Form3 evaluation


Thanks rkagear,
I did print them with quite a few more supports and got the same result.The layer shifting is related to where the supports meet the models.

I just did a print of the boxes at 100 microns and they were better, but there is still some rippling of the surface.I am guessing layer shifting is less likely at 100 microns because the layers are thicker.


I printed the ring from the other post .It was pretty good except for some layer shifting

here is one on the boxes at 100 microns.Better than 25, but still not a smooth surface.


So I printed the ring

with the holes at 25 microns and got the same ripple effect in the supports that others have noticed

I also reprinted the boxes at different orientation both parallel and perpendicular to the front of the build platform .I still have yet to print a box that doesnt have some defect .

If formlabs says this is due to lack of supports how can they explain this pritn.It would be difficult to add more supports to this model and yet there are still defects in the wall of the box

So at this point I am at a loss.My conclusions are if you want to pritna model with no supports on the main structure it will print fine.If you have a model that needs supports you may have issues.Unlike the form2 preform apparently is not good at orienting or placing supports, but even if you manually orient and apply supports you may still have poor quality prints. Not a good situation at all.


Latest experiment. I tried printing the boxes flat. I positioned them with the closed end directly on the platform and then I added supports. My theory is that the supports are contributing to the layer shifting and the parts I hav e printed with no supports dont show layer shifting.

I was worried about cupping and when I tried to print 3 boxes I got a cupping error, but only on one box which is odd. I was able to prtnt 2 boxes parallel to the build plate with the long axis of the box oriented parallel to the y axis.

here are the results. Much less layer shifting, but a weird pattern on the long sides of of the boxes and a small defect on one short wall of one box.I did get some sagging of the back of the box which was expected due to it being parallel to the build plate.

strange pattern on the long side of the box

defect in wall

sagging on bottom of box


Have you tried printing one upside down, directly on the base (no supports)? i.e. Cap resting on the base.


Sag on the bottom in that orientation is unavoidable and is simply the nature of the beast (SLA in general). The rest - defects and “shifts” in walls - is some kind of a Form 3 specific, as of yet unidentified issue.

Check this thread: Bad prints out of brand New Form 3


Thanks .
I have been following your post too.We all seem to be seeing similar issues.I am hopeful it is a software or firmware problem but I dont know enough about the physics of SLA printing to figure it out.
have you guys seen the resin dispense error? I get the impression the cartridge slot on the form3 is bigger than the form2 since my cartridge seems a bit loose in the Form3 slot, but nice and snug in the Form2.


Cartridge dispense error just means there was no resin in the cartridge, or rather, actuating the cartridge bite valve did not produce a corresponding change in tank resin level. It can also happen if the bite valve is not opening for whatever reason (missing slit) or if the resin is very viscous and takes a very long time to pool up in the tank.

It should not happen if the cartridge is full, the bite valve is fine, and/or resin is flowing

The cartridge slot is pretty loose on mine too.


Yes.Mine was fine initially but after about 500ml of resin I was getting a resin dispense error during the prepritning phase. I replaced the cartridge with a new one and the error went away.Well see what happens when this cartridge is half empty.

I am wondering if once some of the weight has gone from the cartridge the bite valve actuator is dislodging the cartridge upwards instead of squeezing the bite valve.


Yeah that sounds plausible. I was surprised to find out how much looser the resin cartridge fits into the Form 3 and initially though there was something wrong. No dispense errors from me but seems like yet another odd design choice.


I heard from the engineering team to add more support.For goodness sake I dont think they even looked at my post.Completely unacceptable response which leads me to believe they have no idea how to fix this.I asked them specifically to have the engineering team take my preform file and orient it and add supports to show me a perfect print and they ignored my request.If you look at the boxes where on earth could you even add supports.The model is failing at the point where there is the most number of supports .How would adding more supports-even if you could find somewhere to put them-actually make any difference.I already did a pritn where I added as many supports as I could and got the same result and I did a pritn with fewer supports and got the same result.Why do I have to waste even more resin to prove to them something I already know?


