Formlabs Form 4 Teardown
bunniestudios.com
bunniestudios.com
As I've written below this is a machine for businesses not hobbyists. You pay for service contracts, the overall "professional behaviour", a reseller network that helps you and whatnot.
I feel like they suffer from the same problem as Ultimaker: recalcitrance. They know they have a reliably engineered core product, but they're too intent on capturing sales without acknowledging the value prop of an incredibly competitive advancing market and are slow to innovate.
In my opinion, Formlabs is all about convenience. The fact that you can directly drop the build platform into the form wash is just so much nicer than having to directly fiddle with skin-irritating chemicals.
That said, my Form 2 is pretty low on speed, rather mediocre on reliability, and I keep having issues with it locking me out of the DRM-ed cartridges because the build quality of the valve motor is apparently rather bad. The printer thinks its dispensing resin while in reality it's not - because the motor is too weak - and then it thinks I drained 3L out of a 1L cartridge and it'll turn on DRM nag mode.
That said, the quality of the resulting prints is still insane. With a tiny bit of sanding you have parts that look like injection molding. I have never seen anyone get even close with FDM-ed prints. The transparency of prints is also so good, people sometimes think it's glass.
I have used both Form 2 in the past and now use Anycubic Photon line (one of the cheap Chinese brands). Main differences I noticed is
- Form software like the slicer is nicer and more polished and more features
- Support is obviously better with the Formlabs printer
- The Siraya resin I use in the anycubic ranges from anywhere of 1/4 to 1/8 the price of the equivalent proprietary DRM’ed Formlabs resin
- The end results of the two printers are mostly equivalent for my use case
- If you ruin something in the printer it’s a lot cheaper to replace it in the Anycubic.
In the end for my use case the Anycubic is way better value, but I could see for a business without in house expertise the support of Formlabs is probably a better pick.
When you're paying each engineer on your team $100k+, you don't want them spending that time fiddling with printer settings. You want to buy a $500 roll of carbon-fiber nylon and a $15k printer, put the two together, and have them call someone to have them fix it when it doesn't "just work."
The expectation that it just works (whether it does or not) totally justifies the price tag to an industrial customer's purchasing department which will be accustomed to 6, 7, and 8 figure price tags for CNC equipment.
Anything with "inspecting" or "composite" or "military" or "aircraft" doubles the price, so you can imagine what happens when you combine all four!
1) Resin stability and reliability
Sure, you don't care if you're printing miniatures. If you're printing a dental appliance, suddenly you care a lot.
2) Convenience features
Heating the tank, refilling the tank automatically, etc. Sure, you can do these manually, but they're a pain in the ass to do manually
3) Repeatability
The mountings on my Chinese resin printers need to be reset all the time. I have to flatten replacement plates all the time. I can't count on the fact that the plate removes and remounts accurately.
4) General engineering quality
FormLabs probably did real engineering, measurement and design on everything with the explicit purpose of characterization and repeatability (light uniformity, release tension, etc.) At $500 a unit, that just isn't happening on the Chinese ones. After the hobbyists poke at them a year or two fixing the issues, the fixes sometimes make their way back into the Chinese units (generally creating other problems along the way).
Nominally, this should all mean way less grief printing things.
And maybe the main LCD is no better, but I would suspect it probably is. As far as I can tell, there is exactly one supplier for the LCD panels in all the Chinese resin printers. It would be really nice if FormLabs actually spent the NRE to create a better LCD with a second source as it would give a competitive kick to the space.
For a hacker hobbyist, the Chinese printers are WAY better value. For someone whose time equates to money, the FormLabs may be worth coughing up as $5K really isn't worth thinking about.
I have a Anycubic mono x from 2020 and it is far more reliable, faster and cheaper. You can buy 5 of them for the price of the Form 3.
The software and firmware is pretty bad from Anycubic but you can figure that out a lot faster than you can fix problems with the Form 2.
I ran into three problems that I was not able to overcome, even after tons of trial and error:
- Warping. With ABS, a heated chamber, ABS slurry, heated glass bed, completely solved all warping. With resin, I was unable to predict or fix warping issues.
