Form 3 vs. Form 2 Faceoff!


#1

DizzyRuff, CuriousSnail and BrilliantBeast lined up and ready to go head to head. I’m running comparisons between my Form 2 and Form 3 to explore any edge cases that could use some tweaks in the latest product.

I’m only testing in Clear, for now, until my backordered Form 3 tanks arrive.

Feel free to suggest prints for me to try. I’ll pick out a few, favoring those you identify as coming out better on the Form 2. Better yet, if you have both printers, do your own comparisons and share the results.

3DMon Resolution Test

First discussed in this thread. Some early Form 3 users reported fused fins, and missing gaps between the diamond and adjacent shapes. Also blurring of the edges of negative features and surface quality issues. More recent firmware has improved things, and in my Clear test the Form 3 did significantly better.

# Printer Resin Layer Estimate Actual Preform Firmware
1 Form 3 Clear v4 50μm 1h 5m 1h 22m 3.2.1 rc-1.4.3-462
2 Form 2 Clear v2 50μm 2h 2m 2h 4m 3.2.1 rc-1.19.12-75

Gaps are accounted for:

The prominent outline seen around edges of positive and negative features on horizontal surfaces printed on the Form 2, is no longer visible on the Form 3:

Note the “nubs” near the top of the circle and diamond in the next photo are remnants from where I added three supports (0.35mm contacts for Form 3, and 0.45mm for Form 2).

Correct dimensions kindly provided by @Joonghyun_Cho

Here’s what I measured. Note I haven’t yet done any X/Y Fine Tuning on this Form 2.

On both machines, the part was printed directly on the base, with the fins colinear to the wiper/mixer path.

Here’s the form file and stl.

Tapered Holes Through Ring

In this thread, @TravisRogers reported problems with layer shifting and holes getting partially filled in, compared to clean results on the Form 2.

# Printer Resin Layer Estimate Actual Preform Firmware File
3 Form 3 Clear v4 25μm 4h 18m 3.2.1 rc-1.4.3-462 F3.form
4 Form 2 Clear v2 25μm 4h 15m 4h 36m 3.2.1 rc-1.19.12-75 F2.form

I saw similar layer shifting in my test:

But the holes looked pretty good:

I printed two rings on each machine, oriented 90° to each other around the Z axis. The layer shifts were present on both Form 3 models, but seemed a tad less severe on the one where the bulk of the ring was more in line with the path traveled by the mixer/LPU.

Syringe Bumper

This is my own design, and includes a tapered channel running through the model, ending with a diameter of 0.5mm. It snaps onto the syringe of a Voltera V-One and is intended to curtail nozzle breaks.

# Printer Resin Layer Estimate Actual Preform Firmware
5 Form 3 Clear v4 25μm 3h 25m 3.2.1 rc-1.4.3-462
6 Form 2 Clear v2 25μm 3h 15m 2h 55m 3.2.1 rc-1.19.12-75

Layer shifting was observed again on the Form 3:

Though the final part still fit nice and snug:

Geometric Shapes

A few shapes discussed in this thread.

# Printer Resin Layer Estimate Actual Preform Firmware File
7 Form 3 Clear v4 25μm 4h 39m 4h 57m 3.2.1 rc-1.4.3-462 RectanglePyramid-F2.zip
8 Form 2 Clear v2 25μm 4h 21m 4h 14m 3.2.1 rc-1.19.12-75 RectanglePyramid-F3.zip

The shelled rectangular prism had pretty significant layer shifting on the Form 3, similar to that reported by @darbyvet.

The pyramids are better, but shifting is still observed on the Form 3 parts if you look closely.

UPDATE Nov 26: I printed the rectangles again, direct-on-base this time with the shell walls pointing away from the build platform. I used 50μm layer height, and applied 0.5mm Z-Compression Correction. I added a vent near the base on one sidewall to avoid cupping, and chamfered one corner to ease removal.

# Printer Resin Layer Estimate Actual Preform Firmware File
13 Form 3 Clear v4 50μm 1h 3m 1h 15m 3.2.1 rc-1.4.3-462 RectOnBase-F3.zip
14 Form 2 Clear v2 50μm 2h 12m 2h 12m 3.2.1 rc-1.19.12-75 RectOnBase-F2.zip

Here are the results:

Dachshund

This post simply wouldn’t be complete without a canine entry. Amidst complaints of the Form 3 being slower than the Form 2, it’s interesting to note this print completed significantly faster on the Form 3. I ran the test twice to confirm the result.

# Printer Resin Layer Estimate Actual Act. 2nd Preform Firmware File
9 Form 3 Clear v4 100μm 2h 30m 2h 40m 2h 45m 3.2.1 rc-1.4.3-462 Dachshund-F3.zip
10 Form 2 Clear v2 100μm 3h 44m 3h 39m 3h 47m 3.2.1 rc-1.19.12-75 Dachshund-F2.zip

Some faint layer shift lines on the Form 3 print.

Thin Cross

An old test from back in the day. Another print where the Form 3 beat the Form 2 time by a wide margin.

# Printer Resin Layer Estimate Actual Preform Firmware File
11 Form 3 Clear v4 100μm 3h 04m 3h 20m 3.2.1 rc-1.4.3-462 Thin-Cross-F3.zip
12 Form 2 Clear v2 100μm 4h 29m 4h 33m 3.2.1 rc-1.19.12-75 Thin-Cross-F2.zip

I though the tall walls might be a good way to explore any layer shifting, but they both came out pretty good. The only significant shifting is toward the base on the Form 3 version. I’m not sure what’s up with the vertical “stripe” down the Form 3 print above - guessing it occurred during washing / handling.

Miniature Cups (added Nov 29)

This test recreates results similar to those observed by folks here.

