My journey to choose a high end 3D printer (Raise3D E2) — Part 3
NOTICE: I’ve been contacted by Raise3D and I’ve updated a few parts of this article since it was first written this morning.
Attack of the flying spaghetti monster
Let me ask you. Is perfection too much to ask for. I suppose it might be. This printer won’t be the most expensive printer I’ll have purchased, but it will be pretty close. And since I’m buying it as an individual, the reason I’m being so detailed on this multi-part purchasing review is that I’m hoping to buy a printer that will be as close to perfect as possible for the best price possible. I want this printer to be a 5 year purchase and yes, I’d want more, but let’s be honest, 3D printing is still far from perfect and it’s really only a matter of time before a much better technology than FDM arises for home users.
I went to the store this morning before they opened and through a little window where I could see the printer, I saw that for the second time now, there’s a feeding problem on the left spool and only the right spool is printing… so there’s a pile of spaghetti.
So, for test printing, we’ve been using the dual color 3D Phil from MatterHackers. It’s a really cool little print and it should really come out amazing in the end. But, we haven’t managed to make him print yet.
The filament
For the right spool, I’ve chosen to use PolyMaker PolyDissolve S1.
For the left, I’ve chosen to use PolyMaker PolyMide CoPA.
The choice of these two filaments was easy. First of all, they’re basically engineered by the same people for the specific purpose of printing an incredible material with the best support material possible to go with it. From what I can tell, they are literally chemically engineered to be perfect companion materials.
Different materials
Rather than printing two colors of PLA or PETG or whatever, I wanted two materials that functioned at different temperatures and had different properties. If you’re only printing the same material with different colors, you’re better off using a two-in-one nozzle solution or possibly a filament bonder like the Mosaic Palette 3 (which I intend to review as well with the same detail) to handle filament transitions.
Printing two different materials at different temperatures has some really important benefits. Also, it provides a great two-in-one test. Not only is the printer printing different materials, but it’s also going to typically suffer a LOT of problems with idle time. Other support materials like simple glue based PVA are amazingly bad at sitting idle in a warm hot end. If it sits too long, it crystalizes and clogs. If it’s retracted too far, the little bit of snot at the end of the filament glues itself to the inside of the heat sink. If it’s retracted too little and doesn’t crystalize, it will make drippy snot rockets.
As for nylon, nylon is always a very difficult material to print with. PolyMide CoPA is maybe the easiest material nylon I’ve ever worked with and Taulman 910 coming in as a close second. But it’s a really difficult material to get absolutely perfect since the amazing properties that makes nylon so attractive to print with are the same properties which cause it to clog, jam, dribble, snot, etc… In fact, nylon is pretty much designed to string. This is why ladies stockings made of the material, it’s supposed to be extremely lightweight, have incredible durability when stringy, etc…
Different temperatures
This printer test was meant to be a test of the printer and not the materials. So, I really wanted to get two materials famous for working together to do my best to eliminate the material from reasons for failure. PolyMide CoPA prints at 250C to 260C and PolyDissolve prints at 215C to 225C. While the materials will do a great job, printing with materials with such a big difference between their temperatures is a great test of part cooling.
When printing PolyDissolve S1 on top of CoPA, we expect that to go really well. We want good and even adhesion from the top, bottom, and sides. But we really don’t want the PolyDissolve to melt into the CoPA. The CoPA should maintain a perfect shape and the PolyDissolve should instead smush into it… and up against it. Basically, we want to PolyDissolve to glue to the CoPA not become part of it.
When printing CoPA onto PolyDissolve, we want the material to rest against the CoPA and to a limited extent, we want the CoPA to slightly displace the imperfections of the PolyDissolve surfaces. It sounds like it would could malformations, but in reality, it’s kind of like smoothing the print chemically rather than through software. But for engineering purposes, we don’t want this to be a substantial shift in shape. So, it’s important that the CoPA prints just hot enough to bond perfectly but we don’t want it to cut divots through the PolyDissolve. So temperature control is critical. The printer should lay the CoPA , create the bond and cool the CoPA before it doesn’t bad things to the PolyDissolve in general.