- Model not sticking to the bed. Behaviour was inconsistent. With a good printer like the Form, that is probably not a problem. With FDM, having an all metal printer, with a glass bed and a good Z probe solved this for good.
- Mechanical performance was garbage. Very brittle.
- Always needing supports. With FDM, if you fully control the design process, you can completely eliminate supports (at least in my case, of course this imposes design constraints). With resin, you always need supports. The supports are more complex and harder to remove.
I still think resin may be the future, because the level of detail with resin is just amazing.
Sounds to me you either aren't printing at room temperature, don't have the resin dialed in or you got a bad batch of resin/
The other criticisms are fair, but not necessarily dealbreakers and can be mitigated.
PLA is brittle as well, about the same as standard resin. Just like with FDM, there are other formulations, but the materials are different, and perhaps there are no good analogs for FDM materials.
Imo resin printing is mostly useful for tabletop minis, not engineering prints.
As opposed to you, I don't think resin's the future, certainly not in the hobbyist space. The fact that it involves highly toxic chemicals and fumes means you need a separate space and be really cognizant about materials safety at all times.
It does have its niches, particularly in mini printing, where the high detail is useful, and you can load up the whole area with minis at no penalty to print time and have an entire army's worth of minis printed in a span of hours.
In my experience with resin, you can get misshapen parts for a bunch of reasons:
1. Resin will change size a bit during and after curing. Not by a large amount, and of course it depends on the type of resin you use etc - but enough. And of course if your design has some thick bits and some thin bits, one might shrink more than the other.
2. Every time the machine exposes the resin, it sticks to the previous layer and to the projector screen/LCD. The machine then lifts the model, to unstick it from the projector screen. This applies force to the model, and can cause it to bend by a tiny amount - which can gradually build up over several layers.
3. Resin has to be somewhat UV transparent, for each layer to cure all the way through in a short exposure time. Because of that, UV curing later layers can pass through the part and cure resin that shouldn't be cured, on earlier layers. This is particularly the case if the shape has areas resin can pool up and won't run off.
> PLA is brittle as well, about the same as standard resin. Just like with FDM, there are other formulations, but the materials are different, and perhaps there are no good analogs for FDM materials.
For many printers, resin has to be brittle because of the way the printing process works.
You cure a layer, it sticks to the previous layer and the projector screen/LCD, then you raise it to unstick it from the screen.
If the resin is flexible, the unsticking process would cause the layer to bend, giving poor dimensional accuracy. So resin is formulated to be stiff and unyielding.
FDM can print flexible materials like TPU - common 'print upside-down' designs of resin printers simply can't. Although you undoubtedly could engineer a printer that could, such as a resin equivalent of an SLS printer, which doesn't face the unsticking issue.
Resin is also much less prone to size changes than standard FDM filaments. If you've printed FDM you should know how much tolerance you have to have in your parts (ex. hole diameters) to account for shrinkage post-print, not to mention the difficulty of printing things like ABS.
Resin has MANY drawbacks and in my opinion kind of sucks as a general user experience, but warping, material quality, and material options are not worse than FDM.
OK, you make a fair point. People will sell you such resin.
But look at the video at the 4m40s mark - the dimensional accuracy is trash.
The makers of these printers will all claim they have 0.01mm layers and 0.03 mm x/y resolutions and stuff like that but the dimensional accuracy of the parts coming off them is far, far, far below that no matter what the marketers claim. That cube barely has 1mm dimensional accuracy.
At that price point you might as well outsource it or at least print a negative and use a real silicone pour.
1. Resin shrinking tends to be uniform, which is something to be taken into account if you plan on making parts that fit into holes, like screw mounts.
2. Peel force is absolutely a problem, that's why you need to minimize contact surface with the print film. With 28mm minis, this tends to be not a problem, but for large models, its recommended you hollow them out. This saves resin and prevents partially cured resin from being trapped inside. Bending tends not to be an issue.
3. Again not a huge problem in practice. If you look at a sliced model, you will see, that below each layer the layer below is almost the same, with most pixels shared between the 2, and with overhangs that stick out by 1-2 pixels. This means that leaking light will likely hit parts of the model that are supposed to be solid anyway. But both underexposure and overexposure can happen, with the former meaning that resin doesn't properly cure, leading to thing features disappearing, and the latter meaning that light leaking will cure unintended resin, which leads to thin cavities disappearing. This can be fixed by dialing in the exposure.
