S3D (and other slicers)
-
I tried Simplify3D when I was 'nowt but a lad' and gave up on it after 2 weeks, as Dual extrusion did not work then.
No I'm a 'grown up' I have just tried it again, as Matter Hackers slicer just would not slice Julian_Vase_002, even the basic model.- "Too many polygons"!!
S3D did, even the detailed version, in no time at all. The interface may seem strange at first but well worth the pain in learning.
BUT my question really is: for Chimera dual extrusion, with two hot ends, does implementing firmware retraction really make such a difference,
or is it only for Diamond/Cyclops heads that it is really needed?
- "Too many polygons"!!
-
I gave up on trying to use my Chimera setup and instead use the Cyclops when I want to do dual extrusion prints. I was having collisions far too often no matter how much I tried to balance the two nozzles out. I realize I could probably have fixed this if I tried for long enough, in the end just wasn't worth it. The need for stuff like ooze shields just made it too much effort for too little reward.
As to S3D, they better update their software soon or I'll be looking elsewhere. Stuff like trying to do top layers over the middle of a hole from a honeycomb infill and having all the plastic fall in. Cura doesn't suffer these issues. Not placing outlines around support but instead through the middle of it….the list is long.
For now I'm still using S3D, it does work with dual extrusion prints with a bit of work. Really not something I do often even though I specifically invested in the ability to do so..go figure.
Jeff
-
I tried Simplify3D when I was 'nowt but a lad' and gave up on it after 2 weeks, as Dual extrusion did not work then.
No I'm a 'grown up' I have just tried it again, as Matter Hackers slicer just would not slice Julian_Vase_002, even the basic model.- "Too many polygons"!!
S3D did, even the detailed version, in no time at all. The interface may seem strange at first but well worth the pain in learning.
BUT my question really is: for Chimera dual extrusion, with two hot ends, does implementing firmware retraction really make such a difference,
or is it only for Diamond/Cyclops heads that it is really needed?
If you have a mixing type hot end, i.e. 2 or more inputs and a single nozzle then you need to retract all filaments by the same amount, otherwise all that happens is that filament is drawn from one or more of the unused inputs, rather than from the nozzle tip. This is how firmware retraction works with the Duet. So if you have separate nozzles, then you should use "normal" retraction, but if you have one nozzle with multiple inputs, then you should use firmware retraction. HTH
BTW, I gave up on S3D because trying to get it to support 3 extruders was a complete PITA. I did ask them (S3D) what was the point of printing a priming tower, then changing the tool before resuming the print but never got an answer. Completely ar** about face way to do it.
Ref Julia vase - you might be interested in my scaled up version, sliced in Slic3r but then post processed to fade from red to yellow to clear
https://www.youtube.com/watch?v=ZLwHOcH_UFkIan
- "Too many polygons"!!
-
I use s3d firmware retraction by using post processing scripts. Works fine, with one caveat. If you're using firmware retraction you have to hardcode the z-hop in the slicer, and have it set to 0 in the FW retract configuration. Otherwise if it does a retract on layer change the next layer prints too low. This is due to an "order of operations" issue with how s3d (and most slicers other than slic3r) operate when a retraction coincides with a layer change.
(I have further info and documentation on this, if needed)
Thankfully hardcoding z-hop isn't a big deal, as that's not something that requires a lot of tuning mid-print.
Edit to say that I think this z-hop issue and the order on layer change is likely why they aren't supporting modern firmware retracts with z-hop, going by a conversation I had with support.
-
Here's the further detail - I went through a bunch of slicers and generated gcode of the same model, and found any instances of retraction on layer change. Here's the "order of operations" for each when this happens:
You can see that slic3r's the odd ball out, and it's no coincidence that slic3r is the only one that supports modern FW retracts. If you substitute (through post processing) the FW retracts into a slicer like s3d, and include Z-Hop, the "unhop" will come after you're already at the new layer z-height, making your next print move start at (new layer height) - (FW z-hop height). Probably not what you want.
As I said, works fine in s3d if you use the slicer settings for z-hop height, and do the rest of your FW retracts with post-processing scripts.
-
If you include Z hop in your firmware retraction parameters, retract on layer change should work ok in all slicers now. This is one of the things I fixed in firmware 1.17.
-
Ah, I didn't catch that in 1.17. I'll test it out!
-
S3D and firmware retraction.
I thought I read somewhere in this forum that 1.17c now supports the S3D required M codes without conversion to corresponding G codes.
