3D Printer Content Warning!
tldr; The
Creality Ender 3 S1 Pro has everything documented below,
out of the box with zero effort. This is me being cheap and using the whole animal.
I've been pushing the limits of my original model Ender 3 for several years now, making basic modifications to keep it somewhat up to date, the most major changes being a Big Tree Tech SKR Mini v1.2 controller to get silent stepper drivers and replacing all the fans to help quiet it down. When I first started I used a lot of PETG filament, and it is good stuff with a nice resilience, but found that carbon fiber filled PLA had better rigidity, a nice look and feel, and less stringing issues and clean up. The carbon fiber is abrasive though, and it really does eat through soft metal parts, so I switched to a hardened steel nozzle (which is much less thermally conductive and does affect print quality slightly.)
But most Creality printers until very recently did not use an 'all metal' hot end, instead they utilized the PTFE bowden tube all the way to the nozzle. This actually works much better at the lower temperatures of most common filaments, so it makes a lot of sense for them to use that method. With a bowden tube design the extruder motor is remotely located and pushes the filament through the long PTFE tube and low friction is critical. But the drawback is that PTFE needs to be kept below about ~240-250C unless you want to generate super toxic fumes, so some of the more high temperature filaments are unusable with that design. It honestly hasn't been an issue for me, PLA is actually stronger than a lot of the high temperature stuff, but the one downside is that it can't handle being used in high temperature environments like the inside of a car or to mount hot motors and so forth.
The go-to material for hot environments is ABS which is actually one of the classic printer materials, being what most people used before PLA became common. ABS has its own challenges, and although it will print on a teflon lined hotend, it is borderline as many formulations really like ~260C, so you are really pushing the safe limits. It is also very sensitive to environmental temperature fluctuations, requiring an enclosure of some sort to stop thermal stresses and ideally 50C+ air to keep the natural shrinkage from badly distorting the part or causing loss of bed adhesion. It also can give off some objectionable fumes by itself, which again an enclosure can help deal with.
I basically didn't want to deal with all that ABS drama, and until a couple days ago I avoided even trying. But my friend has been having ongoing issues with a classic (otherwise outstanding)
Anova sous vide immersion heater he uses in his pub. Under daily use (it's admittedly a consumer machine), the plastic holders keep cracking, and even though they happily replace them every time, the fact that they have not changed the materials, and the inconvenience of replacement has just become a nuisance. So I've been wanting to make him an Aluminum holder for a long time, it was a key project for the CNC I'm attempting to build, but his holder is broken yet again, so I want to get him something that will work for now, and I can modify once I have more advance metal processing available.
To make this possible, a 3D printed part is required to bind the simple Aluminum parts that can be fabricated with my current tools. But using PLA or PETG in a hostile kitchen environment, especially for a heater mount, seemed to be asking for failure, so I decided to stuff some cardboard around the Ender and wired up a 30W PID heater to get a 50C chamber temperature, and gave some eSun ABS+ a try. The results were disappointing, lots of layer adhesion issues where the print would come apart like a stack of pancakes.
On closer examination I started to suspect the print environment wasn't the issue, it really looked to me like under extrusion, something that's been plaguing me off and on for a while. Clogged nozzles (something that carbon fibers can exacerbate) will create back pressure and cause the extruder to slip, and for a while I was running with the extruder gear grub screws not properly tightened
, so those were the first things I checked. But everything was clean and correctly attached. Doing some basic tests, marking a 100mm length on the filament with a pen and extruding that amount made it clear that it was slipping badly. My old eyes aren't any good, and the extruder gear is located in an awkward spot for inspection, so I pulled the whole motor out and sure enough, the abrasive CF had worn a difficult to see groove in the brass gear teeth that grip the filament, allowing it to slip under moderate load. The fix was simple, just offsetting the gear so a fresh area was exposed to the filament, but I'd had enough and it was time to upgrade this system to a dual gear direct extruder and all metal hot end.
There are a lot of good hot end replacement solutions out there, some harder to obtain over here than others, but the unit I chose was the new
Creality Sprint Pro upgrade kit, as it ticked the right boxes feature wise, and was likely to be less fiddly to adapt than the more professional offerings for generic retrofitting. It was also available for next morning delivery from Amazon, which thanks to instant gratification, is often the deciding factor for me. The kit is intended to bolt on to the newer Ender series printers, and although it was adaptable to my original model, it wasn't intended for it, so I had to make a lot of little tweaks to get it mounted properly. The Sprint is an all in one direct drive unit, meaning that the extruder and heat block are mounted together. This places the extruder motor on the print head, which potentially adds a lot of moving mass, a bad thing, and so to minimise that issue they use an undersized lightweight stepper motor and gearing to give it enough torque to properly extrude.
