8 Minute Bonkers Benchy on MK3.5
I modified the gcode for the 8 min Bonkers Benchy print from the MK4s to run on my MK3.5. I had to up the temperature all the way to 270°C to increase flow but there was still underextrusion.
In terms of mods, I've upgraded all of the axes to MK3.9/4 style hardware, changed the nozzle to a Microswiss CM2, and installed the HighV Part Cooling Nozzle. See picture at bottom.
RE: 8 Minute Bonkers Benchy on MK3.5
great work, can you post th g-code file you modded please??
RE:
Sure here you go!
RE: 8 Minute Bonkers Benchy on MK3.5
Sure here you go!
Oh it seems like I can't upload a .gcode file in the reply. Here's a link instead: https://drive.google.com/file/d/1veRtGRvU4t4egoqTr2CvvmQrjYxUcnI7/view?usp=drive_link
RE: 8 Minute Bonkers Benchy on MK3.5
Hey, thanks for this. It was fun to watch! I went ahead and ran it on my non-HF nozzle (too lazy to swap) and it worked great. Print is surprisingly good, even slightly better than your picture (I used the same filament). I printed on my MK3.9S. Also interesting is that I timed it and got 8:17 which exactly matches the estimate.
Here's where it gets weird. I copied your profile exactly so I could use as a quick draft mode and also to see which knobs have the biggest effect. First slice was ~10 minutes. Crawling through your gcode it looks like you used a couple height modifiers? Anyway I had to to match your physical output: infill styles, line directions, etc. Now my slice shows 9:20. I have been going down this rabbit hole to identify the differences. Most of that extra minute is tiny deltas on most extrusions that slowly add up over the whole print. With the EXACT same line width, acceleration, starting velocity, distance the time and peak velocity don't quite match, e.g. 28 ms vs 31 ms. What am I missing? One possibility is that the print time estimator uses an inertial model and has different masses for the 3.5 vs 3.9 toolheads, which are different. But I don't think it does that.
RE: 8 Minute Bonkers Benchy on MK3.5
Glad you had fun running the file! When I made it all I did was change go through the identifiers in the comments of the gcode and change them from Mk4 to Mk3.5. I also tried to copy the profiles from the gcode to use as a draft mode (which you can actually do by going file>>import>>import config). I found that a lot of the time spend added by the slicer versus the file was in machine and filament limits and not actually in the print profile. I ultimately gave up because my stock E3D v6 nozzle with a Microswiss CM2 nozzle cannot achieve the flowrates for that kind of print speed and the accelerations made my machine sound like it was going to tear the machine apart.
RE: 8 Minute Bonkers Benchy on MK3.5
I just mentioned it in another topic.
The MK3.5 seems to have reduced extruder steps compared to the MK3.
Previously there were 280, but in the MK3.5 there are only 266.
Although both use the same extruder.
RE: 8 Minute Bonkers Benchy on MK3.5
I also tried to copy the profiles from the gcode to use as a draft mode (which you can actually do by going file>>import>>import config).
Ha! I forgot about that feature. Well, I have it now. Just need to solve the mystery on the time estimates. Since the original was MK4 then I have the same machine but with 1.8 steppers which can technically be faster (which is not what I'm after). I will hunt down the original to see if that highlights anything.