Speed

Print duration is dependent on two components:

• How fast is your print head moving? I run velocities/accelerations similar to you partially because I have a 350 which is pushing the limits of 2020 extrusion and 6mm a/b belts as well as…
• How much filament you’re laying as the print head moves. This is influenced by your nozzle diameter, which in turn influences what kind of line width and layer height you can expect. It’s also influenced your extruder’s ability to melt plastic (eg volumetric flow). For ASA/ABS I limit volumetric flow to 35 mm^3/sec, or PLA I limit to 25 mm^3/sec, and for PETG I limit to 20 mm^3/sec

My print speed is often limited by volumetric flow - not the actual speed of my print head, so I haven’t bothered chasing higher ceilings. Granted, tend to print I print large/chunky/functional things so my goal is to lay down as much material as possible. If you’re chasing lots of fine detail, a smaller Voron can go faster than what I have but isn’t going to be that much faster than where you are now.

Print Quality

Thanks to a combination of CoreXY (rigidity) and Klipper (pressure advance, input shaping), I have basically zero ringing/ghosting show up in prints. It is worth talking about quality expectations though. Harsh lighting can reveal that layer lines are not perfectly aligned layer to layer. Not sure if this is a Voron thing or is it’s just more obvious now that my layers are a lot more noise free.

First layer

Automated gantry leveling (Klipper will get the bed and gantry to be ‘perfectly’ in plane thanks to 2.4s being able to mechanically move the four corners of the gantry independently - trident does similar, but moves the bed instead), a klicky probe and a Z calibration macro, and bed mesh make my first layers extremely consistent print to print.

One caveat: because the printer is enclosed and big (if you go for a 350), if you print sequential objects without letting the printer fully heat soak, the first layer will progressively get a touch higher and higher between prints as the printer expands in the z-axis.

I replied to another post with a list of mods, so take a look at the other comments in the post for some out of box mods.

As a 350 owner, be aware of two things.

First, big bed = big chamber = heat soak takes a while and you have a lot more surface area to lose heat from. If you want to print big ABS/ASA parts you’re going to want ACM panels, a better sealed/insulated front door, and potentially a radiant layer inside the printer.

Second, the big printer limits your rate of acceleration some compared to a smaller CoreXY. IMO if you have a big printer to print big things you’re probably not going to have small/finely detailed parts that often. Those are the kinds of parts that will go a touch slower. But honestly 5k acceleration is orders of magnitude faster than most bed slingers can achieve and 10-15k is only a 2-3x increase so you’re not giving up that much.

Other than than, no regrets about the 350.

I assume you mean “what mods do I recommend out of the box”?

1. Klicky. I personally think tap adds too much mass and klicky is great
2. Magnetic panel clips to make it way easier/faster to get the panels on/off
3. An under bed filter with carbon. I’m using “the filter”. Even if you’re not going to print ASA/ABS the extra chamber heat helps eliminate warping on large PETG parts s
4. If you’re going to be going to be building a larger printer and print ASA/ABS skip to ACM panels. Also do #5
5. The fridge door is so much nicer than the stock double doors, but isn’t something you need to do out of the box
6. You’re probably going to run into wire breaks in the cable chains - especially the x and y chains. An umbilical makes that much more unlikely. You don’t have to have to USB or CAN to do this

… Off the top of my head, those are the big ones

I got CNC parts from microcenter. They had the chaotic labs kit in a differently branded box for significantly less $$ than available on the web. The QR code for documentation inside the box went straight to chaotic labs and the parts look to be identical.

As with anything, do some level of research before you buy.

For context, I can run prints on my 2.4 what would take something in the neighborhood of 4 times as long on my old ender 3.

I would put the difference even higher between my 2.4 and my old i3 clone, but I’m also running a 0.6mm nozzle and print with 0.9mm extrusion width / 2 walls and 0.3mm layer heights. My limiting factor is volumetric flow, which I’ve found to vary between materials (ASA = way easier to print fast than PETG).

You have four paths to get parts for a Voron:

• self source all the parts individually
• get kits for the things you want and self source what you want to be picky on. For example, here’s a motion kit
• configure a kit from someone like West3D
• buy a BOM in a box. There are a couple of brands that do this

I personally went the West3D route. It seems like the LDO and formbot BOM-in-a-box options are popular. If you live near a microcenter they offer smaller kits if you want to mix and match or use a brick and mortar. Self sourcing tends to be expensive due to our collective tendancy to buy higher quality than necessary parts and shipping.

Yes, you will be building the thing from a ton of parts. Yes, it will take you a while. If you’re comfortable building things there’s nothing particular hard about it. You absolutely will not need to solder. Most kits come with premade wiring harness and there’s plenty of complete wiring options available even if you buy components. Depending on your goals, you might need to customize your wiring some. This means crimping, which isn’t hard per say but you’ll probably need to buy a crimper or two and dial in the right amount of squish for your terminal and wire gauge combination. Too much force and you’ll wind up severing the wires. Too little and the terminal will come off the wire. Again, not hard but you’ll probably need to do it a few times before you develop a feel and get consistent.

