jasonrohrer

Also, if anyone knows where this bug should be reported, that would be helpful. It's not a Cura bug, but a printer firmware bug...

For others who are facing the same issue, the API can be accessed by visiting your printer's IP address with your web browser, like this: http://192.168.88.209/docs/api/ By clicking the various commands there, you can see example syntax, and you can press the Try It Out button to execute a command. In this case, you want to go to the Materials section and run the DELETE command with the GUID of the material that is "stuck" in an invisible state in your printer. The idea is that if a material is loaded that doesn't match your printer (like loading an S5-only material on an S3 printer), your printer will accept the material but then NOT display it. However, if you try to re-load the material again later (maybe after fixing it to support the S3), if it has the same GUID, the printer will reject it, because it's like, "Nope, that GUID is already taken." Of course, it doesn't tell you what the issue is... the material just fails to load. The other amazingly broken thing about this is that even the Materials GET API does not list this "phantom" material. But the printer remembers it. I guess the API only lists usable materials. But then we must wonder why the printer remembers unusable materials. SO... you need to use the DELETE command with that GUID to clear it out. Then after that, loading the material via USB stick (or whatever method) will work. Anyway, if you ever find that a material fails to load---barring errors with the material file itself---it's worth drying a DELETE operation through the API on that GUID. As for where you find the GUID, I'm not in front of Cura right now, so I'm not sure. But I think you can also open the material file directly, with a text editor, to find the GUID.

Sorry that I never reported back.... I eventually found a solution. From an email that I sent to the AquaSys people when I finally figured it out: I'm not sure that I ever got this to work correctly when duplicating materials inside Cura. It has been a while, but I think I ended up modifying the material XML files by hand. But you still need to be careful, because if you upload an S5-only profile to an S3 printer, it won't show up in the printer's material list, BUT it will still block future profiles from being uploaded if they have the same GUID. My original problem was that these third-party material profiles (provided by AquaSys and BASF) were only written for the S5. I uploaded them to the printer, got Material Not Loaded (b/c I have an S3), but the printer still remembered the material in an invisible way. So then when I went to edit the material file to support the S3 and re-upload, I got Material Not Loaded again. But this time, it was because the bad S5 profile was blocking the new, correct S3 profile. However, the S5 profile did not show up in the front-panel material list NOR in the material list as fetched by the HTTP API calls. However, I could prove that the bad S5 material was still remembered by the printer because the "remove material" HTTP API call for that GUID was successful. And after calling the remove API call, adding the material again via USB worked fine.

I got an UltiMaker S3 specifically for printing with dissolvable supports. However, I could not get even simple test prints to print reliably and repeatably with UltiMaker PVA. Prints would start out fine, but have adhesion issues midway, and then go haywire. And yes, I tried every suggested fix, but none of them worked. AquaSys 120 seemed to be 100% reliable, by comparison. However, using it to support PLA is technically "off-label," and it is also very slow to dissolve (especially in 95F water that won't soften PLA too much). And there's no "fine" profile for it, and Cura says it's not compatible if you're using the AA 0.25mm nozzle for your PLA. AquaSys GP is meant for use with PLA, and it dissolves very fast in 95F water. However, the profile situation is even more dismal (no Marketplace profile, because UltiMaker won't allow a direct competitor to their PVA). It took a ton of monkeying to get their beta profile loaded correctly on the S3. It seemed pretty reliable (once a print got going, it would always finish). However, there are apparently some issues with the filament cooking off in the nozzle. Whenever re-loading GP, a bunch of black bits would get pushed out. The nozzle eventually clogged, and I had to go through the whole cold/hot-pull cleaning process a few times. And the beta profile provided (which seems to have some temp problems?) also has no "fine" profile, and again Cura says it's not compatible if you're using the AA 0.25mm nozzle for your PLA. BASF BVOH is the most reliable of the bunch. I've never had a print failure or a clog with it, ever. Again, the profile situation is dismal (no official UltiMaker Marketplace support, and the "beta" profile provided requires a bunch of manual tweaking and fiddling before it will load on the S3). And if you use the BASF-provided beta profile, there's no "fine" setting, and no support for AA 0.25mm PLA. HOWEVER, I'm having excellent results just using the Generic PVA profile when printing with BASF BVOH (as suggested by another user of this forum). That lets me use a "fine" profile and print my PLA with the AA 0.25mm nozzle. So, print after print, BVOH is just working. No black bits when re-loading. No clogs. And it dissolves quickly. Two down-sides to BVOH: 1. It REALLY sticks to the bed, more than the other supports do. I'm using Layerneer Bed Weld on the glass bed, which usually makes removal easy, but I still have to pry really hard to get BVOH to pop off the bed. (Though I suppose you could always soak the bed in water, if it was ever really stuck.) 2. It makes the water instantly cloudy/opaque when dissolving. This makes it very hard to visually monitor dissolve progress (you have to fish the part out, blindly, with tongs, to check it). Finally: Needless to say, this has been a VERY frustrating experience for a $4000+ printer that has dissolvable supports as one of its main selling points. UltiMaker: it seems like you should seriously consider adopting a 3rd-party solution as the "official" dissolvable support for PLA, instead of continuing to push UltiMaker PVA, which just isn't reliable enough. Anyone who tries any one of the three alternatives (AquaSys 120, AquaSys GP, or BASF BVOH) would see a night-and-day improvement over UltiMaker PVA.

