3D Printing Circuits Gets Rid Of The Box Altogether

Many think that the next big step in 3D printing is when we’ll be able to print in metal, well, at an affordable rate. But what about printing in metal and plastic at the same time?

The thing is, most electronics are typically two-dimensional. Layers upon layers of relatively flat PCBs make up the brains of every bit of technology we know and love. The funny thing is, we live in a three-dimensional world, and we like to shove these flat circuits into three-dimensional boxes. Well, what if we didn’t have to? What if the circuit could be embedded directly into whatever shape we want? It’d be pretty awesome — minus the whole servicing aspect of the product…

Anyway we’ve seen some great hacks over the years attempting this, like adding a copper wire strand into your 3D print, embedding components into your print by pausing the job, or even going old school and using the point-to-point Manhattan style circuit construction to add some electronic features to your part. But what if your printer could do it for you?

That’s exactly what Optomec is attempting with the Voxel8 3D printing electronics platform. It is your standard run of the mill FDM style 3D printer, but it has a 2nd extruder that is capable of squeezing out liquid silver ink that dries at room temperature. Just take a look at this quadrotor they were able to make.

http://www.youtube.com/watch?v=zbm2SSql8V8

It’s foreseeable that you could do this with any multi-extruder 3D printer, but current software isn’t really equipped to add in three-dimensional traces to your part… until we heard about Autodesk’s latest foray into the world of 3D printing that is!

It’s called Project Wire and it’s being developed in SPARK, which is Autodesk’s open 3D printing software platform. Yeah you heard us right — it’s actually open. Now is the time to get giddy. Project Wire allows you to easily add three-dimensional traces into your model, not to mention components — oh and multi-material slicing, active pause commands so you can add features mid print… Awesome.

3D Printed Thumb Drive Schematic

Project Wire is set to release with the Voxel8 sometime later this year. But it’s foreseeable that it could be used with other printers as well, since it is not Autodesk’s goal to be proprietary to a single brand.

[via Makezine]

84 thoughts on “3D Printing Circuits Gets Rid Of The Box Altogether

  1. No no no! This is such a showpony product, think makerbot, add solder extruder.

    Ive actually been telling people all over the internet to check out the FirePick Delta (on hackaday.io) to see what a proper printer that can ‘print electronics’ looks like, the ‘voxel8’ will need you to stay next to it, to add parts when needed, the FirePick could work without needing additional input (a user placing parts in a half printed object)

    1. I imagine most regulars maybe shouldn’t be looking forward to this kind of manufacturing becoming commonplace.
      I’m thinking discreet components with unlabelled traces, embedded in plastic so as to be inaccessible and unserviceable.
      It’s cool tech and all, and I don’t mean to be a downer, but I only see this as assisting planned obsolescence.

      1. I agree. Taking “no user-serviceable parts” to a whole new level. I thought it was a pain getting an AirPort Express open to fix the power supply, but at least it wasn’t embedded in a piece of solid plastic with traces scattered throughout.

    2. If it saves me $200 on the cost of the printer to avoid an extra vacuum-tip head for the pick-and-place, I’d happily have a multi-step print with manual insertions. It’s still significantly easier than soldering components of that size.

      Imagine using BGA parts with this. No more careful placing and heating. Pop it into a mold and press a button, watching your matched-length traces route to the HDMI port you placed earlier.

      1. “route to the HDMI port”

        Ahahaha, I can’t see this sort of printer giving you the quality of trace geometry needed to deal with high frequency stuff for several years at best.

    1. It’s not open source, it is an open api/platform vs a closed api. These are terms that have existed for a long time. So it’s documented and available to anyone who wants to use it, vs a closed system where its can be behind a patent/paywall or obfuscated

        1. yeah i wish it were so in reality, what there are some really hard to RE api’s that you can’t get all the info out of without massive amounts of work that might eclipse what you wanted it for in the first place. so enough work to make them effectively closed.

          trololololo on

        2. Tell that to people who have had to deal with all the Windows API functions Microsoft won’t disclose, while using them for Microsoft products in order to have some advantages not available to other companies writing software for Windows.

