Wow larsenstephen, thanks for sharing.
That print comparative analysis is very well done and informative. I’ve noticed in other F3 related posts that there is an overall drop in quality/crispness in 25z and 50z pints produced by F3, besides the primary warping and drooping issues.
Please do tell if you’ll notice any post-update improvements.
I’ll be watching this thread like a hawk - I have 13 more days until my purchase deadline.
I realize that hope is not a strategy, but I really want to give this one a chance.
Wow larsenstephen, thanks for sharing.
The wrong circular shapes of windows and other opening looks for me as parasitic curing of residual resin. If you notice the distortions are appearing consistently on the down side (during the printing) of the detail. It’s natural on such places the residual resin to stay for longer (especially with not so fluid resin) and hence to be exposed from the parasitic laser light. It’s the same effect described in the conversation below but affecting the holes in the detail.
Also ensure that once you take the detail out from the printer for washing there is no strong daylight around - the residual resin which is already sensibilised can harden in few seconds and you’ll be not able to remove it.
I’ve seen you respond with this in multiple threads about bad prints from the Form 3. I don’t understand how this is helpful at all. You call this “natural” yet it was not a problem and IS not a problem with the Form 2. Why would it be different with the Form 3? And even if it is “natural” it is in no way, shape, or form acceptable for the Form 3 to have been released in this state.
This has absolutely nothing to do with daylight curing unfortunately. The filling of holes only happens when the hole is parallel to the build plate as well for the layer shift lines. Holes that are perpendicular print more or less as expected. IMO this is an issue with the flexible film, which in theory is better than a rigid window and may one day prove to be the better way to go, but that day is still a ways off given how long this problem has persisted and how it it’s still a nagging issue despite the number of preform / firmware updates that have happened since this printer’s release.
I’ve had a Form 3 since they were released and received one of the first shipments. I sold a 1.5 month old F2 to get the F3, unfortunately the F2 was the superior machine and still is. If you’re looking to produce the level of crisp detail you get from a F2, save your money and buy an F2. Come back to this thread a year from now and if it’s dead then upgrade.
Ironically I just received a marketing email from Formlabs touting the quality of the F3 over the F2 for jewelry. I’d say this is their eventual goal, but the reality is not yet there.
While the parasitic residual resin curing theory is worth serious consideration, I don’t think that is what is happening with the models I’m printing. Here’s why:
We don’t experience that defect with any of the five Form 2 printers we’ve had (we presently have 2 on hand which operate nearly 24/7). We have experienced circular opening distortion only with both of the Form 3 printers we’ve had. This tends to indicate that the issue is Form 3-specific.
That defect is not occurring anywhere else on the model one would expect to see parasitic resin pool. The defect only occurs in circular openings. This tends to indicate that the issue is specific to certain geometries.
The room in which we operate our printers is windowless. The artificial lighting we use in that room does not emit UV light. This tends to indicate that environmental UV light is not affecting the models.
Based on these factors, it is my opinion that parasitic curing of residual resin is not the cause. IMHO, the cause of the misshapen openings is specifically related to some characteristic of the Form 3 and only the Form 3 (or Form 3 firmware) in its present state of development.
If I am wrong and parasitic curing of residual resin is the cause of the misshapen circular openings, it does not explain the other serious defects we see on our Form 3-printed models such as layer shifting, mushy detailing and surface bulging. FL techs suggested two possible causes and recommended solutions:
the models require more support to prevent those defects from occurring. This suggestion is a bit perplexing since, 1) the models print perfectly on our Form 2 with fewer supports, and 2) the Form 3 is explicitly advertised as needing fewer supports and smaller contact points than required by the Form 2 due to the Form 3 placing lower peel stresses on models. When more supports were added, we continued to experience the same defects.
we should only print at 100 microns, not 50 or 25. But test prints at 100 microns exhibited the same defects we experienced at 50 microns. Additionally, the characteristic “Aztec stair-stepping” visible in inclined and curved surfaces printed at 100 microns spoils the surfaces of the models making them unsellable.
