PoisedWhale reporting for duty

My Form3 finally arrived!

A few observations and criticisms, hopefully i’m not duplicating what’s already been posted here too much…
In no particular order, and bear in mind some of these are pretty nitpicky:

  1. The glue Formlabs currently uses for the blue warning and instruction stickers is pretty horrible. I had to scrub it off the printer case and the top of the tank with IPA. Worse, the sticker instructing removal of the light engine retention plate actually rips apart leaving little slivers all over the inside. I’d strongly recommend changing the sticker/glue supplier.

  2. The side-facing power connector is nice when plugging the printer in for the first time (easier to do than with a connector facing straight back) but is otherwise a nuisance, with the cable either ending up bent 90 degrees if routed back behind a desk/cabinet, or cable running around the printer legs if routed off to the side (annoying when levelling the printer or moving it around)

  3. Light and sound alerts are a nice touch!

  4. Onboarding seems a bit like it was an afterthought - it doesn’t check for completion of any of the steps, it will allow you to skip pretty much any part of the setup. At least it did in my case.

  5. Post-onboarding, my Form 3 hung on “Initializing printer” forever, without moving any of the axes. It just stayed there for 15 minutes, and then i’ve had to pull out the cord. It initialized fine on next boot.

  6. No power switch or sleep/shutdown options is a serious issue. I have it connected to a power strip with a switch, but still, that’s super-clunky. Hope that’s fixable in software and can be implemented later.

  7. And speaking of no sleep, fan that’s constantly on and at what seems like max RPM is horrible for a device that’s apparently designed to stay on constantly. The fan can obviously be at least turned on and off via software, as it revs up mid-boot. Please implement a sleep mode and kill the fan when printer is not in use.

  8. And speaking of fan that’s constantly on - i’m mildly worried about dust ingress. The fan appears to be near the bottom and seems like it’s going to be sucking all the dust off the desk and into the printer.

  9. I really like the new cover design - seems to use WAY less work area, and since i’m in a very tight space, this is a very welcome change.

  10. Speaking of which, the new tanks take up more space (you win some, you lose some) but come with a sealable box which is great. No more stacks of cardboard boxes!

That’s about it for now, still didn’t get around to doing some prints and direct Form2 <-> Form3 comparisons, and i might not be able to for about 2 more weeks (business travel).

4 Likes
  1. Attempt to update firmware results in “Error uploading firmware” every time. Might have something to do with the fact it’s on a “neighbouring” subnet from my office desktop machine (192.168.2.0/24 vs. 192.168.3.0/24).

Edit: nope, happens when they’re on the same subnet as well
Edit2: manual firmware upload works

  1. Printer got stuck trying to connect to WiFi (perpetual “connecting”), then reported “Failed disabling wifi” when i’ve tried toggling it off. Subsequently, the wifi toggle froze in the off position until i restarted the printer. Managed to get it in that state twice so far.

Wireless networking in general seems pretty flaky.
Flakier than Form 2.

  1. New mixer and tank shape seem to produce way less bubbles in Grey resin. There’s still bubbling, but they’re large bubbles, not a ton of milkshake-like foam. The mixer and the tank geometry form a kind of a dam / reservoir on the right side, where the resin accumulates and levels out while filling, then the mixer swipes to the other side, releasing the accumulated resin and spreading it over the tank. The move is much less violent than the Form 2 wiper too.
1 Like
  1. Coming from DIY SLA with dual-Z-axis tilt-peel, Form 1 with tank tilt peel, and Form 2 with shear+lift peel, the peel on Form 3 looks seriously bizarre. After printing, the LPU scoots away almost all the way to the far right (“garage”) side, then the Z axis starts lifting and there’s a wet suctiony shlopt (not POP, not CLICK) sound as the film releases. The sound itself is barely noticeable, as the Z axis stepper is way louder than it. Then, for some reason, the Z axis continues to lift super SUPER slowly. So slowly i couldn’t notice it moving at first. It keeps doing that for about 15 seconds or so. Then everything resets and the next layer starts. The slow Z lift seems to only occur for the first layer.