          Then there are the many single source systems where the hardware, its OS and all the software are either produced by one company or any 3rd party software is tied up with a “no sharing any of the information with anyone, lest we sue you into oblivion” NDA.

          And right here on HaD there have been a large number of chips and devices where interfacing information is not provided to anyone, anywhere, not for any amount of money. So people go to work poking and prying and attempting to figure it out.

          One example (that nobody bothered to suss out beyond the basics) was the video chip used in the Dell Axim X51 and X51v PDAs. Someone got Linux booting on them, another person had Android sort of working, until (IIRC) 3.x was too bloated for it. But with only the basic 2D frame buffer access known, performance was terrible, thus there wasn’t any interest in wiping Windows CE off and flashing Linux or Android. Intel would not release the full GPU documentation, no matter what. Closed, chained up, locked and welded shut API. An X51v with its 480×640 screen and support for an external VGA monitor *would* have been a pretty nice Linux or Android PDA – a few years ago.

          1. Ah, the good old M$ mystery API calls again, so super secret only M$ can find them.

            (Of course the dog just ate the piece of paper you had them written down on…)
            (Hardware ‘secrets’ aren’t too much dissimilar; it’s either ‘we can’t be arsed releasing the docs for the 3 lamers running Linux’ or more likely ‘buy a million chips and we’ll give you a copy under a NDA’)

        1. You do rais an interesting point, the sample print is featured is pretty much a 2D circuit. If the traces did a loop-the-loop or something, it would be a much better example of being capable of printing circuits in 3 dimensions…

  2. Let the new era begin. PCB (Printed Circuit Board) will replaced by PCC (Printed Circuit Cube). Three in one machine (3d printer plastic+3d printer metal+ pick and place) will be the main equipment in the electronics industry. We live in an interesting era.

  3. I don’t think I’ve ever wanted to burn an “electronic” more than I do right now. That thing looks terrible and their approach to the idea isn’t the greatest. How could you ignore the pick&place functionality? Without that, it dooms the “product” to the simplest of circuits, the ones that you don’t need a machine made by a team of people that costs a lot of money to make.

    1. Because sometimes you have to walk before you can run.

      Pick and place can become an addon. Pick and place can be part of version 2.0 or 3.0 or whenever they get to it. You can still use it today without all the extra complexity or cost that they haven’t yet developed; you just have to add the deadbugs yourself. For a hobby printer, it’s a great next step on the way to a fully automated circuit builder.

  4. Nice idea. Looks like it won’t be production ready for a while.

    I’m also not totally convinced about the need for this. There are all kinds of bendable, printable and wearable pcb projects going on around the world. Maybe my imagination just isn’t in too good shape, but I can’t think of any real killer applications for this. What do you think? Does this have a future outside tinkerers workbenches?

    1. The concept of designing a circuit in 3-d instead of on a 2-d PCB has great potential to shrink circuits. The “killer application” would be any complex electronic device that could benefit from being smaller. There aren’t many devices that couldn’t benefit from a 3-d circuit, if the technology were usable, workable, and inexpensive.

      That said, there are certainly issues. Heat dissipation is one, but it’s not necessarily a problem for every electronic device, and with good enough 3-d manufacturing technology, one could even embed passages for liquid or forced air cooling in the device. Maintainability/serviceability is another problem, but many circuit boards are not economically serviceable today, so continuing down that path won’t always be a show-stopper.

      Mass market manufacturing costs may confine this sort of idea to tinkerer’s workbenches for a while, though. We’re already pretty good at using machines to produce printed circuit boards, and then populate them with components, all in a very fast, automated, reliable, low-cost way. And modern circuit boards can have very many layers, which at least partially allows one to work in more than two dimensions.

      1. I expect to see 3D in integrated circuits first. Today, if a circuit on a PCB is too large it can (with some constraints) be integrated in a piece of silicon. Shrinking ICs, however is not that trivial any more because (AFAIK) they can’t work reliably with traces and elements below 5-10 nm. That is why they need to be made (sooner or later) 3D.