So far, the two Form 3 printers we’ve had have not produced a single model our customers would accept. I look forward to running some print tests on our Form 3 with the updated firmware this weekend (the gray resin for the test has shipped, but not yet arrived). The test will include the battleship superstructure model shown in the photos above. I’ll post photos of the results.
Our Form 2s, on the other hand, routinely print beautiful models. The rare Form 2 print failures we experience normally indicate that the mirrors need some cleaning. Once cleaned, the printers resume producing beautiful models.
I remain hopeful that FL is resolving Form 3’s issues. There are features of the Form 3 I do find far superior to the Form 2 (optical cartridge, easier enclosure access for cleaning, bigger touchscreen, etc.) I am very anxious for the Form 3 to be able to produce sellable models like those we are get from our workhorse Form 2s. We want it and we need it.
I think the Form2 vs Form3 prints say so much. I’m regretting my decision to purchase the Form3. I definitely could not recommend to anyone. Maybe Formlabs will fix it eventually, but the pace is extremely slow. Believe it or not, I considered the Form2 when it was originally release but decided against it because I thought the technology needed more time. I was very wrong. The technology took at least one step back.
Yep… this has been a standard response from them. This is the response I initially got, until i submitted file after file and picture after picture where I utilized more supports, changed orientations, etc., with absolutely no improvements. then they finally admitted that the problem I was writing about is indeed an actual issue that they say they are now currently working on. Unfortunately I was told that a firmware update sometime in the future is the only way this may be fixed. So for everyone experiencing these issues… it’s gonna be awhile.
I’ve noticed that if you look very closely to the LPU during printing, you can see that the film sags a bit.
It’s not clear how that would cause the specific artifacts that everyone has been seeing, but I imagine any kind of sagging is undesirable and perhaps they just need to tension the film more.
I’m curious if this sagging means the rollers are being gently pushed up again the part as it’s being printed.
Here’s a video showing what I observed. Might be hard to see exactly what I’m talking about though because of the reflections.
About your comments:
… We don’t experience that defect with any of the five Form 2
Form3 has different laser (maybe more powerful) and different optical system, it might be possible that a parasitic reflection is causing this issue as I can clearly see a projection of the supports over flat surfaces (see some images above). The projections of the supports have very sharp and clear corners, if it’s diffuse light it will be fuzzy
…The defect only occurs in circular openings
It depends on the resin, object orientation and density of supports how intense will be the effect. The holes tend to “attract” and keep the resin inside because of the surface tension, notice that only one side of the hole is distorted (the one where the drop of the resin may stay longer)
Also it depends where was located the detail on the build plate - I found some comments in the forum which suggest the parasitic curing is stronger on the internal side (facing inside the printer).
Thanks for your thoughts and insight.
Three points I wish to emphasize:
circular openings are significantly misshapen, rectangular openings much less so.
edges of all openings and geometric features are somewhat rounded on the Form 3 (“mushy”) but not the Form 2.
changing build platform placement, reorienting models on the build platform, increasing support density and increasing touchpoint size have all failed to correct the wide spectrum of defects we get in Form 3 prints.
Unfortunately, parasitic curing as a possible cause of one kind of defect (misshapen circular openings), while interesting, does not explain the wide spectrum of defects we experienced with our Form 3, including layer shifting, soft detailing and surface bulging.
Reports made by others indicate that the defects found in our Form 3 prints are not unique to us.
Hopefully, the new firmware update will correct the defects.
It’s OK if it sags, the LPU will push up the flexible part of the tray and the part that is currently being printed is going to print fine.
If the film sags right over top of the LPU where the laser rasters, and you didn’t design it to do that, you would unexpectedly end up with >100um or whatever your layer resolution is.
If it was designed to sag and achieve a 100um difference when resting on the top of the LPU glass then you achieve your expected layer height, but then the leftmost roller would bump up against the recently cured area.
