Stress testing the Bambu Labs Carbon X1 (and myself)
There are many other things I was supposed to be doing today, but today ended up becoming the day of throwing everything I know how to do and then some at the 50-pound Carbon X1 3D printer now occupying what was originally a dining room table and later became a desk as covid WFH entered a second year. I am back at the replacement kitchen table again, the inevitable outcome of taking a job that involves entirely too much hardware and clutter.
This is my 4th 3D printer. I started this journey in June 2019 with the Monoprice Maker Select v2, which at the time was considered “full-featured”. What those features were in retrospect, is not all that obvious to me. It wasn’t even the now-default Bowden tube style printer, it had a direct drive that constantly required manual intervention to do much of anything. Other than this:
That printer lasted less than a year. I ended up moving the first week of March 2020 (long story) and put a disproportionate amount of effort into saving this shitty printer and the 3 entire rolls of off-brand filament I had for it and bringing them to my new apartment instead of leaving them in the storage unit where a lot of my stuff sat until that summer. March 2020 wasn’t even over and alas, the print bed on it literally ripped in half trying to print face shields using PETG. There were sketchy wires hanging everywhere and the printer I had just put so much effort into moving in a frantic scramble to find an apartment before the hammer came down ended up being unceremoniously hauled out to the dumpster in a drawer from something else that broke in the move.
What followed that was extremely predictable. I joined an online group that was for makers trying to socialize online during the covid-19 shutdown. Everybody else in the group was using either an Ender 3 or a Prusa, and I realized those were basically the 2 printers anybody else really saw as a viable option. I promptly went and bought 2 Ender 3’s after seeing the massive wait to get the Prusa I wanted and set to work with that.
And I did. I ran those printers 24x7 up until the point when I moved again about a year later. At that point they had been broken and repaired and broken again so many times they weren’t really salvageable, and they too went into a dumpster.
At that point for me the story of 3D printing really felt over: I had simply done what I felt like I could do with it, other than learning CAD, and trying to print things from CAD, which had mostly been futile in past attempts. It was the original motive for purchasing a printer, but not one that had resulted in a whole lot of productive output, and at some point I simply gave up and moved onto other pursuits.
But a few things really changed between early 2021, when I mentally just gave up on 3D printing, and January 2023, when I got the bug again (sigh). During my first go at this I had mostly been using a ThinkPad that had been a gift, but which quickly became just overloaded with everything I needed to be doing the technical work that I was trying to accomplish. CAD wasn’t impossible on the ThinkPad, but it was very laborious and never ran smoothly enough to make it much of a learning experience. It would only have been productive if I already knew exactly what to do to overcome the frustration of doing it on a machine ill-suited to the task at hand. I eventually got an M1 Mac Air and quickly filled that hard drive and everything else on it as well, but it is a much better tool for running a CAD program than even a pretty expensive Lenovo laptop turned out to be.
The 3D printers that are available to consumers at a normal-ish price point also have changed a lot, it turns out. Other people that spend more time obsessing over gadgets have more thoroughly documented what makes the latest and greatest so different from the Prusas and Ender 3’s of the past, but without some fairly intentional context it’s hard to get a sense of what the difference actually is.
The older printers were purely mechanical devices, and almost all of them required you to manually insert an SD card with the gcode for the object you wanted to print. That’s over. The Carbon X1 connects over your wifi via some sort of spyware magic and you can just print straight from the app without fussing with USB cables or SD cards, and even monitor the print progress from your phone or smartwatch.
The Carbon X1 can be set up with a brand-spec AMS, aka a material changer, that has the feel of an inkjet printer in terms of how automated it actually is, it can detect if there is anything in each of the 4 slots, it tells you if there’s an issue with any of them, you can have it just automatically change those colors/materials without any real intervention or supervision on your part. It’s great. That was all impossible a few short years ago. Definitely get the AMS if you feel inspired to buy a printer after reading this post, I think that was really a big missing piece for the technology and it’s hard to know what my experience with this would have been like without it.
The Carbon X1 does all kinds of automatic calibration and troubleshooting that would have been unthinkable on an Ender 3 or similar. It will just pop up on the built-in LCD screen with a QR code if something is wrong, and that directs you to a page with the detailed instructions to fix it. To do that it uses LIDAR and AI and all kinds of other stuff that I will undeniably struggle to fix if there ever is a problem. It is like the difference between driving a beater car without Bluetooth or a backup cam, and comparing that to a brand new vehicle with lane assist and touchscreens. People will point to the downsides and cost of having all of that automation, but also overwhelmingly prefer to have it, because the reality is driving without any of that stuff is even more of a pain in the ass. I’m sure the automated calibration/bed leveling/filament monitoring has saved me a massive amount of time and frustration that in a vaccinated world where I have a much more demanding job than I did in 2019-2020, I just plain wouldn’t be able to do a lot with something that can be as much of a timesuck as the old style of 3D printing.
But in not even a week I had already burned through several rolls of filament using the Carbon XL, which I was rarely able to do over several years on the other printers I had, even though I ran them much more of the day and often overnight while I was sleeping. They were just so unreliable that a lot of the time the actual output or material usage was close to zero. Once I had the X1 dialed in, I have so far only had very sporadic issues, and only by pushing it to the absolute limit of what I know how to do with 3D printing.
