My journey to choose a high end 3D printer (Raise3D E2) — Part 4
Play me some jams!
Some of you might be wondering, why show us this horrible image? I looked at it and I feel as though it was a great start to the day. I don’t want to see the printer only when things work. I want to see it when things go wrong and I want to find out how reasonable it is to fix the issues.
So what? It’s just spaghetti right?
Nope, this was a catastrophic jam. It showed me some truly terrible things about the printer. But the good news is, it’s fixed. The bad news is, it was very difficult to fix.
So, the left extruder clogged. Not only did it clog, it clogged, jammed, oh… the horror. It jammed in the heat sink and it was STUCK SOLID!
This was a problem which was a horrible reoccurring event on the FlashForge Creator 3. Due to multiple reasons, hot, soft, filament finds itself cooling and lodging in the heat sink. I believe in the circumstances we’re facing here, it was a combination of either temporary heat creep, or it was due to retraction and a VERY effective cooler on the heatsink. Either way, it happened.
THIS IS GOING TO HAPPEN!
Before you start wondering if this is a flaw in the printer, it’s not. This is FDM printing and while direct drive, if properly implemented will ALWAYS outperform Bowden, it also makes printer maintenance a disaster.
FlashForge’s design was one of the worse on the market for multiple reasons. The nozzle on the FlashForge is this long solid tube which has many incredible benefits. But it also means that when jams happen (and they will) either it’s time to throw away the nozzle which is expensive, or you’ll have to remove the nozzle, which is difficult, heat it up to insane temperatures and carefully drill out the filament. The reason for this is that you can’t just use intentional heat creep to free the filament.
Raise3D wisely made use of a class/typical heater and nozzle
This is great. I love this. But with the good comes the bad. This design is proven and solid and although I’ve been tempted by BMG Mosquito or E3D Volcano on multiple occasions, I believe this design is generally my favorite.
There are some real problems that come with it though.
They are a pain in the ass when you have really nicely design hot end assemblies like the E2 has. Any maintenance to the nozzle requires taking off face panels, sometimes removing the entire carriage, etc… so to avoid dealing with the common maintenance issues, we tend to try every alternative route to solving this problem without having to perform all-out surgery.
Things I’m concerned about
This is possibly the greatest single flaw in the E2 printer design. A semi-pro or professional class direct drive printer should NEVER ship with a single drive extruder. This is a cardinal sin in 3D printers. Sure this printer was released in 2019 when we hadn’t learned the truth, but a modern 3D printer should never have a problem gripping filament.
The E2CF has in fact fixed this problem and I’m hoping to hear back from Raise3D whether it will be possible to purchase an upgrade kit for the E2, but I honestly would have expected this upgrade from Raise3D in 2020.
There are very many problems that simply don’t occur with dual drive which happen all the time with single drive. And though I haven’t tested flexible materials on the printer yet, Raise3D does clearly state that their printer can’t handle soft materials. They rate the printer for nothing softer than 90A (if I recall correctly).
I make a point of this because the way to clear this clog without a full hot-end disassembly was made much worse because I couldn’t depend on the extruder to help with the problem. In fact, it got in the way.
I want to complement Raise3D on a graceful failure as you see in the photo above. Because of their good design, when the left nozzle was being dragged through the spaghetti of the right nozzle, it didn’t creep up the assembly like the Creator 3 would do consistently. In fact the FlashForge Creator 3 was a fire risk because of how it handled this situation. The filament would slowly creep up and form a cocoon surrounding the entire hot end and heat sink and the Creator 3 wouldn’t even detect thermal overruns. That was a nightmare.
Because of the more traditional design of the Raise3D and it’s somewhat elongated nozzle (compared to an MK8 for example), it failed beautifully here. I had to scrape the snot from the nozzle and clean it. But it DID NOT risk fires or thermal overruns. I would have expected however a thermal underrun that should have been noticed in software, but I believe at no time did the nozzle heat over 250C in a battle against this disaster.
Also, given the nice shield surrounding the hot end, there was never a chance for the filament to attack the sides of the heating block.
Top marks for Raise3D on this. Really. This was great!
