3d Uber-Printer Build (Voron 2.4)

[Netpackrat]

Well, that sucks. It was working fine, and then completely lost its shit on the 4th layer with underextrusion. Looks like the extruder latch is broken, so I will have to re-print on the Prusa (once it finishes with what it is currently printing), and disassemble the extruder to replace it. I saw a beefed up unit on Printables intended to solve this exact problem a little while ago, so I will probably go for that. Not sure if the broken latch is a cause of the problem, or just something I did trying to fuck with it just now. Will find out once I get it apart I guess.

[Netpackrat]

Back in business. I think it was a combination of the latch was already starting to let go, as well as trying to push the filament faster than was reasonable, regardless of what E3d was able to achieve in their testing. Printed the new latch and got it installed, and the extruder seems to be working again.



This ended up being a lot less hassle than I expected. 4 screws to remove the stealthburner housing, and one more to remove the broken off part of the latch, and put it back together. Didn't even have to disconnect any wires.



Here is the broken latch, with the new beefed up latch for comparison. Or rather, with the spare beefed up latch, since I printed two. The main reason I did this was to enhance part cooling by doubling the layer time, which can be a problem with small parts (the last layer doesn't have time to cool before the printer starts laying down the next). But spares of a failure prone part are always good to have around.

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.

So yeah, I think I have the answer to this question now, and it comes down to just because you can do a thing, doesn't mean that you should. I think the profile in the slicer was set for such a low flow rate, because PETG just doesn't like being printed that fast, regardless of the rate at which it can be physically pushed through a hot end. I just never noticed that before, because at Prusa speeds it doesn't matter al that much. Right now I am playing around with spiral vase mode a little before I call it a night, and I have it dialed back to 10mm^3/s (standard in the profile was 8mm^3/s). And I am still getting quite a build up of material on the nozzle, which drags through the print and messes it up. That may also have something to do with the Revo hot end... People going from an E3D-V6 (which is what Prusa uses) to the Revo have reported having to modify their print profiles since they read temps a little differently based on where the thermistor is located.

[HTRN]

Now you can print your killdozer.

[Netpackrat]

Now you can print your killdozer.

Beelzebub himself now fears the Voron two point four.

[Netpackrat]

My new 0.4mm nozzle arrived today and I am going to start actually tuning the printer, first with that nozzle and later on with the bigger sizes. I am printing just the Voron calibration cube right now so I will have a baseline to compare with after I start making adjustments. First up will be tuning the pressure advance... As expected, I am seeing higher print travel speeds while printing with the smaller nozzle. It's hitting 100mm/s at times, and at that speed the flow is only 7mm^3/s, with the max volumetric flow set to the base default of 11mm^3/s, so there is plenty of potential speed there to be had.

[Netpackrat]

So yeah, spent the first week making large prints that look like dogshit quickly using the 0.6mm nozzle. Mainly because I borked my 0.4mm nozzle and had to order a new one, and I wanted to tune for that size first. Later on I will repeat the process for the other nozzle sizes I want to use. This is the tuning guide I have been following thus far:

https://ellis3dp.com/Print-Tuning-Guide ... uning.html

The first step that he mentions, tuning the extruder, is something I did when I first got the printer running, before I printed anything with it. Basically you tell the extruder to extrude 100mm of filament, measure how much it actually feeds, and then calculate a correction factor which will be included in the printer.cfg file the Klipper uses. Mine was pretty close to begin with, so the correction needed was really small. This is all done on the filament input side of the extruder; if you wanted it would be perfectly adequate to remove the hot end and just feed the filament though without melting it just to take the measurement.

Second, the Z-offset/first layer calibration. This is basically how close to the build plate the nozzle is set for the first layer, and determines how much the plastic is "squished" as the first layer goes down. The process is a little different on the Voron than my Prusas but your basically doing the same thing. The setting I had arrived at with the 0.6mm nozzle was somehow close enough (this is supposed to in theory be an advantage of the Revo, but I had expected to have to adjust it when switching sizes anyway) with the 0.4mm nozzle installed, so I proceeded to print a Voron calibration cube (one of the test prints included in the .stl package) just to have a baseline to see how the printer would do to start.

Next, I followed the guide for tuning Pressure Advance. This is a setting which will vary the amount of filament that is extruded as the print head changes directions (i.e. as it goes around corners). This should reduce bulging and artifacts adjacent to the corners of prints.