Which brings us to today. I had already used the XL to print a massive hoard of the infamous parametric boxes. Several of them are beautiful, almost jewel-like. The largest and earliest ones that I printed are pretty ugly, because it turns out that the magnetic print bed that ships with the unit is sort of useless (boooo!). It requires using a gluestick, which sort of wrecks the surface appearance, and in my opinion a printer I spent almost $1600 on should just… ship with the print bed that is actually useful. If you end up purchasing a Carbon X1 because of this post, save yourself some trouble and get the textured PEI plate from the Bambu Labs website to go with it.
I will point out for those familiar with the parametric box design that I did not print the TPU seals for these boxes. There is almost certainly a way to do it, but I struggled to get it to work with PETG either, and TPU apparently requires manually feeding the material in and messing with things that I just wasn’t ready to try on something I spent so much money on and did not want to already have to take apart and fix. I will eventually return to trying to print with non-PLA materials on it, but it wasn’t really the starting point I went with for trying to find the limits of what the Carbon X1 can do.
I got into 3D printing because of things I had heard at my old-old job in the DC blob about its use in aerospace for rapid prototyping. I had talked to all sorts of brainy people from NASA and SpaceX and obscure defense contractors that were using 3D printing to design jet engine parts and drones and robots and thought, well if they can use it for that, certainly I can just get a 3D printer myself and design much simpler things, like a pair of shoes or a phone case or some cute animal figurines or something.
Ha. Ha. Ha. What I found was the same thing that everybody who thinks like that finds out, which is that none of those items are actually simple they just feel that way when they are designed well, and unfortunately the reason mechanical engineers spend so much time in college learning CAD is that it’s actually just not very easy to master either. Yet, here I am trying all of this again… 🤡
I woke up this morning after sleeping poorly and thought, “I don’t feel motivated to do any of the things I need to do, like figure out my taxes, or buy groceries, or fix the broken drawer on the cabinet under the TV. But what I could do is print the biggest and most complicated thing I can find.” And I spent a bunch of time drinking coffee and looking at models of the Eiffel Tower and castles with turrets and dragons going, what the fuck would I do with any of this crap though? Seriously?
And then I saw a scan somebody had made of a sneaker. Like, a really good one. And I thought, that is much more what this was always supposed to be about. Yes. Let’s print a sneaker. A full-sized replica of a shoe I already have, and then I can just compare and see how much the 3D printed scan version really looks like the actual object.
This turned out to be a pretty solid test of the Carbon X1’s capabilities for a couple of reasons. One, a shoe is just pretty fucking big compared to most things people print. Two, it’s something that requires srsbznss support, because actual shoes are mostly made of textiles/leather that can just be sewn together. I never had any luck printing with supports on any of the previous printers that I had, so I was curious to see what would happen if I ran a whole huge print with a fuckton of support on the Carbon X1, and didn’t cheat by using the Support W filament that is always out of stock. I had some luck with that before, but it clearly wasn’t a sustainable approach for printing with supports, ‘coz supply chain. Three, I wanted to print it using some cool color-changey filament from Amazon that I never had any luck with before, because it looks more like a swaggy gym shoe than plain PLA, and that turned out to be a total waste of time even on the new and expensive printer.
But in the battle-tested Prusament in a solid color I had a lot of luck. The shoe looks pretty much just like my own Adidas sneaker, and the supports broke off easily with the assistance of a scissors. That alone still makes the X1 a much better printer than an Ender 3 or anything similar to it, because supports never really worked very well on those even with a lot of fiddling and going into different slicer programs and playing with the settings. It might be the most impressive thing I’ve printed just in terms of the sheer complexity, realism, and difficulty in actually carrying out the task.
But that was an 8-hour print, even on this beast of a printer. While that was running I mostly occupied my day with finally tackling the challenge of learning CAD, for realsies this time, no take-backs. Good god, what an exercise in frustration. I can draw anything, my friends and siblings marvel at my ability to conjure a sketch of almost any concept out of thin air. I very easily mastered turning very complicated P&ID diagrams into SCADA screens. But CAD has always felt just completely ass-backwards to me, like: WHY DOES CLICKING ALWAYS GODDAMN DO SOMETHING? SOMETIMES I AM JUST THINK-CLICKING. I AM CLICKING BECAUSE MY BRAIN IS MOVING NOT BECAUSE I WANT THE SCREEN TO DRAW STUFF OR MOVE STUFF OR DELETE STUFF. God. CAD programs should just have a mute button on the mouse. Maybe that’s pan, but it should just be a real mute, like nothing happens until you un-mute. Like on Zoom. I need mute buttons for every other program.
So I set some sort of simple boundaries around the effort to learn CAD. I need to design a real useful thing, not a plain cube, or rectangular tile with a word on it or something, and I need to at least make an attempt at printing it, but no more than 3 attempts at printing it. It needs to be designed like a professional thing an engineer would draw, and not like my prior attempts at CAD where I would mostly just import an image, trace over it, and extrude. I need to use real techniques I heard about in past-job doing factory SCADA like mirroring and filleting and all of the other fun stuff in AutoCAD and Fusion360 and the other eeeevil professional CAD tools.