  2. While it’s moving into position for the layer, the LPU (X axis) sometimes pulses slow-fast-slow-fast-slow-fast. No idea why.

  3. Laser isn’t visible from the edge of the printer/tank anymore, due to the geometry of the LPU and the tank. Your only indications of something happening are the noises and the movement of the LPU.

  4. There’s a third axis! The whole tank can tilt slightly. There’s a stepper on the right side somewhere that can pull the tank down or lift it up.

  5. The printer will print around 5 layers very quickly, by just swiping the LPU back and forth and moving the Z axis up in concert with the LPU finishing each layer. Then it’ll do a cycle where the tank gets tilted up, then LPU moves all the way right, then the tank tilts back down and presses against it, then the normal layer printing resumes.

2 Likes
  1. Stinks way more during printing. Consequence of heating the chamber with hot air, as opposed to heating just the vat with a PCB heater. Air comes in, has to come out somewhere. And when it does, it’s saturated with the resin volatiles.
1 Like

Thanks for all this great detail! Keep it coming.
Shame about the worse smell. Remind me, where’s the air vent? Any chance of jury-rigging a light-duty external exhaust fan and filter?

The fan vents are all intakes.
The exhaust is simply all the gaps on the case and the lid, as far as i can see.

Managed to do a quick test print (grey resin) before packing for trip…
Impressions so far:

  1. the surfaces are indeed smoother than on Form2; edges are nice and sharp - you could scratch or maybe even cut yourself on an acute-angled edge of a printed object with no fillet

  2. that said, the new LFS peel isn’t magic - you’ll still get layer lines and artifacts on thicker cross sections, as well as the usual odd surface artifact

  3. technical prints can still get slightly deformed, straight lines can still get curved, right angles can still end up oblique instead of right

  4. that said, you can get away with less support than before, and less forethought - i printed a few relatively tiny parts, but non-hollow, with 0 preparation and thinking, using the automatic layout and support, with supports reduced to 0.4mm, density 0.8, layer height 100um; I normally use those settings on Form2 for hollow objects with a 1mm thick shell, and i normally lay things out carefully by hand

  5. supports are still the same, their geometry is exactly the same as Form2’s, meaning you’ll still get warts on your prints - again, it’s not magic, it’s just an improvement over the previous process

  6. curiously, i got some layering artifacts where i normally wouldn’t expect them on a Form2 print (namely, at the very last layers of a (fairly short) object, where the cross section starts to get smaller, not larger

I took some quick snaps with my phone, I’ll try to post them later, probably from the airport.

Edit: I hope the new vats are more resilient to resin, as i’m leaving the resin inside it for 2 weeks until i get back. On Form2 i’d usually empty it back into the cartridge or into a jar, as the resin would inevitably eat and delaminate the PDMS, sometimes in as little as 3 weeks or so.

2 Likes

Thanks indeed for your thoughts! Many things mirror my own experiences, for example our Form 3 also hung on “initalizing printer”,

I do really hope that future PreForm and Firmware releases will take full advantage of the new peel process, higher laser power etc. - because right now, the speed and the auto-supports are a bit underwhelming.

1 Like

Back home after a 24 hour 4-leg plane trip.
I’m not sure what time or day it is anymore and weird ideas are popping into my head.

Like the fact that PoisedWhale means there’s a PoisedPorpoise somewhere out there.

1 Like

Okay, finally got around to uploading the pre-trip photos and doing another test print…
No comparisons to Form2 yet.

First of all, here’s how LFS deals with engineering parts.
This is the print from here, pre 1.3.4 update: PoisedWhale reporting for duty

The good news is, when the surfaces are clean, they’re REALLY clean:

The sort of non-news news is, you’ll still get non-square / bent / rounded vertices and surfaces in the direction of the peel:

The bad news is, LFS won’t magically get rid of layer shifts:

Finger for scale:

1 Like

Now here’s a non-engineering / non-functional, hollow, super-tiny part.
Printed at 100 um.