Raise3D should add spaghetti detection
Since Raise3D is not an OctoPrint based platform, using 3rd party services to detect spaghetti and automatically pause printing needs to be implemented by them. This would require training an ML model and running it on their RaiseCloud service, but it is a logical addition in 2021/2022. I don’t know if there are patents on this, but presenting it as “print right” rather than “spaghetti detection” or even adding another $3 to the price of the printer and paying a licensing fee would be reasonable. This is not difficult tech to implement and once the model is trained, there’s no reason they couldn’t integrate it into an add-on for the printer to run it at the edge.
Why I’m less happy about this than I appear
Notice I’ve been comparing Raise3D more to FlashForge than Ultimaker than usual in this segment.
This is because the UM S5 is a Bowden printer and generally doesn’t have these problems.
Also, the UM S5 hot end assembly is a system where you can pretty much remove the assembly as easy as you’d remove ink from a printer.
The Raise3D E2 suffers some of the same problems here as the FlashForge Creator 3.
- It’s a closed in hot end assembly. This is good and bad.
- Any maintenance operation that requires force inflicts terror as you’re worried it will mess up alignment
- Removing the nozzle causes the hotend and heatbreak to come out of alignment, so there’s no practical way to access access the filament path from the bottom without disassembling the entire block.
- The extruder is only single geared which limits the torque the motor can deliver to clear a filament clog.
- The filament path is very long which makes it very difficult to find rigid enough tools to attempt to force the path open.
- The single drive extruder is not capable of feeding materials such as ProtoPasta’s Iron PLA which is extremely rigid and could have fixed this.
- The path is too long to use cleaning filaments effectively.
So, the end result was, to open the path, I fed a broken off piece of Nylon into the extruder to make the problem worse at first. Then I used a 2mm umbraco key (hex key) to force the filament to feed. I also inserted a steel wire (borrowed from an Ender 3 v2 accessory drawer) through the nozzle in hopes of providing a catheter to transport heat. I then used another piece of broken off filament to stop the heat-sink fan from cooling the heat-sink.
Overall, the only way to solve this problem was to use excessive force and risk damaging the extruder gear (I didn’t… but could have) and risk knocking the printer out of alignment (you should never EVER apply this kind of lateral force to a linear rail). Or, I could disassemble the hot end assembly which honestly terrifies me as I can picture taking out 20 screws and putting back 19 or 21.
Mitigation?
I haven’t had enough time to digest these issues on the Raise3D E2 yet. The way I solved these problems on the FlashForge Creator 3 was to give it to someone else haha. No really, I did, but because there were so many other problems too.
Dual drive extruder
I think Raise3D should simply stop making the single drive extruder. It really just doesn’t make sense to ship single drive anymore and now that they do have a replacement for it, they should retrofit all new models with the new extruder and they should make them available for upgrading existing E2 printers.
Learn to gain access to the filament path
I’m sure that for manufacturing and maintenance reasons, the Raise3D documentation is excellent for showing how to remove the hot end and reseat it properly. I’m also sure that they must have either helped hundreds of customers with this process and I’m sure that their software must handle realignment after this process cleanly.
There are stories about how Toyota took over the world of cars. The most famous story I’ve heard was their invention of screws that would allow cars to be assembled rapidly and cleanly without constant adjustment. This happened in the late 1970’s. Everyone in the entire machine design industry knows this rule (except FlashForge) and I’m sure that the reassembly of the Raise3D hot end makes it easy to ensure that it’s in the right place when it needs to be.
I look forward to when I have the time to go through these steps and become comfortable with them.
The last problem with this is a problem which is worse for the Raise3D E2 than it is for the BCN3D Sigma.
To perform surgery on the hot end assembly of the Raise3D E2, it requires being able to reach deep into the printer and turn screws. This isn’t really a big problem, but especially when doing it for the first time, it will be difficult as it look as though it requires you to work entirely by touch.
Wrapping it up
Overall, other than my disappointment with the single drive extruder, I still think the E2 is nothing short of a fabulous printer at this point. I will do some research to see how Raise3D recommend clearing these horrible clogs. Their support seems to really take these kinds of issues seriously.
