Howie Did It – 3D Printing a Printed Circuit Board [video]
youtube.com
youtube.com
Building a custom PCB by depositing solder on FR4 with a 3D printer hotend is just ridiculous. And if they'd designed the schematic in Kicad instead of drawing the layout by hand with Inkscape, they'd have made far fewer mistakes and needed far less rework. You can get a dirt cheap professional PCB on your doorstep in under a week, and most of that time is in shipping - and if you don't mind spending a bit more for local labor, it can be a couple days. They've already got the CNC for the job too, so alternatively, if they needed really, really quick prototypes, then with a little V-groove bit and some copper-clad FR4 they could have just routed out the design instead of replacing the spindle with the solder hotend. I don't think they needed that turnaround, though, because this project could have been 100% prototyped on a on a test bench with some jumper wires and designed right the first time.
Is that a sort of Gell-Mann amnesia, in that I'm a decent PCB designer but a terrible mechanical engineer and fabricator, and I underestimate the difficulty of the learning curve in my own field, but overlook where they've wasted enormous effort on solved problems with their mechanical designs? Not sure.
Impressive work lads, but entirely unnecessary.
The first problem is attaching components without remelting those solder wires, they didn't show that. Are they doing surface mount or through hole? It looks like it would be for through hole only, but that might be okay for certain boards. But I would worry about over heating the parts in that it might ruin the traces.
The second problem is they didn't get a good enough squish off the hotend. It looks like the surface tension of the solder forced it to raise off the board. I don't know if a cooling fan would have helped here or not, but I suspect that .28mm distance is still too high. I would probably want a 0.10mm height.
I also wonder if there's a different approach that would make even more sense, like a conductive ink or uv-curable paint. It also might make an easy way to make vias and surface mount pads.
Here's one such source, but they seem super expensive.
https://novacentrix.com/product-category/conductive-inks/
Edited to Add: Oh and here's a system already doing just that:
And btw, this does harken back to the heyday of the 3D-printing hype where they hypists were saying that we would "soon" 3D-print entire electronic devices. That never materialized, of course. Maybe that was one inspiration.
Afaik, this seems like it's probably the easiest way of metal 3d printing at home - I wonder whether that's the more interesting thing about this achievement.
I don't really know much about in-and-out of melting metals, toxicity and so on but there are some lower melting metals like zinc at 420*C which could be reached with a 3D printing hotend with a resistive cartridge and copper block I think. That would be rather intersting.
The journey is the reward.
The board won't last the first few minutes of a rally course.
I do wonder if there is a practical reason to do it this way.
Custom PCB fabs are so commoditized at this point.
Is solder a good “wire” in this application?
Would neat wire bends on a breadboard, ala Ben Eater, serve the same or better purpose?
Maybe the turnaround on experimentation is faster when done this way? Definitely looks nice to be able to “set it and forget it”.
Either way I love the build and think it’s a super cool hack to solve their problem.