
And did you get any information on which magnet to use on the printed door?
1 Like
Just a quick follow-up from my sideā¦ In the last few days I have been printing few objects.
With the same setup as before there have been no clogs at all and the only difference was that I used PrusaSlicer for slicing and not Luban.
Are you all using Luban for slicing when clogging or others?