4. PLA in my experience is just as prone to shattering as resin is. While truly TPU-like rubber resins might not exist (at least I've never used them), there are ton of resins with a slightly rubbery texture, which are solid, but are flexible enough that thin features don't break when you manhandle the models. Most official wargaming and board game minis tend to be made of a similar resin (which is probably not UV resin but has a similar feel)
I experienced every issue I mentioned first-hand, in practice.
I suppose it's possible I was uniquely naive in attempting to get engineering quality prints out of a consumer resin printer? It seems there's a good reason every resin 3D printer promotes itself with pictures of figurines rather than anything more demanding.
My experience was that getting good results when printing something like a plastic bottle cap wasn't just a matter of dialing in the exposure - I had to dial in the exposure, the room temperature, the conditioning of the resin, the support placement, the orientation, and design the item geometry with printing in mind. Even then the results were adequate rather than impressive.
Engineering prints are certainly possible I think, but you'd need some specialist resins.
Personally I'd stick with FDM for making functional prints.
Did you try engineering resins, like Siraya Tech's "Build", "Sculpt", etc.?
I make some small parts that I FDM print, then CNC mill for accuracy. Was hoping to switch to resin printing, instead.
Align the bed, clean it thoroughly with alcohol before printing and increase base exposure. Make sure there is no debris in the vat. It'll stick.
All printer resins are variation of acrylic and you have to design your parts with material limitations considered. However there's a range of premium resins for performance prints, look into Henkel Loctite or BASF offerings. We use BASF RG 35 and RG 9400 for small volume production parts. Also, make sure you do not overexpose: print an exposure target for any new resin to zero out the settings.
Supports are certainly unavoidable but are largely a skill issue. After hundreds of prints it takes me minutes to depanel very elaborate designs.
Crank the exposure time for the first 10 layers up. Way up, like 10x the exposure time for normal layers. They'll come out noticeably oversized, maybe 0.3mm bigger than you wanted. Adjust the model to compensate.
(Assuming you've already levelled the print platform etc in line with the manufacturer's instructions and your resin is OK)
- have GOOD ventilation and don't bother with water washables,
- nail down Resin XP2 Matrix test prints,
- use BOTH stirrer wash and ultrasound, and dry prints for hours before curing to mitigate swelling(marginal effects),
- and use 10-15x longer exposure for first layer curing time,
for SLA prints. Otherwise resin crumbles apart, leaves ugly marks on surface for first half inch, and so on. If magnetic beds are used, Z limit switch trigger arm must be adjusted too. Support-free is possible, people doing skyscraper towns are to some extent cargo culting.
PDF from TI from 2019, though I think they have updated their line up since: https://www.ti.com/lit/sl/dlpt019e/dlpt019e.pdf
Normal resin printers use a UV LED array with lenses to produce a more or less uniform light field. A display is then stuck underneath the vat to block light at certain spots. Since there is some distance between the display and the resin inside the vat, the resulting image has a slight blur on the edges; I think isn't an issue and is countered by using high resolution displays.
However, they consume a lot more power and require active cooling.
The display on a LCD resin printer has a limited lifetime, while the DLP is rated for 20000h. Being able to focus the image on the resin also allows for using arbitrarily thick (and thus stable) glas plates, while on a LCD the distance between LCD and resin needs to be minimzed.
Mind we're talking about a device that costs 4500US$ (before taxes).
Or am I missing some good reason to use a LCD display over a DMD in that device class?
>The past few years has seen a step-function in the competitiveness of products out of China, and SLA 3D printers have been no exception. The Form 4 sits at a pivotal moment for Formlabs, and has parallels to the larger geopolitical race for technological superiority. In general, Chinese products tend to start from a low price point with fewer features and less reliability, focusing on the value segment and iterating their way towards up-market opportunities; US products tend to start at a high price point, with an eye on building (or defending) a differentiated brand through quality, support, and features, and iterate their way down into value-oriented models. We now sit at a point where both the iterate-up and iterate-down approaches are directly competing for the same markets, setting the stage for the current trade war.
