HI, all
I’ve got about 2275 hours (mostly PETG at 240) on my Ender 5, and have had to do major hotend work (more than re-seating the Bowden tube) five times. One was myowndamnfault, and I’ll ignore it, but this still leaves the MTBF (Mean Time Between Failures) at 560 hours. If I’m ever going to do production printing, I’m going to have to improve this - I would hope at least to double it.
I’ve tried installing a titanium heat break, but it immediately (within half a dozen hours) suffered heat creep and began under-extruding - and it has also proved difficult to get a leak-proof seal between the bottom of the heat break and the top of the nozzle.
Does anyone have MTBF data on replacement hotends? I’m wondering if the fact that the Bowden tube stays cooler would lead to better reliability on, say, a MicroSwiss all-metal replacement.
I’ve also been getting the impression here and elsewhere that all-metal hotends require more force from the extruder. If this is the case, is it worth replacing the hotend without putting in a dual-drive extruder or converting to direct drive?
…and regarding the seal between the titanium heat break and the nozzle - How does one deal with this? A very thin washer of annealed copper could make a gasket for this, but where do you get annealed copper washers?
for production, there are 2 things. Buy production printers (but they probably have strict maintenance schedules) is one but you could also look for swappable head solutions or equip your printers this way. if your MTBF is 500 hours then replace the head with a freshly refitted one at those intervals. Keep spares on hand to refit the heads when it’s not stopping your production then the downtime is reduced and you have a predictable routine.
How 'bout that - they exist after all, and are made for exactly the purpose I envisioned.
The top end of the nozzle and the bottom of the heat break meet in the middle of the heater block, and they’re both made of relatively hard material. Unless they’re machined to very tight tolerances, they won’t mate perfectly, and there’s a good chance that molten plastic will leak from here and migrate to the outside of the heater block. In my case, it migrated to the top, where it gathered around the heat break, reducing its effectiveness. Annealed copper is very soft, so putting a washer of it between the two before tightening them together will form a gasket. It would be a one-time thing. Copper work-hardens a LOT (*)so you would need to replace the washer every time you reassembled the hotend. There may be other materials for this job, but the high temperatures disqualify most gasket materials.
(*) I remember my metallurgy prof demonstrating this. He brought in a half-inch-diameter rod of annealed copper, and asked the smallest person in the class (a woman of maybe 5’ 1" and a hundred-odd pounds) to bend it, which she did with no great effort. He then handed it to the burliest guy in the class, and said “straighten it”. Of course he couldn’t.
The reason I’ve been keeping logs is that I have ambitions eventually to scale this venture into a “retirement” income, and need to know how to plan maintenance.
If it’s successful, it may well be worth switching to industrial machines, but they’re quite a lot more expensive than the stuff in our end of the pool.
Keeping an assembled hotend ready to swap in is basically what I’ve been doing (ok - I don’t always have one ready).
I’d love to know, though, whether an upgraded hotend would increase the service interval, and whether upgrading it without changing the extrusion set-up would just cause problems.
I print with my prusa around 55 hours a week,I just did a clean lube and new nozzle on mine for the first time in 5 months or so. I spent 3 hours of maintenance (super dusty, my basement is to blame). I clean the bed every few days, an IPA wipe down.
they cost a bit more (or a bit less depending if you calculate your time or not) and are super reliable. mine is stock but for the stainless steel nozzles.
I am around 3-5 hours of maintenance every 1100 hours of printing. I have few failed prints and the failures are often the stl not fully touching the bed.
It may well be that a Prusa s the best choice for a small print farm, precisely because of their wise choice to manufacture them by testing them.
Still - a Prusa is an Ender 5 plus $500.00 in upgrades, and one is left to wonder whether the upgraded Ender might not be more reliable yet.
I’m not a fan of the geometry of most 3D printers, by the way. It’s just bad engineering to move the entire bed, plus the workpiece, back and forth to create one of the axes of movement. There’s no question that it can be made to work, but it’s basically just not The Right Thing.
I’d love to have an Ender 5 with dual z-axis lead screws, and with the X axis stepper moved to the frame to lighten the gantry. (It still wouldn’t be coreXY, since it would retain the Y axis mechanism from the Ender 5, which is a really good design)
Add a well-sorted-out and reliable hotend, and you’d have a pretty rock-solid machine.
Yeah, if we had any ham, we’d have ham and eggs, if we had any eggs.
I can’t say if it would be worth it for an ender I have not owned one.My thought is if they vary much you might have new issues with each, time is money. The Prusa has a one piece gantry It is the core of the machine it is absolutely rigid. I believe it is the reliability begins.
Perhaps a Voron is better?
I would suggest genuine e3d hot ends. I have the real deal and clones, the clones I have a poor imitations and the minor flaws I saw might be where you are having more maintaining issues.
My experience with the sidewinder has left me very jaded. I replaced most of the wiring connectors (who mis matches connectors and uses hot glue to seat them???) One stepper, the whole extruder, squared the lower frame, 2 boards, firmware, bed, one idler, belt and added a BL touch, and supports to try to make it rigid. I remade the printer and it is still flawed because of the frame being able to move.
Thermal paste is a must! Both of my OEM Creality hot ends have been delivered without any thermal paste. Cringeworthy.
Keeping the feeder (direct drive or Bowden) clean and free of dusty chewed up bits and stringy filaments.
Titanium heat breaks provide a very sharp thermal profile. My best results come from using a slightly higher temp and slowing down print speed by about 20%.
You should always keep your hot end fan clean as well, and on high speed. Even small drops in fan cooling as would be if blades are dusty or speed slow will permit heat creep. I find that the cheaper fans need replacing every six months at least. At the rate you’re printing, likely every quarter, because worn fan bearings will imperceptibly slow down fan speed.
Large nozzles of .6mm and .8mm and even 1mm, will give you stronger prints and print much faster. For most mechanical parts (useful prints) this is the way to go.
Time on the print bed is a quality factor in my experience. The longer on the bed, the greater the chance of corner retraction (resulting from uneven part cooling) in the model.
I know this is an old thread but for future people looking if you can’t get any annealed copper washers, use a propane torch and pliers heat a standard copper washer till it gets to a dull red and dip it in water till it stops boiling the water around it. (wear gloves and safety glasses, I hold no responsibility for your injury if you do this wrong)