This perennial "great taste" vs "less filling" argument is entertaining. But the fact of the matter is that in the US, it is impossible to walk into a hardware store and purchase a metric tape measure. In fact, I've never seen a metric tape measure with my own eyes. Most of our "desktop rulers," for art or drafting, have both cm and in. And our glass measuring cups for cooking have both ml and oz. But those are the only places that the metric system occurs in everyday life in the US. My non-electronic Starrett calipers, which I use for engineering projects, are in inches only. It hasn't changed since 1790, and there's no sign of it changing any time soon. As for why it won't change, people who use the old system seem to like it. Esperanto is also a "better" language than English---but like the metric system, it has no soul.

deleted

In fact, I just heard from Infinite Material Solutions (the maker of AquaSys 120) that even THEY are having problems getting custom material profiles to sync on UltiMaker printers----and they're actively developing materials.

Also, in the ".umm" export archive saved to USB, I can verify that my custom ".fdm_material" file is present. So it is getting exported. But the printer is ignoring it when I try to import. However, in that similarly generate .umm file containing a marketplace material (AquaSys 120), the .fdm_material file is found and used by the printer.

I'm trying to make a custom material profile for BASF BOVH support material. I duplicate the Generic PVA material profile in Cura, then edit the information to look like this: Then I edit the print settings to look like this: Note that I leave the material set as PVA, since otherwise Cura will complain that there are no matching profiles (for Draft, Normal, Fine, etc). I also hit the Unlink button (not shown above) to unlink this from the Generic PVA profile. This generates a new GUID for the material (and I've checked the material file to verify that a new GUID is present). However, I cannot get this profile to Sync on my UltiMaker S3 printer. Cloud Sync doesn't transfer it. USB Sync shows the following message on the printer screen: "No new material profiles found." This is after going to Settings -> Maintenance -> Import Material Profiles on the printer. I've tried two different version of Cura (4.13 and 5.2.2). Same result. I've tried a Factory Reset on the printer itself. I've tried completely clearing my Cura settings and starting over in both 4.13 and 5.2.2 I'm also able to connect to the printer API (like this: http://192.168.88.209/api/v1/materials ) and see the full list of materials as XML. And the new material isn't listed there. HOWEVER: I am able to sync Marketplace materials (like AquaSys 120) via Cloud and USB. With USB Sync for a Marketplace material, it says "Found and installed 1 new material profile(s)." The Custom material file is attached. The material file itself is attached. generic_pva+%232.xml.fdm_material.zip

Here's the official UltiMaker guide from Nov 2022 which talks about 35c: https://support.makerbot.com/s/article/1667410781980

The recommended dissolve temperature for PVA is 35c. Does the new PVA Removal Station keep the water warm, along with stirring it?

Also, have you ever compared AquaSys 120 to BVOH? I'm having quite a bit of solid reliability and good luck with AquaSys 120 so far. I just kicked off a 17-hour print, which I'd never trust to PVA, and 2-hours in, it's humming along perfectly. I just ordered the BASF BVOH, though. I was sold on the (very expensive) S3 with the idea that PVA-support printing "just works," as long as you keep the PVA dry. That hasn't been the case for me. The fact that the printer SHIPS with a spool of PVA... I even had the salesman demo a PVA-support print for me before I purchased the machine. It's noteworthy, in retrospect, that he printed an upside-down mushroom that needed *very little* PVA support. He had the classic "nested gyro spinner" on hand, which they printed in the store at some point in the past... but it might be one of those "roboticists treasure their videos" kinda things---here's an amazing 30-hour print from the one day that PVA supports actually worked for us. If AquaSys or BVOH is a more reliable option, I think UltiMaker should move toward recommending those, and probably just discontinue UltiMaker PVA.

Yes, I've been investigating BASF BVOH. Is there a particular material profile that you like to use for it? I see that it's not present in the UltiMaker Marketplace.

52 minutes left, and it looks like this now:

3 hours into the 4 hour print, it looks like this. It's getting messed up enough there near the top that I doubt it will properly support the next layer of PLA when needed.

