3d Uber-Printer Build (Voron 2.4)

[Netpackrat]

The carbon filter is pretty bad-ass too. When I open the door of the printer, I can't smell the ABS fumes at all.

[HTRN]

The carbon filter is pretty bad-ass too. When I open the door of the printer, I can't smell the ABS fumes at all.

Then this might interest you:

https://blog.prusa3d.com/were-launching ... use_75919/

Prusa is introducing a tungsten filled PETG who's main application seems to be medical: radiation shielding. It seriously not cheap at $200+ a kg.

Currently printing. I still have a lot of tuning to do but it's off to a good start, with relatively few problems in the setup compared to how it could have gone. There's a seed starter for gardening that I downloaded off Printables and made for my wife to use in her mini greenhouse... The basin took over 16 hours to print on the Prusa, and total with all the pots was over 30 hours of print time. I sliced it for the Voron at the default settings, which shows it as being a 6 hour, 10 minute print.

That's not apples to apples since I am using the 0.6mm nozzle in the Voron, as opposed to the 0.4mm in the Prusa. I could have switched the Prusa to a 0.6 also and saved a lot of print time, but not over 10 hours worth. Plus I haven't run Input shaping (basically a resonance calibration using an accelerometer mounted to the print head) yet, or done anything else to tune the Voron for higher speeds.

How fast are you actually printing? 150mm/sec? Faster?

[Netpackrat]

Then this might interest you:

https://blog.prusa3d.com/were-launching ... use_75919/

Prusa is introducing a tungsten filled PETG who's main application seems to be medical: radiation shielding. It seriously not cheap at $200+ a kg.

That's kind of cool. Not sure what I would use it for but I can probably think of something. I'll bet that stuff's hell on a brass nozzle. I actually ordered a tungsten carbide nozzle last night for the Prusa because I have a few things I want to print in PC-CF.

How fast are you actually printing? 150mm/sec? Faster?

Perimeters are set to 100mm/s in the slicer currently and up to 125mm/s on infill (varies by type). There's more to be had there once I run input shaper but the slicer defaults are pretty conservative. Max volumetric flow speed on the extruder is set to 11mm^3/s for ABS, which IIRC is what the Revo is rated for, although others have reported that it is capable of more. E3D is supposed to be coming out with a high flow variant of the Revo later this year which may end up being the way to go. Right now I just want to focus on learning the new machine and getting some time with it. Later on I may switch to a higher flow hot end.

I switched to PETG a little while ago since I am having some trouble getting the larger ABS prints to stay stuck to the build plate. That may be just a matter of cleaning it better, part cooling settings, or I may need to try an adhesion promoter. I am pretty sure I have the first layer calibration pretty well dialed at this point, at least for the 0.6 nozzle. The PETG has to print slightly slower, but it sticks to the bed like nothing else. I'll mess with the ABS more tomorrow.

[Netpackrat]

Watching the console in Mainsail, the mm/s in the slicer are apparently just maximums, and it's clearly being limited by the volumetric flow restriction; in the case of the PETG print the highest travel I am seeing on perimeters is only 45mm/s as it hits the 8mm^3/s limit. ABS is going to be higher... I will have to watch and see if it hits the max or not. Bumping up the max volumetric flow parameter a little may end up being the first thing to experiment with. Still a lot faster than the two bed slingers.

I suppose with a smaller diameter nozzle I could probably hit higher linear speeds without maxing out the hot end, but that's still going to be slower overall. I think maxing out the hot end should probably be the goal. At least I bought the Revo Micro instead of the Revo Voron. It will be a lot easier to find another use for the Micro if I decide to go with a higher flow hot end in the Voron.

[Netpackrat]

Did some searching and found this:

https://e3d-online.zendesk.com/hc/en-us ... low-Rates-

Looks like there is definitely a bit of speed to be had there just by bumping up the one parameter. Not sure why the defaults in the slicer have PETG flow restricted so much, unless it has to do with E3D's older V-6 hot end that the profile was originally designed around. I had thought they were officially rated for the same flow, even though everyone knows the Revo will do better.

[HTRN]

One thing I've read that the frequent limitation is not the physical hardware, it's the electronics - apparently one you get beyond 80mm/sec with an Arduino, it starts having issues with accuracy loss, due to the steppers and the 8 bit Arduino. Something to be aware of, before deciding to push the speed on a 30 hour print.

[Netpackrat]

My Voron uses a BTT Octopus board which is a 32 bit board:

https://biqu.equipment/collections/cont ... 2758765666

It gets its orders from a 4GB Raspberry Pi 4B. With the Klipper firmware, most of the computation takes place on the Pi, which then feeds instructions to one or more controller boards, so the actual board itself doesn't matter much. Klipper can get good performance out of older boards since the Pi does the heavy lifting. Spec for the Voron 2.X used to be two boards run by the Pi, in order to get enough stepper drivers prior to the advent of boards like the Octopus. My setup is kind of overkill since Klipper will run fine with a Pi 3.

I am using the same exact Pi4B to run Octoprint on both of the Prusas.... Eventually I will probably convert both over to Klipper since it is mostly a matter of reflashing the controller board and reflashing the SD card for the Pi. Octoprint is great (and in fact can be used as the interface for Klipper), but Prusa's latest firmware is borked and I have not upgraded the factory machine for that reason. The scratch built unit is now 2 revisions behind, and in order to upgrade it using Prusa firmware I would need to analyze the new source files to see how the changes they made affect the changes I had to make, then recompile again, due to the 100mm higher Z axis. Klipper is probably the best upgrade path for that machine since it won't require any new hardware, and it makes updates a lot easier.

Technically the Glacier Bear (my scratch built Prusa Bear) will print taller objects than the new Voron... The Voron is billed as being a 300mm cube but in reality it is more like 300x300x260... The Bear will do 310mm high prints; it will just take forever to get there.

[Netpackrat]

Just started a relatively large print, having raised the max volumetric flow for PETG through the 0.6mm nozzle from 8.0 mm^3/s to 15.0. Print time per the slicer should be 2 hours 49 minutes, down from 3:34 with the default flow of 8.0. Same object prints in 7:18 on a Prusa using a 0.4mm nozzle. We'll see how it goes. Right now it is laying down the first layer, which is speed limited to 30mm/s, and it's only flowing 7.0mm^3/s at that speed. So I should know how well this will work once it gets the first layer down.

[Netpackrat]

Well, that didn't work. Once it hit the second layer and maxed the flow out at 15, I saw 84mm/s travel speed, but it was underextruding and making a rough surface with buildup on the nozzle. I killed the print and resliced it at 12mm flow, and started over. Slicer shows a 3:01 print time now. Increased the spring tension on the extruder gears also.

[Netpackrat]

It's a lot happier at 12mm^3/s flow. I am seeing 67mm/s travel at that flow rate. I will probably bump it up later on to try to find what the actual limit is, but either way it's a huge improvement over what I have been used to. Of course at the moment I am printing relatively big stuff with non-critical dimensions (because I can, now). I think where the higher linear speeds and input shaping will really show their value is more detailed prints with tighter tolerances, using smaller nozzles.