I might be missing something here as there are two layers of the film and what I saw is likely the bottom layer…but just calling out that it looked a little odd.
Yes, I have tried re-orienting the models, repeatedly. Yes, I am aware that the Form 2 and Form 3 are different machines. I am aware that the Form 3 uses low force printing.
*.stl files and photos sent to FL for evaluation. FL did evaluate them and provided feedback.
FL tech support acknowledges there is a problem(s) and reports that only a future firmware update is likely to fix the problem(s).
- Re-orienting models had no affect on print quality
- Adding more supports had no affect.
- Changing model position on the build platform had no affect.
- Changing model orientation on the build platform had no affect.
- Increasing touchpoint size had no affect.
All of this is true for two Form 3s we have had in hand. We get the same kinds of defects on models no matter how we orient them or which Form 3 printed them.
All of the models print perfectly on our Form 2s.
From a practical perspective, an iterative trial-and-error reorientation test for each of the 1400 models we sell would be a business production-crippling practice and not at all helpful or sustainable. No such process was required to get perfect prints from the Form 2s we own. Nor would it be practical or affordable in time and money and resin to test print each of the over 1400 models we sell until we can get a good print with a Form 3.
Hopefully, the recent firmware update is the fix we need. Will conduct additional print tests on the Form 3 with the new firmware shortly. I would very much like to report that the new firmware is a success.
Yeah not to sound rude but have you read anything in this thread? Because it is clear from the numerous posts by me and others that orientation, supports, etc… none if it matters. There is an inherent flaw here that Formlabs has recognized and stated they are working on. And unfortunately it is now just a waiting game for the fix, which has absolutely NO timeline at this point.
1/ There have been two firmware releases since these prints were done on 6th January. Both release feature ‘Various bug fixes’ which could mean anything.
2/ My understanding is that Formlabs was working on tuning grey resin @ 100μm before turning to other resins and layer thicknesses. Presumably they have found from their dashboard data that this combination of resin and layers is the most populars.
3/ It looks to me as if the resin is over-curing, ie the laser is passing through the working layer to cure resin where it shouldn’t be cured. If you like the laser is trying to to cure 100μm layer while you are trying to print 50 or 25μm layers.
4/ The Form3 will work with 0.2mm touchpoints. At 100% density there maybe more supports so there is a trade off between number and size of supports. In general the smaller touchpoint allow the model to be just twisted off the supports. (I’m using 0.3mm touchpoints @ 80% on many of the print on my Form2).
OK I know this last bit doesn’t help with your problems, but I think that that is entirely in the hands to the Formlab techs. Hopefully they will having it all sorted by the Form3’s first birthday.
Surface Artifacts on 2nd Print
I am the one who started this thread… and again… if you just check back in the thread you’ll see that for one of the files where I had issues, with MY OWN files, I made them available to others, and the problems were reproduced by others, over and over again. Not only that, they were reproduced in different resin types, at different layer thicknesses. Again… this is an identified problem by Formlabs with no identifiable cause, and no current fix issued, and no time table for when that fix may come.
So DoggieDoc83, where can I find the part where formlabs states this issues are already identified?
I’ve gone throught the entire thread but didn’t find where someone from formlabs statede this, or is this somewhere else?
I ask about this because my family owned company recently bought the F3 and we already had it replaced because the first one failed misserably to perform any part. I’m running a test print right now hoping to get a good result but with very little faith on it. If this machine fails to produce usable parts again, we will send this printer back. I feel so dumb because we sold an F2 perfectly functional to get the f3 and until this day nothing usable has come from it.
It’s right here… it wasn’t something where Formlabs officially came into the thread but it was the official response I got in my support emails with them. I just stated it and didn’t copy/paste the actual text from my emails.
I came here hoping to read about a solution to this problem. Thanks for keeping this alive.
Thank you, this is very useful. Currently we are testing the machine and results are not what expected.