Speed on a core-xy, especially acceleration, can be a ton higher than a bed slinger. I have a 350mm^3 Voron keep a 0.6mm nozzle on it, and print with 0.9mm line width and 0.3mm layer heights. I have a Rapido HF and volumetric flow winds up being my bottleneck most of the time.

Also note that when you’re going fast, material matters a lot. I can melt ASA faster than PLA/PETG. But… ASA can be a bit more melty so things like overhangs can suffer. And the whole needing a heated chamber thing.

One of the most prevalent mods for my old i3 clone is a z-brace, so…

Ah, galvanic corrosion. Sealing the heat sets would be a pain as they’re open to water on both sides. I also don’t tend to take these apart, so I’m tempted to just leave them alone.

I do! But I also know that rebar tie wire is intended for static situation. This whole thing moves around quite a bit when it’s covered in leaves on a windy day

All has been revealed.

Your tuning looks great btw, look pretty nice even in the worst case lighting conditions, adhesion not an issue doing this way?

Thanks! I will make a “same print, terrible lighting vs good lighting” post in the next day or two. No, adhesion wasn’t an issue. I run klipper with z calibration, so my first layers are very consistent.

My dad asked me to print some stuff he designed for his beekeeping tools, has a bearing surface that’s awkward to print accurately, I’m probably going to resist that with this as inspiration, other than the helper ears I see on the build plate anything else you did?

Most of what I print is self designed. I do my best to make sure the designs are (reasonably) easy to print. For overhangs, printing a part at 45 degrees is basically a cheat code, but it comes with the tradeoff of more iffy bed adhesion.

I printed these with a skirt and some tree supports to improve stability. Beyond that, I didn’t do anything special and don’t put anything on my bed.

To ask questions, for the application does dimensional accuracy actually matter? AFAIK rebar isn’t exactly the tightest wrt tolerances (I know flat products, not long products, but knowing what hotroll coils look like I’m assuming it’s similar), could probably have gotten away with a different orientation and could probably have avoided supports (I find arches print nicely).

These are ASA parts with 0.3mm high layers. If you haven’t printed ASA before, it’s a bit more melty than PLA. You’re right that rebar isn’t super dimensionally consistent, but a 14mm radius cylinder wouldn’t have printed very well flat.

Having said that though, thinking strength might be another reason to print the way you did, face down and you have shear & torsion in between layers, thinking that’s still a concern if you printed it standing, but yeah, just thoughts.

I don’t think layer lines matter a ton in this particular print. Most of the stress is going to be torque caused by one bar trying to rotate relative to the other bar.

also spy kapton tape, did you find the bubble insulation made much of a difference? I’m putting what’s basically heat barrier fabric on the interior as a first try, I grabbed some rock wool and bubble insulation but it’s thick enough that I’m mildly concerned with it interfering with the gantry, having everything off for some refurb and wow I forgot just how close everything is, they really didn’t waste space eh?

A few answers here.

First, swapping to ACM panels bumped my chamber temps. If you dig through my post history you can find a temp graph comparing before/after. Second, adding a radiant barrier did help, but was less significant than the ACM panels. I do need to do the back panel and want to make my fridge door double pane using one of the stock acrylic sides. Third, I have magnetic panel clips that I modified to give me a bit more space to accommodate the radiant insulation. You would never be able to fit this stuff in with the stock panel attachments.

Happy to hear the photos revealed the mystery, lol.

Believe it or not, the outside fence, which is super jankey, is pretty effective at keeping deer out. Rabbits, not so much :(

I hope so! The zip tie solution only made it one season :(

ASA

45° is a great cheat code for overhangs assuming you can keep the parts on your bed.

It’s the center section of a rebar clamp for a plant support structure in my garden. I’m hoping to get a photo of it installed today.

First try. I’m running klipper with z calibration with a bed mesh, so my first layers are very consistent.

I manually painted the trees in two locations. The first was below the cylinder cutouts even start to give the print a touch more stability, but I probably could have gotten away without them. The second was for the small overhang at the top of the cylinder. This is ASA, which tends to sag on overhangs a bit more than PLA. A bevel would have been a great way to eliminate the need for the second set of trees.

The prints are the middle section of a rebar clamp for my garden. I’ll try to post a photo of the completed unit in a day or two. So not a jig per say, but functionally very similar. Good eye!

I updated the OP. The dimensions in the arch is the feature that matters for this print.

The prints really do look pretty good under normal lighting. The orientation did not impact print quality. Elis has a nice demo of this on his site

Only kinda sorta dangerous, I’m not leaning into chaos random. I’ll update the post with a better view of the final part. The two cylindrical recesses are the dimensionally important features and bridging would hurt that.