Check out the top of that main tower:

Also this tree:

The resolution / resolving power is definitely there, even though the edges and textures are a bit on the soft side.
I’ve noticed grey comes out almost NOT in green state and requires very little post curing, so this might be just an overcuring issue.

I’m not 100% convinced i’d get the same result from Form2, especially not with as little support. This was with 0.80 support density, 0.3mm touch points, only on the bottom of the castle, printed very slightly angled.

The model in question was hollow, 1mm thick shell.

You can find it here if you want to print and compare:

2 Likes

This was fixed in firmware 1.3.4.
Unfortunately, the LCD and the formlabs logo (front light indicator) still stay on - would be nice if those timed out and blanked in idle mode. I’m not sure, however, if it’s possible to blank the display and turn off its backlight, but leave the digitizer on so Form 3 can be ”woken up” via touch. It normally is on most displays, but who knows ¯_(ツ)_/¯

Another test print:

Something weird going on here, on the back of the head:

Those surface irregularities aren’t part of the model, as far as i can tell, and aren’t layer shifts.
Maybe the geometry was screwy.

50 um seems less overexposed than 100 um, but i have no measurements or quantifiable tests yet.
Again, i’d say the resolution is definitely not the problem, nor the laser spot size.

Edit: Apologies for the ugly fingernails.

Updated Preform, 50um (realized the update was for 100um grey too late), more test prints:

Printed in two different positions (middle of Y axis and beginning of Y axis - i.e. the place where preform puts it by default):

Why don’t you try @Joonghyun_Cho’s test print? from this thread Print Quality Issue with Black Resin - It seems like a good test model to standardize on for apples-to-apples comparisons of Form3 print quality.

Form file
https://drive.google.com/open?id=1Tw7IvSAeKpqEZEKYSgECHPFVKwIXqodf

STL
https://drive.google.com/open?id=1w6-CBlZY_1HJ9eU-QtMex1KBFGC6cujK

It’s already printing.

That model’s not quite a perfect test, though.
If printed on the platform, there’s a lot of overhangs (good for testing the peel, not so much for other aspects of the quality). If printed angled, you’re bound to get separation issues at thicker layer heights (even with a perfect resin profile).

Ideally, i’d like to reprint some test prisms with holes that we used to test the Form1+ and Form2 with quite a while ago, but for some reason, i can’t find the model anymore…

Edit: That test object is also a nightmare to wash, thanks to all the nooks and crannies.

Here we go:

Left to right - 50um slits along the LPU motion axis; 100um along; 50um perpendicular

All positive features i’ve checked measure as they should / dead on.

Stuff i’ve noticed:

  • the “3DMON” text is much clearer on the 100um print, suggesting 50um setting is overexposing more
  • the smaller rounded pillar is thinner at the bottom than it should be, then thickens by about 0.1-0.2mm - not sure if this is a peel issue, a compression issue (this was printed directly on the platform)
  • prints took 1h30m at 50um, 0h50m at 100um
  • perpendicular is worse than parallel - the movement ratio of X to exposure/Y galvo motion might need adjusting
  • there’s still slight overexposure everywhere, and those smallest slits should be cleaner, even though they aren’t THAT bad right now

Your results seem to be noticeably better than the results JC and I have been getting. Congrats. Anything special you’re doing or maybe your printer is a lucky one?

I’d be curious to see your results with https://pinshape.com/items/35775-3d-printed-hole-and-shaft-test
(You might want to hollow it out first, I didn’t)
On my system I couldn’t fit the peg in any of the holes (as mentioned in JC’s thread).

Anything special you’re doing or maybe your printer is a lucky one?

Nothing special, apart from a very thorough rinse (before and after 15 minutes of Wash). There was a lot of resin stuck in all the crevices and in between the various protrusions, even after the Wash cycle.

I’d be curious to see your results with 3D Printed Hole and Shaft Test by Zach Frew | Pinshape

I’ll give it a go.