        1. ICs are already made of multiple layers; last time I read something on the topic (which was a very long time ago) it was a discussion of a 17-layer process. The problem is that every layer added increases the chance that a chip will have a manufacturing defect, and there are heat dissipation problems if a junction is covered with more material. While wafer yield info isn’t generally made public, it’s enough of a problem that it makes a big difference in chip cost.

    2. One killer application that came to my mind: disposable kids toys, such as the kind given away by fast food restaurants.

      I think the technology, which is definitely promising, needs to be developed a long way before anything more serious will be feasible.

      1. Maybe I’m just not visionary enough, but it’s hard to imagine any sort of 3-d printing technology competing with injection molding for low-cost mass production of cheap trinkets like giveaway toys. Injection molding ties up the expensive machine for very little time each time a gadget is produced. 3-d printing ties up the expensive machine for a very long time for each gadget produced. 3-d printing may have cost advantages for prototypes or low volume custom objects, but injection molding is great for mass production of the plastic enclosure, and conventional PCBs are cheap for the types of electronics that go into toys.

        I’d see it first having practical applications on the opposite end of the scale — highly complex, low volume gadgets, especially if they have unusual constraints on volume. Things like electronics to go into satellites come to mind.

        We’re in agreement that the technology needs to be developed a very long way first.

        1. Well, if I can print my own kid’s toys with files I got from a torrent then the injection molding argument doesn’t hold water. McDonald’s isn’t going to print a toy next to the fryer. It’s like the CEO of Foxconn saying a few years ago that 3d printing isn’t going to change anything, we have been doing it since the 70’s. Well, he is wrong and this is a step in the right direction to prove him wrong. He was thinking about his company printing parts/products for you to buy, not the fact that we will be printing these things for our selves.

          If you thought the MPAA and RIAA were monsters, just wait until Foxconn, Apple, Samsung, etc are all clamoring for piracy to end because their bottom line is hurting.

          Some one needs to start this, more people need to step in with ideas, sooner than later this will catch up with the big boys.

    3. on your keyring, hybrid keys with both mechanical and electronic security

      kitchen appliances like stirring spoons with built-in sensors for temperature

      smart phones that are actually durable and will withstand falls onto concrete

      decent prosthetic devices instead of the terrible hacks we have today

      laptops that are much thinner and lighter

      automotive hoses and manifolds with sensors built right in

      i haven’t even had my second cup of coffee yet, I could go on and on

  5. Honestly, for any sort of signal integrity, I’d sooner just drop in an entire PCB as they did in the quadcopter at the end of the video. I can’t think of very many things I could do with such coarse pitch on the traces.

      1. You do have such a signal in a computer or smartphone as you suggested above, also, serial connectivity has uses high rates now days, so I guess than in a quadcopter you need to have reliable signals, and those traces are no good at all.

        Also 3D printer layers are no match to normal PCB manufacturing process with multiple layers, we are talking of micros, while 3D printers talk in milimeters.

  6. Here’s a novel idea… why not completely seal your project inside a 3d printed case with a LiPo battery and wireless charger… completely water and dust proof! just so long as it contains some sort of radio transmitter… or a translucent casing if you need a display on it ;-)

      1. Agreed, it’s a direct clone of the circle jerk over at reddit’s r/futurology. “Frigg off with the hype,” I say, show me something amazing like a pick and place machine that lays fine copper traces. With free software and free plans to build. I expect things as inaccessible as this elitist piece of shit out of a Harvard lab.

  7. If future corp’s were to improve and adapt this methodology. On one side of the coin It would make many items smaller more compact and cheaper to manufacture Since the build would be one run instead of multiple component runs and then assembly atop that. Plus, (and this one is the pro/con depending on the perspective) any disassembly of the device would more then likely break or render the device inoperable.