Okay, so what though? Here I was at a loss. I do not like junk or clutter or gimmicky things. I was never ever going to be the type of maker that wants everything to be about putting blinky LED lights in things, no disrespect intended toward the people who are finding fun and enjoyment with that, but it’s just not me. I want to design real things and use 3D printing as a tool to advance other skills and talents in my arsenal.
I puttered around my house for a bit and noticed perhaps a dozen books scattered around, some of them just laying open, others with pencils or pens jammed in where I had left off, or merely dog-eared. Then I remembered: yeah, actually, before covid, before all of the chaos of the past several years, probably the last thing I really did with any of this before it just sort of became “dissociate from the chaos by PRINTING! ALL! THE! THINGS!” was to design some simple bookmarks using my crude “trace and EXTRUDE™” method. That was in January of 2020.
Okay, so, from there the need, and the challenge, was obvious: I was just going to do all of that again, but I was going to do it a lot better, with a much better computer, and more patience for using the CAD tooling, and a completely different type of 3D printer. I cast about the living room looking for inspiration, something to lift an interesting texture or motif off of besides yeah, rectangle with a sloped edge or some kind of words or something, and found it in a Chinese fan that a friend had brought back from a trip to the mainland.
If I can just get these little circles… I thought. And scrapped about 20 different versions of “these little circles” before eventually figuring out how to use the sketches in Fusion 360 effectively. I didn’t even try to pattern them the same way they are on the fan, just doing this with my n00bish skill level took several hours of iteration to get the way I wanted it and then actually designing the full bookmark was a whole other thing.
Once I had something I could use though, I set about just tiling it and punching a zillion tiny holes. I did not figure out quite at first that I could just select all the holes at once and punch them until I inevitably fucked some piece of this up and had to start over, at which point I was like, “there has to be a better way!” (yes, I just didn’t know it, because I don’t know what I am doing).
I settled on a design vaguely reminiscent of OXO-brand kitchen gear with the little raised pad instead of the pattern on the entire bookmark, thinking, it’s nice and ergonomic and great for people that fidget a lot, which is me when I am reading (or doing pretty much anything else). I designed it to be fidget-friendly to again, encourage actually using it instead of just being a gimmicky thing that “ooo it’s 3D-printed!” and then gets discarded in a drawer somewhere with other gimmicky swag from old jobs and glossy brochures from old apartment buildings.
One of my goals with buying the Carbon X1 was to get away from the “omg but it’s 3D printed” mindset, and I left up some of my old Christmas decor that had been 3D printed while I was awaiting its arrival to try and hold that thought while I was daydreaming about all of the things I was going to do with my new toy. Nobody ever remarked on the 3D printed provenance of the Christmas decor, or even noticed that it wasn’t from Target or Home Depot or something, and that to me is the bar for success NOT “ooo it’s 3D printed!”. That was fine with a $270 Ender 3, but at the >$1000 price point the bar just plain has to be a lot higher in my opinion.
I had about 30 minutes left on the shoe print when I started writing this blog post, and it took about 30 minutes after that to print the first of the bookmarks. I’m THRILLED. This is by far the happiest I’ve been with 3D printing since I got the Ender 3’s, and the first time **ever** that it feels like the technology is starting to live up to the potential I saw in it when I got into it a few years ago.
I could tell as soon as I opened the enclosure of the printer that a gentle slope on the raised piece of the bookmark may have further enhanced the design, but that is something you can only figure out by actually getting something printed, fast, successfully the first time and not by having 3D printing just be the whole point of a design project. That instant, UGH, I love it, but XYZ… emotional feedback is the promise of a paradigm like rapid prototyping, and it’s just not something I really felt like 3D printing delivered on at all when it was as time consuming and low-automation as it was up until this year.
The feedback is the point. Being a little bit mad, at the design, at yourself, and feeling something about your own skills and ideas and not just feeling frustrated with the printer or with the filament or the temperatures or the support or the slicing or any other fiddly annoying bullshit is the point.
At least for me. I’m sure other people will Year of the Linux Desktop this one and still be printing on Ender 3’s in 10 years, and more power to them. I love Linux. I can even tolerate the other people that love Linux. But sometimes, you just have to use the correct tool for the task you are actually doing, and if the thing you want to learn is CAD and rapid prototyping and design, I’m going to disappoint you and tell you that this is the correct tool and just buying a crappy $300 printer from the nearest Microcenter or off Amazon is not the right thing. If what you really enjoy is just fixing stuff and maybe fixing 3D printers is the bug you have instead of fixing old cars or fixing on a house or reinstalling Linux 80 bajillion times, then keep on Ender 3’ing or build your own printer and have a bunch of fun with that.
But I feel like I really accomplished something today, I learned a whole bunch of things that I was simply blocked from doing in an effective or efficient way before now. And I’m going to keep doing it, because now I have the right tools to do that.