jasonrohrer

And I found the change here, on June 2: https://github.com/Ultimaker/Cura/commit/0f48878bc7ced68ca4c121088d8b2e54c4872c0f The BB nozzles all defaulted to Triangles before this change.

In 4.13.1, if I toggle the support extruder back and forth from 1 (PLA) to 2 (PVA), it auto-switches the default Support Pattern from Zig Zag (for PLA) to Triangles (for PVA). But in 5.2.1, the default Support Pattern is Zig Zag no matter which extruder (PLA or PVA) is chosen. I've successfully printed a very intricate PVA-supported PLA print by slicing with 4.13.1 (there was a bug in 5.2.1 that prevented the slice from completing). Now I'm printing a simpler PVA-supported PLA design, and 5.2.1 is slicing it just fine. However, the PVA supports are detaching from the raft after several layers, and the print is failing. I've tried the print twice, with the same thing happening both times. I'm trying to figure out what I'm doing wrong on this new, simpler print. So I went back and sliced it in 4.13.1 and found this difference in default support patterns for PVA. And indeed, the Zig-zag shape seems really thin and weak, and seems to be what's causing it to detatch so easily (one end of the Zig Zag fails to adhere fully, and then there's nothing to really hold it in place, and it eventually works the rest of the Zig Zag free). Can anyone speak for the reason for this change in default Support Pattern for PVA? Admittedly, the failure could be caused by some other issue. I'll know for sure once my 4.13.1 slice finishes printing. But 14 minutes into the print, it's doing much better. The Triangle PVA supports look well-defined and well-adhered, like a little glass house building up. Where the Zig Zag supports generated by 5.2.1 were already looking quite birds-nesty and wiggly by this point. (I'm also noticing that the 4.13.1 print time is about 30% longer for pretty much the same settings... 2:09 vs 1:39 for 5.2.1)

In this case, since the bases are so tiny, they all released easily, even the bare glass. A release test could use objects with very large bases with full contact to the bed. However, it's much harder to test "easy release" because measuring "how hard you need to push" isn't that straight forward. If you have any ideas about how to do that, I'm open to hearing them.

This is UltiMaker PLA Orange filament.

Results. Magigoo lost two of the taller towers. Bare glass lost three taller towers and two shorter towers. Layerneer Bed Weld lost no towers (there's some bird's nest over there, but it was transported there from the other towers that fell).

Actually, scratch that... one of Magigoo's (45 degrees, 40 mm tall) just fell off. 37 minutes to go. Layerneer towers are still holding fast.

And we've already lost Uncoated Glass, with 2 hours done and 39 minutes left to go. But both bed coatings are holding fast.

This print is in progress now, with three different "stripes" of glass bed treatment. Cylinders have a radius of 4 mm, and are 20 and 40 mm tall (before being angled). Cylinders are angled at 15, 30, 45, and 60 degrees. I will post the results here when the print is done.

Yes, I just printed some stuff with Magigoo on the S3 glass bed, and it was wonderful experience. The parts indeed came right off in my hand, with no tools, after cooling down. I did a Benchy myself, and it looks just like that on the bottom. The print quality is really stellar (and like night-and-day compared to the CreatorPro, even when printing at the same speed---1 hour and 9 minutes for Benchy on both). Also, the "tall thin Y" shape, with a tiny base, adhered and printed just fine with Magigoo, and also came off very easily when done. I guess my days of rafts and brims are over! I'm curious about a really wide-based piece with Magigoo. Like what if the base covers the whole bed? I'll have to try it. Anyway, I'm now 100% sold on the UltiMaker design philosophy, and I think they made the right choice with the glass bed.