Water is EVERYWHERE
Both PolyMide CoPA and PolyDissolve are materials that will self-destruct when in humid places. This is why I use PolyBox and make sure that I have a ready supply of Silica packets to constantly keep my filaments as dry as possible. Sure I can use a filament dryer as well, but cooking filaments to dry them out takes forever and I’ve ruined a few great spools doing it since not enough heat doesn’t work and too much heat ends up turning a spool of filament into a solid block of plastic.
If you’re wondering why this review should just stop here, I should cancel my order and spend nearly 3 times as much on an Ultimaker S5 rather than on this printer, the Ultimaker S5 Material Station is why. For all the money it would cost me to deal with filament management over the next 5 years, and the struggles that I’ll have screwing with the Mosaic Palette when the time comes, I’d probably save money by spending the $10,000+ rather than the $4,000 plus a billion nickels and dimes. Material handling is critical when printing with more advanced materials.
Stratasys does some really cool things by storing their filaments in cartridges. Ultimaker has the material station. What does Raise3D have?
Here comes the bad
Raise3D has nothing. Yup, they’ve made the printer absolutely suck for material handling.
I don’t have the printer in front of me, so I’ll use some pictures and show what happens when they get weird.
So, Raise3D make almost all the same stupid mistakes that were made on the FlashForge Creator 3. The filament storage for the E2 is just a plastic door, without a seal, with no humidity control, etc… it’s basically begging to destroy your filament.
No external boxes or holders
On top of that, there’s no allowances made for using an external filament storage solution like the PolyBox. In fact, unless you’re going to snap the doors off and hack a solution, the E2 is the anti-filament.
Large Spools
You can’t use large spools. You’re limited to buying smaller spools or winding your own spools.
You’re going to be limited on print size because it will require changing filament while printing if you want to fill the volume of the printer.
Non-removable doors
Either you’ll have to risk breaking the plastic by snapping the doors off, or you’ll have to leave the doors open.
No Mosaic Palette… probably
Since there are no allowances for external filament feeding, it’s unlikely the Palette 3 can be used without busting off a door or leaving the top open.
HATE HATE HATE
Since the doors are on the sides of the printer and you’re likely to need to work from feel to use this printer unless you give it an island somewhere… or you happen to want to leave 30–50cm open on each side of the printer… which means it’s basically on an island, installing the filament should be pain free.
It’s not.
This printer is heavy. You don’t want to have to slide it around often. My intended goal was that I’d put some support material on one side and change it as rarely as possible. I’d leave the other side open to change filament every time I felt like it.
Well, the angles required to change the filament are reasonable, but given the odd angle that you need to feed the filament into the bowden tube, it’s surprisingly common to end up with a lose filament that during retraction could cause the filament to slide off the side and wrap itself around the spindle holder instead.
A less than perfect solution for this is to use a peel-and-stick cable holder to make sure the filament stays aligned with the spool.
But this is terrible.
Raise3D screws up
Notice: I will not change the words in this article as I believe my initial impressions are still applicable. However Raise3D contacted me and made at least an initial clarification. So following my rant about the E2CF which may be based on my lack of information at the time, please read the text marked in italic afterwards.
Raise apparently realized that they screwed this up. And not only did they realize this, but they ended up building a whole new printer… sorta… that addresses many of these issues. But they left it in the same format as the E2 (announced two weeks before I’m writing this) and rather than fixing the filament storage in the printer as they really should have…
They instead removed the internal spool storage altogether and drilled holes in the side of the printer. WHAT WERE THEY THINKING?!?!?!?
These dry boxes are amazing. I mean really, they set a whole new standard for over-engineered and they are truly a thing of beauty.
They considered spool stability.
They considered filament routing to avoid slippage
They considered keeping the spools dry.
And they did it in my personal opinion all wrong.
