.png)
6 hours ago
2
Just as the laser printer delivered the benefits of a printing press to personal computer users more than 40 years ago, 3D printers have made it possible for individuals to turn digital designs into physical objects. Some printers cost less than US $1,000, and they can be used to create one-off objects or low-volume production of items.
3D printed objects have one major limitation; it’s not easy to make them “smart.” Adding digital processors and other components remains a challenge, as does adding conductive traces to, for example, detect when someone is touching the object. This is particularly true for the average hobbyist or even university lab—printed electronics technology exists that uses flexible silver inks and semiconductor materials, but the materials and machinery are way out of reach for such groups.
Oliver Child, a Ph.D. candidate at the University of Bristol, has been working on the problem. After a stint in the semiconductor business, he has found a way to merge his digital electronics experience with his love of 3D printing. He has created a process that allows someone to embed a physical microcontroller—such as an Arduino—in a 3D printed object. He calls the result “printegrated circuits.”
IEEE Spectrum spoke with Child to learn more about his work and his hopes for the future of the technology.
Oliver Child on:
- His background as a hobbyist
- The niche that “printegrated circuits” fill
- Experimenting with different conductive filaments
- His wishlist for future developments
- The importance of open source
I was looking through your website and it looks as though you came into this from the 3D printing side; is that fair to say?
Oliver Child: I did a computer science undergrad program and then went into the electronics industry, but I have always been a bit of a hobbyist.
I went into the semiconductor industry with a startup that had about 300 people working for it. We still couldn’t be competitive, because iteration was so expensive compared with establish products already on the market. Every time you wanted to make a new chip, you had to pay to license the software and then the tape-out fee cost several million dollars as well.
I have always had this self-sufficiency attitude, wanting to be able to do everything on my own. 3D printing gives me access to affordable tools with incredible precision right at your fingertips.
Obviously, it’s nothing like making semiconductors, but I can imagine a world one day where we can print all sorts of weird electronic things so that we can try out new things much faster without necessarily being part of such big structures.
When I look at your “printegrated circuits” projects, my takeaway is that this is an opportunity for a low-volume way to bring smarts to objects. Is that correct?
Child: Clearly there are many limitations with this technology, especially with the electrical resistance of the plastic and the physical resolution that you can get from off-the-shelf 3D printers and materials.
It starts with your Arduino projects that you’ve spent a long time crafting and you’ve got wires sticking out everywhere. And then you want to make a demo unit that you can show to lots of people and maybe give a few out to participants who want to study them. It needs to be low cost and it actually has to work. So this solution falls somewhere in the middle ground between individual Arduino projects and high-volume mass production.
There’s some really exciting stuff going on in this middle. For example, getting custom PCBs (printed circuit boards) in low quantities in now easy and affordable.
My go-to example is the TuneShroom, which is a mushroom-shaped MIDI controller. It connects using a USB-C connector at the bottom, and these black bits on the top are touchpads. Conductive traces go between the pad and a microcontroller that’s embedded in the base of 3D print. The PCBs used in these projects have through-holes that are designed to have component pins and copper wires soldered into them. With printegrated electronics, the 3D printing process pauses at a point so that the PCB can be placed in a cavity of the object. Then we use custom code that instructs the 3D printer to inject conductive PLA [polyactic acid] filament into the through-holes.
Conductive Filament Challenges for Printegrated Circuits
Was it difficult to get the conductive filament material to work?
Child: We had a bit of a learning curve, to be sure. At first, we tried just pushing the PCB pins into the plastic. It turned out that the connections were not reliable because any deformation of the plastic resulted in unreliable connections. And if you removed the component and pushed it back in, the connections were even worse.
Then I saw a 3D printer using two different types of filament material, and I wondered why we couldn’t do something like that. I found that you can get Protopasta Black PLA filament that is infused with carbon black. It has enough carbon to make the material conductive. Interestingly, the material is anisotropic, so its electrical resistance is different when printed in one direction than it is in another.
With a two-printhead printer, we could switch back and forth between the standard and conductive filaments. It turned out that a single-head printer wouldn’t do, because you need pure material when printing and extensive purging between filament changes wasn’t sufficient to completely clear the printhead.
Even with two heads, it took a lot of experimenting to get the injection printing to work reliably, and to minimize contact resistance between the PCB and printed trace. We worked hard on getting the right amount of material, the timing, how fast you inject it, and how much you have to retract afterwards so that you don’t pull the material out when the nozzle comes up. It took a lot of testing.
Have you explored using any other materials?
Child: We tried TPU [thermoplastic polyurethane] filament which is more elastic than PLA, and can also come infused with carbon. It’s really cool because you can use it to make pressure-sensitive devices. You can create a lattice, and when you squish it, more of the stuff comes in contact with itself, which changes the resistance.
We also tried copper-infused filament from Electrifi 3D, but it’s about twenty times the cost of the carbon-infused material. While it has much lower resistance than the carbon material, the resistance is still much higher than pure copper wire, so it’s not really useful for powering anything. I hope to do more experimenting with it, but it’s too expensive to make using it accessible at this point.
You mentioned Arduino processors. These are certainly popular with hobbyists and experimenter, but aren’t they a bit large for some applications, such as wearables?
Child: I’ve been using the XIAO series microcontrollers from Seeed Studio. These are Arduino compatible, but are only about 21 by 18 millimeters in size. They come in a variety of designs for different purposes, and they work with our material injection process to connect with the through-holes on their PCBs.
We’re using off-the-shelf 0.4 mm nozzles on our printers. These work fine with the 2.54 mm hole pitch on these devices. You can get smaller 0.2 mm nozzles for printers, but going smaller with a higher pin density is pushing things a bit tight at this point.
What’s on your wish list? If I could wave a magic wand, what would you want to move this process forward?
Child: I’d wish for a low-cost filament that has the same conductivity as copper wire. That would solve so many things.
My next wish would be for a way to use multiple boards or components in a single printed product. And we’re working on modifying a standard 3D printer so that it can pick and place solid components such as PCBs so that we don’t have to interrupt the print process to place them by hand.
Also, I’d like designs to be more transportable between different 3D printers. Right now, the inner settings of many printers are locked down and the user can’t adjust them. I’d like to be able to publish a design and have people be able to print it out on their own, anywhere in the world.
It appears that you’re a fan of open source.
Child: I’m a big proponent of open source. Currently, too many interactive devices are too complicated to build, making it difficult to make the same physical device somewhere else. It’s not worth the creator’s time and effort to document the entire process because there’s not much chance that others will be able to replicate the builds.
I think we’re going to see more people finding it easier to share their designs. If we can abstract our work in a digital form, that would make it easier to share and modify if you want.
We’re currently running a big user study, getting lots of people to print their own units of a certain design. We’re working with maker communities to see if we can share designs and have others make them on their own printers. In time, my hope is that people will start incorporating functional materials in their open source designs so that anyone, anywhere in the world, will be able to print these interactive devices, and even modify them for their own specific needs.