My first print on my S3 was a 3D Benchy. I was told by UltiMaker that "if I use UltiMaker PLA, I don't need to use any glue." This was the Tough Black PLA that came with the printer, left nozzle, AA 0.4mm core. Default settings in the latest version of Cura for this setup (which is FINE, with a brim for adhesion). And sure enough, it was very difficult to get the print off the bed! I let it cool, but it wouldn't come off. Then I put it in the freezer for 10 minutes, and it still wouldn't come off. I eventually chiseled it off with a plastic scraper, which required some violence. Meanwhile, the default settings on the FlashForge CreatorPro included a raft for Benchy. With the Wham Bam bed on the CreatorPro, the print popped off in a snap. But it was impossible to get the raft removed entirely from the bottom of the print (this varies from print-to-print on the CreatorPro.... sometimes the raft comes off easily). On the other hand, the brim on the S3 left some jagged edges after it broke off the bottom of Benchy. And the print quality was DRAMATICALLY better on the S3. The contrast is pretty stark. Though the default settings on the S3 are much finer (0.1 vs 0.18 layer height). But there was distortion all over the place on the CreatorPro Benchy. Though the CreatorPro finished the print 2x faster. The only artifact on the S3 Benchy was ringing on the bow portals and around the cabin doorway. I will have to run an apples-to-apples comparison where I tweak the CreatorPro settings until it takes the same amount of time as the S3. Next step is Magigoo, of course. BUT, to see the advantages of good adhesion, I did one more experiment: I printed a 20mm tall cylinder with a 4mm diameter base and 45 degree "arms" sticking off the top, like a skinny letter "Y". I turned off skirts on the S3 and disabled the raft on the CreatorPro. And this print was simply impossible on the CreatorPro with the Wham Bam bed. The little "disk" for the bottom layer didn't stick at all, and just moved with the nozzle. But it worked FINE on the S3. In fact, I was able to print some tall towers with even smaller bases (2mm, 1mm), and they adhered fine (though the skinnier ones started "wiggling" a lot as they got taller, getting distorted, and eventually getting knocked off the glass bed). That's a pretty good demo of where the S3 beats the competition, I think. It can print a tall, skinny "Y" with no special adhesion add-ons. Of course, for such a tiny base, the part pops right off after cooling when you pull on it. I will try this with Magigoo today and see if that makes adhesion for tiny parts less reliable. Not to derail the thread too much.... but care to state the case about why Layerneer is better than Magigoo?

So do you just use "office" binder clips to hold the thicker glass in place?

I've found a few borosilicate replacement plates on the market. Is there one that you recommend? And you mentioned "thicker" plates. How does that work, in connection with the stock UltiMaker bed clips and auto-leveling procedure?

I had a zoom call with the S-series product manager at UltiMaker today. He said there's are two good reasons why they don't ship the S3 and S5 with a flex plate: 1. UltiMaker's bed leveling procedure involves "hot probing," to ensure that filament residue on the nozzle doesn't mess up the measurements. This hot tip, pressing into the PEX coating on the Wham Bam, can damage the coating (make a divot in it). I also heard from the CEO of Wham Bam today, and he's aware of this potential issue with hot probing. Obviously, there's no danger of a hot tip damaging a glass plate. 2. UltiMaker want 99.99% "press PRINT and it always works" reliability on the full spectrum of materials that they officially support. They've found that existing flex beds (like Wham Bam) don't work 100% perfectly with their full list of supported materials. For example, heavily warping materials sometimes "warp" the flex bed up with the material. Other materials have adhesion problems. Thus, they err on the side of "printing correctly every time" and sacrifice "easy removal" on the altar of reliable, versatile printing. Their market niche is reliability. And I must admit that tiny-base, tall parts often get "knocked' off the Wham Bam bed for me on the FlashForge. Thus, I'm in the habit of using a raft every time for smaller parts. I'm guessing that the glass bed on the UltiMaker S3 will be more reliable in this regard (small parts will stick so well than they won't need a raft). Their logic is sound to me, and I think they've made the right decision from a wide-spectrum reliability point of view. UltiMaker doesn't want to adopt a new technique if it's not up to their reliability and versatility standards, and I respect that. Making a flexible build plate that works perfectly is an open problem. I will keep an eye out for "hot probing" problems on my Wham Bam surface. I do plan to keep one glass plate as plain glass, so I can test both.