As mentioned, the Sprint comes from the factory on the new
Creality Ender 3 S1 and
S1 Pro, the
Pro version being 'all metal' which as I mentioned trades off a little bit of print quality for higher 300C temperature support. The kit version comes with a new mounting plate and the ribbon cable with a quick connector to properly retrofit it to an older Ender, they also sell the unit a little cheaper without these parts for newer machines that use compatible mounts and cabling, make sure you order the right unit if you get one. It has a built in mounting location for the BLtouch (and their clone the CRtouch) automatic bed leveling sensor, but doesn't include the little
Molex 5-pin PicoBlade jumper cable required. I ordered one for a few bucks from RS-Electronics, but cobbled together a workaround patch cable to get everything running while it gets delivered from overseas.
The geometry of the new print head is very different, and it required a number of changes. The new mounting bracket had clearance issues at the extreme travel positions that required taking about 5mm off of the mounting bolts. A while back I got an Irwin (Made in USA) metric tap and die set that uses hexagonal dies, which I was a bit ambivalent about as most dies use standard round sizes and this gimmick felt like lock-in, but this does come in handy for quick bolt cutting.
I thread the bolt into the die, which itself is easily clamped in a small vice. Cut the bolt to length, give it a few whacks with a file, then unthread it and have the die clean up the threads as it comes off. A normal nut can do the same thing, but not nearly as well. That's your hot Irwin Tools
tip-o-the-day
.
I also countersunk the screws of the X axis belt tensioner to free up even more travel. This wasn't really required for the nozzle to reach the edge of the bed, but was needed to allow the BLtouch to access it, as it is mounted far to the left of the nozzle. While everything was torn apart, I thought it opportune to investigate the cause of occasional X axis layer shifts I've been experiencing. These were usually subtle, but occasionally were bad enough to ruin prints. They were only in the X axis and when the head was moving to the right, so I knew it was some movement issue. After some fiddling around with it, I suspect it was the toothed timing belt engaging with the aluminum extrusion near the lower side of the motor pulley. The belt runs inside the v-way slots of the extrusion, and on the tensioning side the roller is large enough diameter to keep it parallel clear of the edges and interference free, but the motor side seemed to have a smaller diameter pulley and that presented the opportunity for the belt teeth to grab at the sharp edges of the extrusion. I thought of trying to smooth out those possible engagement points, but instead chose to simply put a half twist in the belt between the lower carriage mount and the motor pulley. This causes the belt to be twisted perpendicular through that danger zone and have ample clearance without any obvious negative effect.
The new print head mounting bracket had an extension of the stamping to engage with the X axis home end stop, but it was too long for the location on my machine so that it prevented the nozzle from reaching the left edge of the print bed, losing about 10-15mm of precious printable range. I relocated the end stop by rotating the X motor mount 180 degrees and flipping the end stop switch which increased the offset and this greatly increased the range, but this meant the home position for the nozzle was now about 10mm
off of the bed, but at least this was something I could resolve in the firmware.
Speaking of firmware, it had been a long time since I compiled a new version of Marlin for my machine. Once I have a working version, there is little reason to upgrade it, it tends to be pretty stable, but getting everything synced, sorted and compiling clean is always a trudge. All these changes certainly required an update, and every time I need to go through and validate all the customizations to the default configuration headers to support the somewhat outdated and memory constrained board I upgraded to. There has been a lot of iterations of change, compile, load and test as I got all the new offsets and limits dialed in to support the different geometry and performance of the new print head.
So after all that, what do I think? It's not bad honestly. I changed from the industry standard 0.4mm to a 0.6mm nozzle, that's kind of the new ideal size and is more carbon fiber friendly. I've just finished printing a
dimensional calibration print with the end of a spool of very old and moisture laden PLA, and although there is some expected stringing from the moisture, it printed just fine. Time to crunch the numbers and compile the FW with the corrections.
Sadly I don't have anyone (yet) to hand it down to on this side of the world. I am waiting for an affordable marking laser (I hate stickers, but maybe they are a better choice) so I can put tiny QR codes on everything that are linked to a database of descriptions and rough prices for my wife, so she at least has the info to sell it (although I bet she just dumps it somewhere, not that I will care at that point!) Even if I had a Kohei, the sheer volume of what I have would overwhelm the average Japanese living space by a lot. If I actually figure out a business model, maybe it could be sold off or handed over to employees to manage. It's a conundrum, but it will work itself out somehow. A lot of my most treasured items came to me by way of fate, I suspect they will find someone who loves them after I am gone.
show, but it is done. As I mentioned, I've been trying to finish a long abandoned home maintenance issue to reseal one of the outside verandas. Stripping and preparation was difficult as the elastomeric coating someone had previously applied had degraded and was blistered and gooey, and the area itself was narrow and exposed to the beating sun. There was a lot of material that needed to be removed from hard to access areas under the flashings and sliding glass door jams, so a lot of tedious cycles of scraping and scrubbing with IPA to get down to the crazed but still effective fiberglass underlayment.
