So I’ve been thinking of building my own 3d printer. Either SLA or DLP, as the hardware is becoming more available, and in many ways these are simpler than FDM machines
Resin tray, bottom lined with optically clear teflon film. A few are doing this, and it seems to last longer than the silicone coating option
DLP system? They are getting cheaper and cheaper, but still super expensive
SLA system? The Muve3D mounts their laser on a x-y carriage, but this limits drawing speed.
So I had a though of building a SLA system using x/y galvonometers as used for laser light shows. You can get them fairly cheaply online, or take them out of a laser light show box. They are driven by analog voltages usually, though some digital are around. I know dispersion might be a problem, but even the cheapos ship with optically coated single surface AL mirrors. Using a laser like this, you should be able to get phenomenal draw speeds. You’d have to tweak the path generation to prevent overshoot/overdraw, but thats just software.
The next issue is laser type. UV LED laser is possible, but I consider it the most dangerous because it simply isn’t visible. Many resins will cure with blue light just fine, so this may be a better option. As for laser loiter/draw time, I’d guess I’d have to test that out. Being near UV, it may cure fast even for low intensity laser light. Focused to a point, the photon flux even for a low power laser is incredible.
So am I nuts? Do you think a 3D laser printer using galvonometers is possible?
As for moving parts, the only one required is the Z carriage.
Well its gotten better, but miniscus in the resin causes defects judging by the print videos. Also wary of floating resin on water as most resins are moisture reactive.
Just a note: Point source deflection/reflection does not at all result in a rectilinear scan at the image plain. It results in a non-linear scan from center out along with a considerable amount of pin-cushion. There is a reason the high end SLA’s are so large and that is to keep as long a light-path as possible to minimize needed correction in SW. While a lot of these units exhibit prints of the small bird-in-a-cage and the Eiffel Tower, these are very non-linear prints. High res, yes, but again non-linear. Some new units promote a “max” X-Y resolution and this may hold true for the center of the image plain, but not for the corners. This is why the term “max” is used as a qualifier. Resolution falls off fast as you transition from the center out to the corner. While the mUVe may be a slower unit, it is accurate across the entire image plain. While the math is rather simple, it will be in firmware much like the Delta.
Although I wouldn’t recommend buying a pryntech machine. The laser, galvos, drivers etc are all straight from China. I didn’t ever find somewhere to get lens, iris and mirrors from though. I think @Anderson_Ta has got some leads on good mirrors.
I bought some of the cheap galvos and a control card, but never got round to putting anything together. I think it is doable, but you may end up with a slower scan speed to get decent quality, in which case it might be better to use a simpler setup like X-Y gantry.
Pryntech are yet another company that claim open design and say how much they love openness, but the hardware design is not Open Source. They even use the OSHW logo in their company logo, surely that is abuse of a trademark if I ever saw one!