I have a brand new UM S3, and I'm printing PLA with PVA supports I've had quite a bit of luck printing with the Normal and Fine profile with PVA supports, but it hasn't been reliable with the Visual Extra Fine profile. The project file is attached. One time, the bottom layer of the triangles didn't print correctly... one or two of the "bars" of the triangle just weren't there. Another time, the bottom worked, but later on, an upper layer got gummed up, and the support stopped printing correctly. This is relatively new PVA filament that has been stored consistently in PrintDry vacuum-sealed boxes (that my local UltiMaker dealer sent me home with when I bought the S3): https://www.printdry.com/product/vacuum-sealed-filament-container/ I also leave the silica gel pack that came with the filament in the vacuum box. And I seal/pump and vacuum the box when it's empty, to keep the gel from absorbing from the room while the filament is being used. It's also winter here, and quite dry in the house. Below 25% humidity. But I haven't been leaving the PVA out between prints. Still, it has been on the spool holder *during* prints. So, last night, I put it in a Sovol 3D drying box, which heats to 50c and has a little fan. I left it in there for more than 12 hours. And now I'm trying the print again, leaving the PVA in the drying box, and holding it at 50c during the print. There's a little hole in the lid of the drying box for the filament to pass through. That leaves only 1 foot of filament exposed to the air, between the drying box and the bottom of the filament feeder on the printer. And it's still not working great. Bottom layer missed a few struts, but I just let it go to see what would happen. It seems to have recovered somewhat, but now some of the upper layers are missing support areas too. The model itself seems to be sufficiently supported so far, even with the incomplete supports. But it's definitely not what I want to see. I don't like the feeling of "taking my chances" with derelict supports. It's now 2 hours through a 4-hour print.... we'll see what happens. I've had much more reliable results using the AquaSys 120 dissolvable filament. However, Cura doesn't have finer profiles by default for that filament. Not sure of the reason for that, but I can't even get down to 0.1mm (0.15mm is the finest it will go in Cura). UMS3_supportTest_v4_PVA_ExtraFine.3mf

This is not a bug, so I won't report it as such. But it is risky. In general, I've found 4.13.1 to be much more reliable than 5.2.1 And I'm a brand new user (just got an S3 for the first time a few weeks ago), so I have no previous experience with either version. I started using 5.2.1 right away, as recommended by UltiMaker, but I ran into one problem after another. Every time I reverted back to 4.13.1 as a sanity check, I found that the problems are resolved. This has not been a great experience as a new user of a $4000+ printer. If I was stuck using 5.2.1 exclusively, my new printer would be useless to me. Fortunately, through many hours of investigation and experimentation on my part, I've found my way back to 4.13.1, and it seems to be quite reliable. Maybe for users that need cutting edge features, like metal printing, the should be steered to Cura 5. But beyond that, it doesn't feel like it is quite ready to be called THE version of Cura that everyone should be using. Seems like it might be wise to continue flagging it as experimental or beta or something, and calling 4.13.1 the official "Stable" release recommended for users who need a reliable workhorse and don't need cutting-edge features. I'm speaking as a software developer with decades of experience shipping systems to end-users, so I know this might be a painful pill to swallow. But part of the UltiMaker brand is that it "just works," and that brand image is being weakened by Cura 5, at least in its current state.

After more layers are printed, you can see a pretty dramatic 1.0mm overhang on the second-layer PVA supports. This is asking for trouble, I think.

And here's layer 92 from 5.2.1:

Also, in case it's not clear, here's what the model looks like. Obviously, in real life, you'd never print this model vertically like this. However, it was designed to be a good test of multi-layer supports. In real life, I am printing much more complicated structures that have compound-overhangs like this and no possible "support-free" print orientation.

Here's what Layer 92 looks like from 4.13.1 You can see how the supports above the PLA shelf "bridge" to connect with the PVA scaffolding on the outside of the model.

On left are samples printed with Cura 5.2.1 On right are same models printed with Cura 4.13.1 The second image shows a close-up of how the second-layer supports came close to failing. I think this is because of the reduced horizontal expansion. This causes the supports to "skip" the layers where there is solid PLA in the middle. Then the supports try to "start up again" above that PLA shelf, but they have trouble adhering. Whereas, with 4.13.1, since the model is encased in PVA, there's a nice wide "foundation" for the second-layer PVA to adhere to. That foundation includes PVA at the edges and PLA in the middle. I'll post some screens from the slicers in a minute. supportTest.stl supportTestShort.stl

I've been wiping it down repeatedly with 70% Isopropyl alcohol, until there are no visible fingerprints or dust. I hold it an an angle in the light, to get a glancing reflection, to make sure it is clean and streak-free, etc. I don't let the IPA dry on the surface, but wipe it dry with a paper towel.

Thanks for the info! Zig Zag isn't exactly "new", since it was an option in 4.13.1. But it's definitely new as a default. Maybe they also tweaked Zig Zag? And it's certainly faster, and I'm sure it dissolves quicker too. However, I don't see them making an argument that it's "more reliable" They say: But... it's way less reliable. Less material put down, less cross-lines holding it together, no external wall for added stability, less horizontal expansion, etc. I have opened more specific bug reports on this issue already on Github. The lack of reliability isn't a bug, so I haven't reported that. But there are floating bits in the support structures: https://github.com/Ultimaker/Cura/issues/14184 Another change is that horizontal expansion defaults to 1.0mm now, instead of 3.0mm. This is much riskier in terms of print failure. I'll take some images of side-by-side finished prints from the 4.13.1 defaults and the 5.2.1 defaults. 4.13.1 "encases" the model in support material (thanks to the 3.0mm horizontal expansion), where 5.2.1 has second-layer supports completely "on top of" part of the PLA, which is risky, and comes close to failing.

Also, this contradicts this article from November 8, 2022: https://support.makerbot.com/s/article/1667337600656