Let's talk about '3D printed' circuit boards

If you haven't yet used a consumer-grade desktop '3D printer' you might still be under the illusion that they are going to 'change the world' as the hype used to promise. Someone once remarked to me about the dynamics of a hackerspace, "they come for the 3D printer but stay for the laser cutter". I certainly did, after fighting with the 3D printer that produced inconsistent, and poor, results, no matter how much effort I put into it. The next time I wanted the wonder of 3D printers I designed a model with OpenSCAD and sent it off to Shapeways and got my item in three different materials a week later. I saved a ton of faffing time, and money, that way. Then, if I wanted twenty, they would be the same quality and shape, and someone will be there to answer if they are not.

So on the heels of that hype arrived a slew of '3D' desktop circuit boards printers — Voxel8, Argentum, and Voltera to name some — which promise the same but for circuitry. The advantages, they say, are faster iteration of prototypes, extended material possibilities, and mechanical construction that isn't otherwise available.

It's easy and natural to get excited about these prospects. But there is only one case I can think of that a consumer-grade '3D' circuit printer justifies its cost: a hackerspace. Other than there, these machines are near useless.

How does a circuit bring-up work? We research the technologies and ICs that we may use. We'll read the datasheets and order evaluation and development kits to play with. If the circuit is simple we'll wire it up with a breadboard, cables, or an Arduino. There's no need for a custom circuit board at this stage. If the circuit is more complex, we'll need a board that can meet our needs -- this may be a 2-or-more-layer board with plated vias or a controlled impedance board with 0.5mm pitch devices. Crucially — and this is what those who promote these devices fail to mention — we want the prototype boards to be the same technology that your production unit. Otherwise, we may be doubling our work and spend more money.

"But you'll have to wait a week for the board to arrive!" It's called 'planning'. If an engineer cannot plan their project to fill their time between sending the boards out and waiting for them to arrive, then something's already wrong. There is so much more to do in a project to fill that time up. Save the two grand you'll spend on a printer you'd use twice for a 'quick-turn emergency fund'. Besides, PCB manufacturers are not sitting idle in the face of demand for quicker turns; most offer a day turn, and that should be good enough for prototypes in a bind. Or, find a local PCB manufacturer and save yourself some shipping time and money.

As for materials, I personally have worked with 'exotic' circuit materials that are production ready and well characterised. Even if I could produce novelty items with one of those printers that would look good on Twitter, how would I scale? How would I make them robust? I couldn't unless I re-do it with a different technology and have to go through a new learning process. As for new 'paradigm changing' constructions, I still haven't seen an example for this that isn't contrived. 

In conclusion I'll say that these consumer-grade desktop circuit printers could have a loving home at hackerspaces for educational and fun one-off projects. But please don't buy into the hype that they are all that useful elsewhere.

Saar Drimer4 Comments