# Printer Resin Layer Estimate Actual Preform Firmware File
15 Form 3 Clear v4 50μm 2h 05m 2h 22m 3.2.2 rc-1.4.3-462 MiniatureCup-F3.zip
16 Form 2 Clear v2 50μm 2h 12m 3h 08m 3.2.2 rc-1.19.12-75 MiniatureCup-F2.zip

The Form 3 print shows severe rippling / layer shifting on both sides of the cup. For scale, note the cups are about 2.5cm tall.

Notes:

  • Where time estimates were unavailable in Preform, they were read from the machine at beginning of print
  • Actual print times were taken from Dashboard print history
  • Brand new tanks were used for both printers
  • Prints were washed by hand in IPA (approx 15-20mins). No post-curing was performed on the Clear prints.

Form 3 blurred details and layer shifting
Print Quality Issue with Black Resin
Bad prints out of brand New Form 3
Form3 evaluation
Form3 evaluation
Form3 evaluation
#2

Formlabs have admitted that they have tuned the Form3 to print well at 100 microns, and that there are problems printing at 50 and 25 microns. So is printing test at a resolution that is known to have problems a fair comparison?


#3

In the mean time they have released several firmware updates to improve printing quality. So I think it is fair.


#4

I am having the same issues on my Form 3. Tons of layer shifting. Direct comparisons to identical prints on my Form 2 show the Form 2 prints to be completely free of these issues.

Here is my post about it with pictures: Bad prints out of brand New Form 3

I contacted support and was told that I don’t have enough supports. I am having a hard time accepting this considering how well these printed on the Form 2. I have been nothing but frustrated with the Form 3 and am regretting my purchase so far.


#5

Where did formlabs admit there was a problem with 50 and 25 microns.I didnt get any notification of that. I did print my boxes at 100 micron and they still had layer shifting.I tried adding more supports and I still had layer shifting. This machine is clearly not ready for re;lease and yet here we are wasting time and money on test prints to prove something to formlabs that they apparently already know.Thank goodness I still have my form2.I sell aftermarket parts for sci fi models and I cant use my form3 for those parts because the print quality is just not there. I can live with slower print times and/or firmware tweaks, but the Form3 fails at printing simple geometric shapes. Anyone used to a Form2 is going to be very disappointed with the Form3.

At this point I have used almost a complete liter of resin and have wear on my build tank and I dont have a single saleable print from it.


#6

Thinking about it, there are only really two ways these layer shifts can be happening if extra supports make no difference.

  1. The LPU unit is pushing against the prints causing distortion, if this is the case then Formlabs saying add more supports makes sense as it sounds like they know this might be happening.
    or
  2. The drive system which has different drive types for X and Y will have different acceleration - deceleration factors when sharply changing direction and if these are even slightly out will cause errors.

Can I ask, not sure if it’s just the way the light is hitting, but in this pic

The supports on the perpendicular print look more uneven?


#7

Great question! Here are a couple more closeups. The supports are identical in Preform. The layer shifts largely occur at the same height (makes sense; they were printed at the same time), but due to the rotation their effects predominate in perpendicular directions. In fairness that might contribute to one looking more severe than the other depending on the angle viewed. Examining the supports more closely, I did notice the two right next to each other on the parallel print have a pretty bad “wobble” in the middle.


#8

Thanks for the detailed write up. I also appreciate you noting all the specifics including the PreForm and firmware versions used in your comparisons. Our biggest opportunity to improve print performance will be through shipping subsequent versions so this is helpful to understand side by sides over time, strive for apples to apples, and I hope, give us an opportunity to show a positive trend through subsequent releases.

Your dog also kinda reminds me of Charlie the original Formdog:

image


#9

@JenniferMilne.

Do you have any explanation for the printing issues? Are there firmware updates being worked on to address print issues? It would be nice to know these things since I suspect there are form3 buyers like myself considering returning the machine because it simply does not work as advertised.


#10

Hi Jennifer,

Thanks for acknowledging my post, and the kind words. I’m hopeful this helps contribute in some small way toward your team isolating aberrations and effecting improvement. I’m looking forward to running tests in other resins, and would be happy to rerun several after future firmware updates to compare.

The dog does kinda resemble Charlie! I thought him when I took the photos. Fun fact: There’s a Golden Retriever named Charlie in the neighbourhood who’s pals with my pup. And his owner has a Formlabs printer.


#11

Deeply concerned about Form 3 layer shifting.

A Form 3 is expected to be delivered next week. Worried it will not perform.

My designs are small scale models that I sell as a primary business. They are all printed at either 50 or 25 microns on a Form 2 and generally the Form 2 does very well. If the same models are printed at 100 microns on the Form 2, surface smoothness and detail are too rough to be acceptable by customers.

For my business, we need the Form 3 to print models at 50 microns with no appreciable layer shifting. If not, I can’t sell what it prints. Perhaps, for the time being, a refurb Form 2 is the better choice.

I bought a Form 3 earlier this year but canceled the order after repeated delivery delays. After not being able to find a suitable substitute at a comparable price, I recently re-ordered a Form 3 last month. I was delighted to learn that the printer would be delivered the first week of December after just a few week’s wait.

But now, after seeing these posts and the significant layer shifting issues and possible inability of the Form 3 to print at 50 microns, I am very concerned the Form 3 can’t print the products I sell. I wish I had ordered a refurb Form 2 instead.

@Formlabs, please get the Form 3 working at 50 microns and fix the layer shifting issue.


#12

I hate to say it but, if I were you, I would be rightly concerned. I am in the same boat as you… I print mostly miniature… 3D models that I do myself and sell as finished products. The Form 2 has been a workhorse champ fo rthis stuff, but so far I have not gotten a Form 3 print that would be acceptable for a final product. I did hear back from support today but we’re still in the “working on it” stage.