    However, On the reverse side of the coin,

    This methodology would make most items created by it remain disposable at best. Meaning if any internal component failed or otherwise ceased functioning correctly the device would be operable or unrepairable and thus remaining a over priced paper weight.

    Sure to a corp that means more sales less it fails to often. it also would eliminate repair teams inside corp’s but would also just be a full replacement in a warranty claim. profits go up then they come back to earth….

    Would make nightmares for Tinkers/reverse engineers, not impossible of course (acetone bath comes to mind) but not nearly as easy or sightly afterwords…. Tho were resourceful devils, If there’s a new manufacturing security invented, there will always be a vulnerability exploited and a hack invented to follow suite. ” Necessity is the mother of Invention”

    So Lastly in my opinion I believe this would only show it’s face in cheap novelties and large corporate a**hole devices like gaming consoles….. “The last part just made me think about throwing a controller and not being able to repair it.”

    1. “if any internal component failed or otherwise ceased functioning correctly the device would be operable or unrepairable and thus remaining a over priced paper weight. ”

      Well you pretty much describe just about every laptop and tablet computer built today with existing technology!

      In fact a 3-D printed device can be MORE repairable because there will not be a big huge circuit board inside the device that can’t be removed without destroying the device.

      And really when you return your broken remote control or your broken digital watch or your burned out power supply, do you really think that they open these things up and repair them? NO, they put them in the bin to be sent off to a poor country where poor people can get cancer from the chemicals they use to separate and recover the materials. With a 3-D printed item there is no circuit board, just some plastic and metal and you can just toss it right into the recycling bin along with your cans and botlles.

  8. They say they want to combine circuit and houses as one unit. What kind of stupid idea is that? I can pirnt my Drone and then have to reprint the whole drone because I wanted to change anything on my circuit? The idea of being able to remove the circuitboard from a housing and have design mechanic and electric seperated is a good thing.

    1. This is what I thought too. I suppose you could design the plastic portions to have a break-away structure so you could retrieve the expensive electronics when the rest of the device was toast. I guess this would be great for making the “trinkets” that the CTO dismissively called the output of current 3D printing into “electronic trinkets” but it seems farfetched that this particular technology will be useful for very complex, electronically sophisticated devices.

      1. Prototyping, experimentation and low-volume custom devices are possible good matches for this technology. As described in a post above, mass produced devices like TV remotes won’t be until the per-unit manufacturing time [cost] falls below what it is today using existing, mature manufacturing technologies.

        1. Unless you are the type that design and simulate your project and 95+% certain that the prototype works the first time, you are not likely to want to use this. i.e. pretty much your average “makers” that uses breadboard have to change their iterative way of tweaking/learning/debugging cycle and kiss their breadboard goodbye.

          Of the ones that do their homework and get the design working the first time, they might as well make a PCB, so this “printing” toy is of no particular use for them.

  9. These toys are supposed to target to people that aren’t ready to put in the effort to design everything correctly upfront in one go. i.e. people that breadboard, learn and tweak their designs as they go. By having the printer print a big lump of circuits into plastic means that the iterative process no longer works because you would have to have 100% confidence that everything are going to work first time (or 2nd time). That requirement alienates the user base.

    For the group that do the design correct and have the confidence to commit to a built, they would just make a batch of PCB and a 3D printer with pick & place option would be way more useful to them.

    1. Stop spamming. Your reading skill is lacking. No where I talk about consumer grade throw away electronic.

      What a way to completely miss your target market that have to breadboard their projects and people that don’t bother reading a datasheet and do their math.

      It is the end of endless tweaking your design because now you have to make sure that they work 95+% on your first try. Having to wait a few hours standing next to a slow printer and manually placing parts is going to suck if you have to do a dozen tweaks just to get a simple circuit to work.

  10. I have an Atari with a power brick… literally. Atari thought it was a good idea at the time to encase the PSU in solid epoxy and mold a plastic shell around it. The thinking, at the time, was that the epoxy would act has a heat sink and effectively wick heat away. No one really considered failure when (not if) the overheating components needed to be replaced because the epoxy turned out to be an insulator. Non-potted replacements are available but besides the point.

