Large Scale Central

Affordable SLA Printing

With 3d printing, one process is great for one preferred outcome, another for another.

FDM (Fused Deposition Modeling) is great for larger parts requiring lower fidelity and cost, and this is what most home-use 3d printers are. I have one, and have had good success while it was still working. Some people call it the “toothpaste printer,” because it’s laying down a tiny extrusion of melted plastic wherever the print head moves.

SLS (Selective Laser Sintering) fuses a layer of particles in a big tub, one layer at a time, resulting in a hard part in the middle of a lot of powder. This requires no supports. Shapeways is great for this; they know how to nest parts from various customers, print the big group, and clean up the big mess of leftover powder. This process isn’t really appropriate for home use, but Shapeways has made it accessible to us.

SLA (Stereolithography) is totally different. UV light precisely cures, layer by layer, a liquid resin. Very high resolution. And until recently, the process was nowhere near the realm of home use. Formlabs is a foremost example, but by the time you buy all their stuff you’re into multiple $K.

I learned from a friend last weekend that he was SLA printing, with a printer that cost less than $300. I didn’t believe it, even though I saw the nice parts he got out of it. But now I believe it. Several Chinese manufacturers are offering these printers, a prominent one being Anycubic Photon; and this is what he is using [link].

Recently, Anycubic has upgraded their SLA printer, and this is the one I’m considering [link].

Since this SLA printer would be for small / precise parts. I’ll still have a need for larger / structural parts requiring less detail. So I’ll be needing a new FDM printer, such as a Prusa, as reported by Joe, or the Anycubic versions.

Between those two kinds of printers (and it’ll be a while until I obtain them), I’m hoping to get away from the high cost, long lead time, and shipping cost/time required with SLS/Shapeways.

I’m not a pro, so take all this for what it’s worth. I do want to reinforce Joe Z’s point though, in that these technologies are becoming increasingly affordable for “the masses,” even high-res SLA.

===>Cliffy

Cliff,

looks interesting. if I had not already invested in a form 2 over a year ago would drop the funds in an instant. only downside is that it is of Chinese origin, so support is questionable IMHO. I have use the support functions with form labs such that I would want someone available to readily answer my questions. it appears there are a lot of resin colors available for this machine as well. build size is a little small but will work great for detail parts.

this should make a lot of FDM machines obsolete.

AL P.

I agree Al, thanks for your comments. I hear the Formlabs products are great, with all the advantages you mention. I have a friend who’s used A Form 2 heavily for production purposes (he sells HO kits), and swears by it. I’ve been thinking of getting one when / if I get the extra $3k (but that’s for a refurbished one; the new Form 3 starts at $3.5k).

My friend with the Anycubic though says their support has been quite good, and swears by it. And he has the earlier Photon (for $300); the Photon-S is said to be much upgraded.

I know there will be shortcomings with something costing 1/10 the price. Yet it will get me into the SLA market now, vs. a few years from now.

Cliff,

At this point I would go for a form 3. new tech there doesn’t use a separate tank for each resin. not clear to me how these knock offs deal with that. does the machine hold a tank full of resin and meter it out? , or are you limited to what you can pour in a tray?

I have had tray issues that were resolved by formlabs, and at least one cartridge failure when the machine told me the cartridge was worn out, again replacement supplied.

Looking forward to your experiences.

I’d look at the PRUSA SL1 printer, $1,400 also. Good price and great print results. Quite a bit cheaper than the Formlabs

Al Pomeroy said:

Cliff,

At this point I would go for a form 3. new tech there doesn’t use a separate tank for each resin. not clear to me how these knock offs deal with that. does the machine hold a tank full of resin and meter it out? , or are you limited to what you can pour in a tray?

I have had tray issues that were resolved by formlabs, and at least one cartridge failure when the machine told me the cartridge was worn out, again replacement supplied.

Looking forward to your experiences.

Thanks for your further observations Al. But I don’t have even $1k to invest, at least for some time. I’ll probably get the $400 unit and assume it’ll be a 2- or 3-year throwaway.

Thanks Greg for that Prusa info, I’ll have to look into that for a future buy.

So doing the math, and correct me if I am wrong, the super cheap SLA printer uses an LCD display, not a laser.

The display has a resolution of .047 mm

The Prusa SLA printer is .01 mm, almost 5 times better.

So the LCD instead of a laser is the clever bit that gives you resolution better than FDM, but way less than the typical SLA driven by laser.

There’s the price/performance tradeoff.

Greg

You’re right Greg, I’ve heard it takes hours to do a tall print for the cheapest Anycubic Photon, but the -S version has higher wattage / faster print rate. Both that and the Prusa SL1 are UV LED, if I’m reading correctly.

The cheapest laser-based SLA unit I’ve seen is this kit from Matterhackers for $1300:

https://www.matterhackers.com/store/l/peopoly-moai-laser-sla-3d-printer-kit/sk/MP86LX2U

Formlabs says they’re out of Form 2’s, and only offer refurbished ones for $2,500 and a 2-month lead time. Looks like they’re going all in on their Form 3, which is a great product I’m sure.

My longest print job at work was around 110 hours.

Wow! How tall, and what kind of printer, Chris?

Another factor is the resin. I haven’t gotten SLA prints (from Shapeways) in quite a while, but they were somewhat brittle. That’s a plus for SLS, where the parts can be gotten somewhat flexible, but very tough. I’ve been wondering if resins today can be the same.