What they should have done
While the filament storage on the E2 is far from perfect, one thing that is quite nice is that there’s plenty of room to fix mistakes.
They could have offered all the benefits and features of this new solution and also fixed the external feeding problems by replacing the door and spindle holder of the E2 with a dry-box insert.
Either through a vacuum formed insert or through a sealed acrylic/plexiglass container, they could remove the doors from the sides of the printer and then provide two options.
- The ability to place a spool and silica as well as a humidity sensor into the slot where the spools are currently stored.
- A secondary path that would provide side external access to the bowden tube for supporting external filament management.
What they did instead was to limit the spool size again and place two enormous plastic boxes on the sides of the printer.
Lack of upgradability
At this time, the 3D printer store I’m working with has a single unit in stock for testing. They haven’t even sold it yet. They are evaluating whether to sell it. Raise3D could have made the perfect choice and made a proper upgrade kit with two new hotend assemblies and two insertable filament storage units.
Now I’m torn between a few choices.
- Do I look at alternatives? Should I spend the additional funds and just get the Ultimaker S5 and throw away all my 1.75 filament and just rip off the bandaid?
- Do I ask the store to get this new monstrosity and replace the E2 dry boxes with PolyBoxes instead (which I already own)?
- Do I ask them to send this one back and go back to searching for the perfect printer?
I’m making these comments because I honestly had no idea that the E2CF printer existed until I went looking for images for this article and Raise3D made it well known that
- This printer costs the same as the other printer
- This printer seems to have better hot ends
- The modifications are modifications to the design, not simple changes.
- This is really a new printer… but it looks like hell thanks to the poorly considered dry boxes dangling off the side
- Raise3D now will support yet another damn printer. This means slightly different parts, little support nuances, etc…
I’m left wondering what other improvements have they made to this new printer which makes the other printer inferior?
What will save this sale and review?
I am relieved at this point that I haven’t bought this printer yet. It forces me to reconsider and come up with an alternative plan. The E2CF is a slap in the face which says “We know the E2 is not what it should be. We made the E2CF as a better alternative”.
So what I’d like to see is:
- An upgrade to the E2 with all the goodies on the E2CF except the dry boxes.
- A clear comparison between the E2 and E2CF which makes me feel as though I’m not buying a product that even Raise3D considers to be broken.
And here’s the promised update.
Raise3D asked me to look at the original announcement of the E2 which was published in September 2019. It clearly states that the E2 was never meant to be just a product. Rather it states it is a product series. They had intended all along to release new printers based on the platform. So, at least from their perspective, while the E2CF seems very much to me as an E2 that not only prints CF, but also addresses the obvious issues I’ve found independently with the printer. They claim that alignment is even better now, I can’t imagine how, but if they say it, based on my experience with camera assisted alignment on the printer, it must be beyond perfection on the E2CF. And they also upgraded the extruders, supposedly for CF, but the dual drive extruder I’m certain would allow the printer to print flexibles with Shore ratings down to 75A (minimum 90A by spec).
I’ve left my initial words in place rather than editing or softening them because I think it’s really important that not only did I think these things when I lacked information, but also that Raise3D did very quickly contact me and ask me why I thought them. I believe they will proactively address these issues. Remember, they didn’t send me a printer, I didn’t borrow it. I’m evaluating it before purchasing it. I actually believe from the way their mail was written that it’s very important to them that their customers (or at least potential customers) feel well treated. And this is precisely why after watching Joel Telling’s trip to Raise3D I took Raise3D seriously as an alternative to Ultimaker.
Continue as normal
So, now that I’ve ranted, I want to continue writing the review as though I didn’t just get slapped with this crazy new information.
Open Filaments gone Bad
Raise3D was great when they chose to do the Open Filaments program. The idea was that the community would provide filaments and profiles which could be easily accessed online and added to the IdeaMaker software.
I believe they were hoping that the filament companies would consider testing their filaments and adding their own settings to the database.
I think this failed.