    I really want nothing to do with this. We already have problems being a throwaway society. Do we want to continue moving the entry bar for newcomers to this hobby, doing more to expose less to future hackers? Yay… let’s move to make the world a little bit crappier for the next generation of hackers.

    1. ” We already have problems being a throwaway society. ”

      With a 3-D printed device you can just chuck the whole darned thing in the recycling bin, no need to send it to a third-world country for disassembly

  11. 3d circuits are fairly common in the automotive world, some embeded in plastic usually formed at the same time as the connector, some free floating with spot welded joins, VAG electro hydraulic power steering pumps are a good example, had to replace the can tranciever in one recently and it was a huge pain in the arae. Ended up just buying a used unit and re-coding it.

  12. Screen print PCB design onto Pyralux. Wash with ferric chloride. Add additional layers by gluing together with Super 77 spray adhesive. “Drill” holes by burning with a chinese bought laser cutter. Ever try using Risograph master foil, burning your design into the master using the lightest raster setting on your cheapo 40W laser cutter, and using that as a disposable master for your screen printing process? This process is way easier than toner transfer and way cheaper than wasting money on fab houses.

    Enough with half cocked solutions. Let’s quit giving commercial guys money to perform “professional” jobs. This is Hackaday not plug your commercial product hour.

  13. If you’re worried about closed-source software, solve it the old fashioned way.

    Find the author and convince them that they’d much prefer their health over keeping the source code secret.

    *shrugs* It works for the Mob. It worked for Taiwan for decades.

  14. Instead of conductive ink, why not print the circuit with a low melt metal like “Wood’s Metal”? Ink seems like a weird choice when you have 3d print head that is more than 100F hotter than the melting point of these alloys.

  15. Good idea, the only snag here is that the low melting point alloys have a tendency to fracture.

    59C alloy (BiInSn) is handy for removing stubborn chips without damage to the board though, rescued some expensive hard to find 74C926 chips from the dead boards they came from.

  16. Very funny article…. wait, am I supposed to take this seriously??? Whoops.

    The “research engineer” at 0:27 obviously has no experience in PCBA manufacturing. Apparently he has never heard of flex hybrid PCBs? And he says “this whole paradigm will change”… Really? in 10 years, maybe, but with this sloppy Voxel project, not a chance.
    I assume they expect companies like Altium to support this by creating proper design support software like schematics, layout, and simulation tools so that you are 3D printing circuits that don’t work because your design was done visually? That may work for a USB powered LED, but it is severely limited to anything really useful.
    Another killer was the comment by the CTO of Autodesk saying that “we will be printing things that can ONLY be printed on 3D printers”. Umm, yeah, that future showed up a few years ago…. then again, maybe not with the Voxel.

    I really only decided to post because I wanted to let the Hackaday folks know that while I use to visit the site daily, I barely hit it twice a week now. It only takes me about 5 minutes to get caught up, when I used to spend ~30 minutes a day on the site. Kind of tired of the same old junk or latest “cool project with an Arduino” that has already been done previously by other microcontrollers. Apparently, if you didn’t make it with an Arduino or a RPi, you never actually made it. The ATTiny music player in the Jewel case project was the first post in a couple weeks that I actually read completely. That’s the kind of clever and well executed projects that should be on this site.

    1. This could well be useful even in a crude form, it could be used in auto repair for instance, or on the ISS to patch things with remotely supplied designs (I hear they recently did a repair for the first time with a 3D printed part)..

      But sure it needs to develop and be copied and improved by various parties.over time.

    2. You would think they have a standard procedure for doing emergency patches for space mission. NASA have very strict guide lines down to bending component pins, soldering etc. Things also have redundancies.

      To think that they would use an unqualified process with less reliability than a dead bug is very funny and taking advices from non-qualified armchair “engineers”.

  17. This is something that indeed needs to be developed and be available, but I agree, it’s going to be annoying too when you can’t fix things easily so it’s best reserved for specific cases.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.