I just saw this SLA resin, which claims to have those qualities (tough / a little flexible):

https://www.matterhackers.com/store/l/liqcreate-resin-for-sla-dlp-3d-printers-clear-impact-1kg/sk/MNKF9QY9

On be recent derrick project, I blew $400 in Shapeways parts for their “engineering plastic” process. But with this resin and an Anycubic, it looks like I could have made those parts for around the same amount – paying for the printer in one (admittedly involved) car project.

Production speed won’t matter much to me for a few years. If it takes a week, that’s still a week faster than Shapeways. Speed will matter more if I can ever get to producing kits, like my friend who’s buying the Form 3. But in a couple more years, who knows how much the laser SLA units will come down by?

Having said that, of course I’d love to get the Form 3, but there are too many other big expenses in the queue at the moment.

Cliff, you might have been able to make the parts for $400, but they would not be the same quality, so not quite apples and oranges.

Greg

Cliff Jennings said:

Another factor is the resin. I haven’t gotten SLA prints (from Shapeways) in quite a while, but they were somewhat brittle. That’s a plus for SLS, where the parts can be gotten somewhat flexible, but very tough. I’ve been wondering if resins today can be the same.

I wonder the same. Unfortunately, there is very little info about how SLA resins hold up outdoors. Formlabs has said their SLA resins are for prototyping - not finished products. This is interesting https://forum.formlabs.com/t/which-resin-to-chose-for-toys/22778/8

I just saw this SLA resin, which claims to have those qualities (tough / a little flexible):

https://www.matterhackers.com/store/l/liqcreate-resin-for-sla-dlp-3d-printers-clear-impact-1kg/sk/MNKF9QY9

That looks promising. Only resin I’ve seen that claims to be “weather proof” and perfect for outdoor applications! This is pretty cool https://www.youtube.com/watch?v=qycM8KDEaH0 but I’d like to see real-world tests of parts similar to G scale bells, horns etc.

This is why I use 3D printed parts as masters for molds. It is cheaper to cast X number of copies vs printing the same number. Plus, the castings tend to be a little more durable.

Greg Elmassian said:

Cliff, you might have been able to make the parts for $400, but they would not be the same quality, so not quite apples and oranges.

Greg

I agree Greg, the better SLA’s have better resolution.

But the comparison I had in mind was between my model (SLS “engineering plastic”, basically HP Multi Jet) and my friend’s HO SLA model from the cheap Photon. The latter had noticeably better detail.

Cliff

Craig Townsend said:

This is why I use 3D printed parts as masters for molds. It is cheaper to cast X number of copies vs printing the same number. Plus, the castings tend to be a little more durable.

Craig, a guy I mentioned earlier that has a kit business (Bitter Creek Models) does that. He’s graduating from the Form 2 to Form 3, and uses SLA to make his masters. Then makes vulcanized rubber molds, then white metal castings. I’d be doing it like you describe, SLA for masters, RTV molds, resin parts. At least for multiple pieces.

I have been using two formlabs resin consistently. one is called grey pro, which gives slightly better resolution that the regular grey. this I use for masters for spin cast parts and one off models. depending on the design the parts can be brittle, but experience has shown what minimum wall thicknesses can be achieved. the second resin I use is called durable. I use this for anything that has the potential to be handled or damaged in a derailment. it takes a very severe bend to break it. I first used this resin to make an outside frame for my Connie pilot wheels, and has held up like a champ. granted your experiences may be different. I model finescale 1:20.3 and my models do not live outdoors.

Al P.

Greg Elmassian said:

So doing the math, and correct me if I am wrong, the super cheap SLA printer uses an LCD display, not a laser.

The display has a resolution of .047 mm

The Prusa SLA printer is .01 mm, almost 5 times better.

So the LCD instead of a laser is the clever bit that gives you resolution better than FDM, but way less than the typical SLA driven by laser.

There’s the price/performance tradeoff.

Greg

Greg,

Both the Prusa SL1 and Photon S claim an X-Y resolution of .047mm, and a Z resolution / layer height of 0.025–0.1 mm.

Anycubic is dumping the non-S Photo, I suspect partially due to this needed Z res improvement.

Power to the LED’s is 25W for Prusa SL1, 50W (apparently*) for Photon S.

*They say 50W, but don’t explicitly say all that is going to the LED’s

I also wanted to look at speed, which is a little more tricky to compare but I wanted to do this regardless. Prusa says the SL1 will do 6 layers per minute, and they give the range of layer thickness. So it’s no big deal to do the math. The Photon S has speed listed as “20mm / hr”, but don’t say what layer thickness / res that time is good for. I have to assume it’s for the least res / thickest volume, but have just emailed them for clarification.

Anyway, assuming the worst in that regard for the Photon S, it appears to be 3 times slower.

But if Anycubic comes back and says, naah, that speed was using .025mm layers, (or anything better than .1), that would change the speed comparison drastically.

So for now, I’d say both do the same resolution.

The Photon S costs a third of the Prusa SL1, but is 3 x as slow.

Cliff, reading the prusa site, it said minimum layer height was 0.01, recommended 0.025 to .1

The thing that seems strange, and maybe it’s my ignorance, but the x and y resolution is “buried” somewhere and not stated.

The closest I found on the Photon was that the UV led screen was 2k pixels, assuming in the x direction, maybe .0575 mm?

cant find it very straightforward on the Prusa site, but a review said .047 mm…

Greg

Thanks Greg. Neither site makes comparison very easy…

On the Prusa SL1 page, their specs say “Size of a pixel (Resolution): 47um in XY axes”

The Photon S page refers to it as “XY DPI : 47um (2560*1440).”

So, .047mm for either’s claims.