The structure of the Open Filaments program looks like there’s no real peer review process involved.
They made a huge mistake by making each Open Filament profile a manifest which includes filament settings as well as print profiles in one. The result is that the Open Filaments catalog is getting littered with a bit of a mess. Each profile which uses a given filament seems contradictory.
A better solution would have been to make the filament profiles including things like density, optimal temperatures, optimal retraction, optimal speed, and anti-oozing settings one thing. And then print settings separate. Then the filaments would be more easily configured and uploaded and tested print settings would be uploaded separately.
I really don’t like the Open Filaments design and I believe I’ll end up creating my own profiles instead.
Also, the Open Filaments Program is a weird IdeaMaker link to a web browser and a kinda clunky integration into the software. If they wanted to take this approach, they really should have embedded the browser into the software rather than triggering an external browser and making the process a little awkward.
Back to the good… no awesome!!!
Alignment
Independent dual extruders have always been a really terrible idea for 3D printers because they require X, Y and Z alignment.
FlashForge proved that there were at least 1000 things (exaggeration? maybe, maybe not) that you can do wrong in designing a IDEX printer. FlashForge did it so poorly, they should have just closed shop, changed their name and hid from the world before trying to come back.
I was expecting a wizard requiring me to align x, y, and z for offsets by looking closely at different prints and guessing the right offsets. I was expecting this to be a disaster from hell… because FlashForge convinced me that this was going to be a miserable process.
So I went into the menu and kept looking for alignment settings. And looked. And looked. And didn’t find it.
I was confused. I couldn’t see how this printer with nothing more than “camera assisted” alignment could not have a menu where you would diddle and pray to get everything aligned.
So I gave up looking and figured, let’s just rip the band-aid off and I tried printing without aligning the heads.
It JUST WORKED!!!
The experience with IDEX printing was as smooth and as perfect as printing on an Ultimaker S5. I mean really, it didn’t just do a perfect first layer for one filament and then half assed on the second. Where the Ultimaker S5 knows the precise offsets of the X and Y and finds the Z through a touch pressure test, the Raise3D finds the X, the Y and the Z magically. I still haven’t figured out what magic this is. I’ll need to sit and stare for a while and guess which way it had to do it.
This picture is a bit messy, but I’ll warn you, I forgot to take the picture before I started peeling up the print. So, this is not a fair representation. This was an intentionally stopped print.
Using the default settings we see that there are purge/wipe towers. And we also see that those wipe towers are REALLY important.
Wipe Towers from hell
It appears that there is, during filament change what appears to be (using the Open Filaments PolyMida CoPA profile) a very generous purge between filament changes… on the PolyDissolve, it’s much less. Overall though, there is quite a bit of filament waste.
This is possibly the biggest disadvantage relative to an S5 and Cura I’ve seen so far. I much prefer purge towers or walls over purge buckets and brushes like on the FlashForge, but this is a bit messy. The problem with purge/wipe towers is that for them to actually be useful, they have to contain enough mass that the wipe process doesn’t just knock them over.
For a large print, they are far more efficient than using a wall around the print. But for small prints, like this one, it is very likely a wall would have made more sense.
Also, to get the required mass while being filament efficient, printing either two conjoined semicircles or maybe for fun, a yin and yang would have been better. The goal is to wipe snot. It’s not important whether the tower is pretty. So, this goal could have been easily achieved with maybe half as much filament wasted.
A little note about PolyDissolve
Anyone who has printed with PVA knows that it’s ugly. It looks like a hardened bodily fluid. PolyDissolve on the other hand is this crystal white beautiful material. I’m honestly tempted to make things out of it and put them in glass boxes to keep them from getting wet. It’s one of the prettiest filaments I’ve even seen.
Closing it out
I’ve written enough for now. I’ll continue once Phil is properly printed. I think I’ll also experiment with my own dry box solution for the E2. I may CNC a vacuum mold and see if I can make a better solution than they did.
See you in the next installment.