From somebody who obviously knows more about things like this than the general consensus (or lack of it) in places like Washington, DC.
No surprise, that's a place with very little ability to understand complex things, especially technology and its impact.
Formlabs makes some bold claims, eg. 99% of surface area within 100um of spec when using Precision Resin [1].
[1] https://formlabs.com/store/materials/precision-model-resin/
Ie: this [0] is a $300aud (maybe $200usd) LCD SLA printer.
[0] https://au.elegoo.com/products/elegoo-mars-4-msla-resin-3d-p...
Quality-wise, model detail is impeccable even on my years-old Elegoo Saturn.
For tiny models like table-top miniatures, then, I guess quality is subjective, but I certainly think it is more than 10x better. The numbers do not nearly represent the difference in surface finish and detail.
When you get a workflow down and need quantity, though, resin is cool. Since it exposes the whole bed at once, I can, for example, pull 24 small prints off of a small printer every 1 hour. 2 days of production, and I have hundreds, using just one printer. With FDM, this particular part, FDM couldn't even make, but if it could, it'd be more like 15 minutes per part - would need a small print farm to keep up with literally one $150 printer.
I don't even have a space for the resin printer right now, though. It really needs a dedicated and well-thought-out station to keep the toxic stuff from ruining your day.
Buying a "premium" model with proper support, a service contract etc. is totally fine for a business investment. I would get lots of raised eyebrows if I were to buy a "hobby machine" for 1/5 of the price.
So... There's a place for machines like these.
The differences between hobby-grade and professional-grade printers are less obvious at this point, more about reliability and throughput than they are about fidelity at this point.
A very succinct description of the Raspberry Pi past trajectory.
If you use complex devices like cameras, you can probably find recent drivers and readme files for rpi. For other SBCs you will probably end up trying to adapt documentation for rpi and getting stuck because your OS is using a weird version of libc, or something like that.
Nope. It was all about the subsidy.
Because the RPi was subsidized, competitors like the Beaglebone series couldn't achieve a critical mass since they were always being undercut. This set up a decade+ of mindshare.
And then the RPi foundation knifed everbody in the shortage by shunting all the RPis to commercial companies like FormLabs.
Even now that low-end x86 machines are cheaper than the RPi's, everybody still sings the praises of the RPi's.
Marketing always wins.
ARM MPUs became popular because they are actually easy to deploy.
The Foundation did no such thing because it doesn’t manufacture the boards and it’s not their decision.
The Trading company manufactures the boards, and it chose not to destroy its commercial contracts (all of which experienced limited stock too, particularly of CM4s, even though they were prioritising CM4s), which kept the lights on. They were already selling dramatically more units into commercial channels (Citrix for example) than to hobbyists.
The two entities also introduced the Raspberry Pi Pico and Pico W in the pandemic era, which were wildly popular and kept their educational goals on track, as well as giving suppliers like Pimoroni and Adafruit something to design absolutely amazing products around, with a microcontroller that wasn’t supply-constrained. IMO they more than made the case that most people don’t need a whole linux box to run a physical computing project.
I don’t get the whining. Do the little Intel boxes work for you? If so, use them and leave Raspberry Pi in the dust; there’s more than one way to add GPIO, SPI, I2C and PIO to your projects.
(Side note: the Beagleboard is a Texas Instruments product, is it not? I am not convinced its origins are all that different to the Pi, in that regard. If TI couldn’t compete on price and marketing, that is on TI, surely)
Back in the day (like 12+ years ago) I used some intel SBCs with GPIO, i2c etc. At that time they were nice because they just worked with your favorite linux distribution and put your binaries on it. At the time it was painful to setup an environment for cross-compilation. Today that's easy.
Raspberry Pis are never the best SBC available, and have some very closed aspects but they are usually available, have good community and corporate support and are supported for a long time. I've bought a lot of the other SBCs out there and after a couple of years you find that you're almost the only user left, and the vendor has 30 new products and don't support yours. Then one day you're trying to update your linux kernel for feature y but it breaks feature x, and searching for help yields workarounds and patches, or even OS images for the same issue... for rasberry pi. After a few times, you learn to stick to raspi, or maybe Jetson if you need more.