Firmware retraction does indeed work but how do I suppress the "unsupported M102 messages".
Also, as a relative newbie to dual extrusion, how do I eliminate the deadly pause over the print when once colour finishes and before it moves to the prime/wipe tower? I have the Tpostn.g files set with M116 Pn. -
S3D and firmware retraction.
I thought I read somewhere in this forum that 1.17c now supports the S3D required M codes without conversion to corresponding G codes.
Firmware retraction does indeed work but how do I suppress the "unsupported M102 messages".
Also, as a relative newbie to dual extrusion, how do I eliminate the deadly pause over the print when once colour finishes and before it moves to the prime/wipe tower? I have the Tpostn.g files set with M116 Pn.Add the following line in yoour post processing areaa atthe bottom of the Scripts tab in Process settings
{STRIP "M102"}
That will remove them at source
HTH
Doug
-
S3D and firmware retraction.
I thought I read somewhere in this forum that 1.17c now supports the S3D required M codes without conversion to corresponding G codes.
Firmware retraction does indeed work but how do I suppress the "unsupported M102 messages".I don't know why S3D generates M102 or what it expects the printer to do. I guess I could make the firmware ignore it.
Also, as a relative newbie to dual extrusion, how do I eliminate the deadly pause over the print when once colour finishes and before it moves to the prime/wipe tower? I have the Tpostn.g files set with M116 Pn.
Set up the tpre#.g files to move the head away from the print.
-
Before David made these recent changes to the firmware, I did a bit of a write up on how to get S3D to do firmware retraction. It's here https://www.duet3d.com/forum/thread.php?id=396.
-
I'm messing with Cura 2.4 beta 2 at the moment, wow I must say it has sooo many options. I notice it also has adjustable acceleration. I think I'm going to dump S3D for a while, kind of getting sick of the bugs.
-
I could easily be wrong but I don't think the acceleration options in Cura are compatible with the current Duet firmware.
Also, I don't think it supports firmware retraction for all printers.
However, I am using the Cura master branch at the moment and getting pretty good results.
I have S3D also, I hate it!
-
@(In)Sanity:
I'm messing with Cura 2.4 beta 2 at the moment, wow I must say it has sooo many options. I notice it also has adjustable acceleration. I think I'm going to dump S3D for a while, kind of getting sick of the bugs.
Have you figured out where to enable Firmware retraction and change between Absolute and relative extrusion?
-
I could easily be wrong but I don't think the acceleration options in Cura are compatible with the current Duet firmware.
Also, I don't think it supports firmware retraction for all printers.
However, I am using the Cura master branch at the moment and getting pretty good results.
I have S3D also, I hate it!
Does firmware retraction really add that much in terms of quality ? Doesn't it basically just retract if not printing ? Seams like the delay the slicer would add for the retracts shouldn't be noticed with the Duet WiFi speed.
Trying to solve this issue at the moment with Any slicer at all, as you can see it's trying to print a top layer over top of a hole with nothing to connect to. This just makes a complete mess.
Know any tricks in Cura to not have this happen ? So far it appears to have the same flaw S3D does in this area.
Thanks,
Jeff
-
Well, quite some time back, I wrote an email to the S3D technical support people complaining about how it failed miserably because it didn't extend skin far enough under perimeters that are placed above. It's essentially the same problem as you have. At that time, the only slicer that seemed to get this right was slic3r. I haven't used that for quite a while but it's probably worth a try.
-
At least Cura is open source and so we have the possibility of fixing stuff we don't like. I have been making some trivial contributions to it. It has occurred to me that a fix to this problem would be really good so I shall go back to thinking about a solution.
-
Yes for mixing type hotends firmware retraction is an absolute must
-
A Search on there Web page gives no results for firmware retraction therefore it is a no go for me with Cura
-
Well, quite some time back, I wrote an email to the S3D technical support people complaining about how it failed miserably because it didn't extend skin far enough under perimeters that are placed above. It's essentially the same problem as you have. At that time, the only slicer that seemed to get this right was slic3r. I haven't used that for quite a while but it's probably worth a try.
I actually tried the same scenario/part with slic3r and it failed in the same way. I've wasted more filament then I care to mention or ended up with terrible looking parts because of this exact issue. I do lots of ABS work and I'm not about to print too dense and have stuff try to curl or delaminate. I'm going to mess with the beta version of Cura for a bit and see what I can do. The S3D folks just don't seem to care anymore, if they do they are failing to relay that to the customers very well.
Jeff