Turns out that Wham Bam has flexible magnetic build plates specifically designed for the S3 and S5. They have cut-outs on the corners to make room for the glass bed clips. S3 here: https://whambam3d.com/collections/frontpage/products/copy-of-258-x-230-kit-with-pre-installed-pex-build-surface S5 here: https://whambam3d.com/collections/355-x-275/products/copy-of-355-x-275-kit-with-pre-installed-pex-build-surface I have two glass beds for my S3, so I will install the Wham Bam on one, and then test with identical prints on both, and report back with any differences that I notice.

Well, my target here is lost-PLA casting, and Nylon isn't going to work for that! 🙂 I'm also running this printer in my house with kids and a pet, so I'm trying to avoid too many fumes. PLA smells delicious!

I have a tube of regular (not pro-series) Magigoo here. I got it for the Creator Pro, but never needed to try it after I got the Wham Bam plate. I know that the UltiMaker S3 comes with a glue stick in the box (which itself is kinda silly), but maybe the Magigoo is better, because of it's "magic release" property. I'll find out!

Well, the question still remains: Why hasn't UltiMaker addressed this issue? I'm hoping someone from the company can chime in here. Since everything else about the printers has been designed with such careful intention, there must be a good reason for not having a flex bed. But I'd like to hear it. (It also might be good to know.... if they tested flex beds, and print quality was slightly degraded, then some of us could be aware of that, and keep two glass plates around, one with a flex for quick prints day-to-day, and another plain glass when we want the very highest print quality possible.)

Yes, I came home with the S3. It was that good, despite this one flaw. Print quality and reliability is more important than ease of removal. After all, if it can't print well, there's nothing worth removing from the bed! I'm only printing PLA with PVA supports. Nylon was just a scare story that I heard from the salesman. But he demo'd PLA with PVA supports for me in the store (a tiny mushroom print). And he could BARELY get the thing off the bed. He knew this was a bad demo of the product, and he was kindof embarrassed about it. I came home thinking, "I'm definitely going to find a flex bed mod for this thing."

I just got home from the Ultimaker showroom in my area. I was convinced enough by the demo that I saw that I came home with an S3. The showroom printer.... well, it just worked beautifully, and felt like how a 3D printer should be. Except for one glaring problem: getting the print off the damn bed. To see this elegant machine work its magic on a TINY demo print.... and then watch the salesman try fruitlessly prying off the print, then watch him stick the plate in the freezer for 10 minutes, then watch him try to pry it off after that.... he eventually got the print off by VIOLENTLY jabbing at it with a putty knife, and the print itself went flying across the room. Oh yes, he tells me... some Nylon prints get stuck so bad that one customer pulled a piece of glass off the plate with the print. This guy has been selling UM printers since 2016... he's not a novice. Granted, this was a tiny demo of PVA supports, and PVA sticks worse than PLA (or so the salesman told me). "Oh, we always use gluestick on the bed," he told me. Apparently it didn't help. Or if it did help, I can't imagine how much worse it could possibly be without gluestick. But the whole experience was ridiculous.... especially when shopping for a $4500 printer. I recently upgraded an ancient FlashForge Creator Pro with a Wham Bam flexible magnetic bed, and the thing now prints perfectly AND the prints come off gently and easily by just flexing the bed. No more frustration. No more violence needed. (Though it can't print PVA supports to save its life...) And YES, there are Wham Bam mods for the Ultimaker S3 (I told the UM salesman about them, and he was eager to learn about them). But the question: Why doesn't this $4500 printer COME with a magnetic, flex-bed in the box? There must be some reason Ultimaker is avoiding the obvious solution.... maybe it reduces the quality of bed leveling? Or... what? (Or maybe this salesman has his demo printer tuned wrong for the first layer... that very well could be...)