If anyone’s interested, I tried the box upside down directly on the build platform (with a vent in one of the sidewalls to avoid cupping). Here’s the form file.

More photos and details are on my Faceoff thread. Let me know if any of you try it as well.


Excellent.So your results confirm mine.With zero supports the form3 actually does better with boxes than with supports. And yet the recommendation from Formlabs was to add more supports? Of course there is a very simple solution here.lets have the Formlabs engineering team take the files we have all posted on the forum and orient them and add supports to prove to us the problem is related to supports and orientation.They can then post the files on the forum and we can print them. I simply cant understand the reason why this has not already been done.Formlabs is asking users with ZERO experience of using the Form3 to know how to orient and support parts when it is pretty clear that our experience with orienting and supporting on the Form2 is of no benefit . The alternative is for us to keep fumbling around and wasting time and resin and continue posting failed prints.I cant believe that is in any way good for the sales of the form3.


That’s normal, the Form 3 tank is huge, volume-wise, and fits about 500ml from what i’ve seen.
Once you’re down to 500ml, the cartridge is empty.

You can just dismiss the message and continue printing (unlike on Form 2).


That makes perfect sense because when I swapped out the cartridge at the direction of formlabs support the cartridge I took out felt empty even though I had only used about 500ml of resin.



@darbyvet The direct-on-base print still wasn’t perfect. There was some rippling on the sides - see the additional pictures in my link. The Form 2 had one severe layer shift partway up, while the Form 3 had less severe artifacts (mostly localized ripples) scattered along the walls.

I tried the direct-on-base print because I recall it mitigated / avoided a similar issue I had on the Form 2 with layer shifts and bowing on what was essentially a box.

All that said, I completely understand your frustration and share your sentiments to some degree. The print quality needs improvement, in particular to eliminate the layer shift lines many of us are seeing. The working theory that layer shifts are correlated to the tank lift and stir action which occurs every once and a while is interesting and might have merit.

I expect Formlabs is working hard on this in the background, but could do a better job of sharing visibility into those efforts. I wholeheartedly agree I’d like to see their team take some of the popular prints users here have reported problems with and reproduce/investigate/resolve the aberrations.


after how much prints optics degraded? Is it possible to change optics


AMEN , amen, amen.

I didn’t pay $3500… let’s call it $4000 considering resin, tanks, etc., to be a beta tester. And that’s EXACTLY what is happening. Several users here are FAR more willing than I would ever be to happily print and reprint using materials THEY BOUGHT to help Formlabs along. And I commend you for that, but man, it’s not right.

There’s no reason, as @darbyvet said, why Formlabs can’t be running their own prints to see what they reproduce. And maybe they are… but all we’re hearing is “add more supports per our engineering team”. And this is clearly not the issue. I would also be more accepting if it was happening on certain TYPES of prints or very specific orientations… but from my own experience, and as seen in MULTIPLE threads posted, it doesn’t seem to matter WHAT is being printed… round objects, square, oriented this way, oriented that way… whatever it is these shifting artifacts are there. Mostly related to supports, but even present when an object is printed support-less.

And as @Darbyvet said, being blanket-statement told to add more supports is just silly. Because at what point am I to know, on any given print, when I have added enough. Clearly the software is not to be trusted, as told to me straight from Formlab’s mouth… rather it should be used “as a starting point”. But I have to have SOME indication of what is good enough, and I can’t afford to randomly guess on every. Single. Print.

That’s. Just. Silly.


My optics degraded after 1 year of pretty heavy use.You can clean the window and mirrors.One of the benefits of the Form3 is that the optics is easier to swap out if there are problems.


So I heard back from Formlabs support and they said the orientation of my boxes was good and they suggested adding more supports inside the box. I have added more supports and also added a solid box and will do another test print.

I also printed a large figure on the form3 at 100 microns and it looks pretty good.I cant detect any layer shifting, but will have to prime it to be sure.So maybe the issues we are all seeing are related to small parts only?