The calibration website contains a form into which you input some data for your printer, and it generates a g-code file that you print. The chevron features are printed at different increments of Pressure Advance, and you choose what you think is the right setting based on that. I ended up choosing .035 after a bunch of head scratching, but to be honest I probably could have gone one step in either direction. I printed another calibration cube with the new PA setting, and I did see some improvements around the corners.

The next step is to adjust the extrusion multiplier. This differs from the extruder calibration in that it is filament dependent. So if I changed brands, and sometimes even colors (or individual spools) of filament, in theory I might need to re-tune this setting. I had to tune the EM on my Prusa as well when printing the Bear parts, but the process that the designer of the Bear upgrade is different from what the Voron crowd follows. I'm not sure if one way is more correct than the other, but I think it might have as much to do with differences in how PETG and ABS behave, since the Bear is designed to be printed in PETG, while the Voron parts are ABS. I followed the Voron recommended process, which is more based on visual appearance, vs. the Bear procedure, which is based on actually measuring extrusion widths with a micrometer. When in Rome...



You print these little 30mmx30mmx3mm calibration cubes, at incrementally differing EM settings and then choose the one that has the upper surface the most like what you are looking for. I tried adjusting the brightness and shadow of the above photo to make it easier to see the differences between them, but I only had limited success at that. Anyway, the first square I printed was at an EM of .98, which was way overextruded based on the rough appearance of the top layer. Basically it's putting out too much plastic, and the nozzle makes ridges as it is dragging through it. I dropped to .94 and that was closer, and then printed another square at .92 and settled on that. Based on what I have read, that is a common extrusion multiplier to use with a Voron.

Currently it is printing a new calibration cube so I can see what total effect the changes have made. I know already that I will need to re-do the first layer calibration, which is normal after changing the EM, since now it is only extruding 92% of the plastic that it was before. I will probably do some printing with it like this, and then my next steps I think are going to be to re-adjust the belt tension (they loosen up for a while as the printer is used), and then run input shaping using the accelerometer that came with the kit.

[Cobar]

That looks to be a pretty useful site. I may have to make use of that, especially since I am running klipper.

[Netpackrat]

I am definitely going to need more ABS. Installed a tungsten carbide 0.4mm nozzle in the Prusa this evening and re-did the PID tune and first layer calibration. Going to print the nozzle brush/purge bucket/sheet stop mod on the Prusa for the Voron out of the PC-CF filament that arrived the other day. A couple parts of that end up in contact with the heated bed, so the designer of the mod recommends polycarbonate or nylon for the heat resistance. Hardened nozzles are not really available for the Revo yet and after a year of printing the Prusa needed a new one anyway.

[Netpackrat]

The carbon fiber reinforced polycarbonate was interesting stuff to print with. The reason I chose the CF variant is because PC (and nylon) have a reputation for warping off the bed mid-print, which can lead to print failures an ultimately print head damage if not caught in time. I read a while back that one of the benefits of the reinforcing fibers, is the print is stiffer and a lot less likely to warp. And oh by the way the parts printed from it are stronger too. I printed the sheet stop and nozzle brush holder for the "Decontaminator" mod from it, as well as a new cooling fan shroud (plus a spare) for the Bear. I swapped out the fan shroud but once I got the ASA one off it seemed to be in good shape still, so the PC-CF one is probably just overkill.

Decontaminator mod: https://mods.vorondesign.com/detail/eiG ... 2GapXbQVgA

It's pretty cool; you set a macro to have the printer extrude into the purge bucket to prime and then rub across a bronze wire brush a few times to clean the nozzle right before it starts printing.



Here are the brush holder and sheet stop printed in PC-CF.



This is the nozzle brush installed in the holder, on the printer. The 2 screw heads at the corner of the brush holder serve as sheet stops for the powder coated, spring steel build sheet. The top of the heat bed is magnetic, and these sheet stops make it MUCH easier to install the build sheet accurately.



Here's the purge bucket that is supposed to install around the nozzle brush. Only problem is I need to shorten this in height by about 10mm because of the insulation I put on top of the deck panel. This will reduce the volume of the bucket (I will need to empty it more often) somewhat, but I don't really want to remove any of the insulation from the deck. This was printed out of ASA on the Prusa (you can see a litte overextrusion evident), but the new one I may as well print out of ABS on the Voron since I have it printing pretty well now.

[Netpackrat]

I still need to modify the purge bucket in CAD and reprint, but I did complete the "Rock N Roll" mod last night. This makes it easier to tilt the printer onto its back to access the electronics bay:

https://mods.vorondesign.com/detail/tiI ... HJY0JNJK9A