Conductive filaments and printing solder are one thing, but what if you could spice up your 3D prints by embedding wire right inside the filament? That’s what [Bas] is doing, paving the way for printable electronics, PCBs, coils, and odd-shaped antenna.
The general idea of [Bas]’ technique of embedding thin copper wire inside a single layer of a print is to lay the wire down in front of the nozzle, effectively turning bare wire into insulated wire in whatever shape you can imagine. The trick, however, is figuring out how to put wire down in front of a nozzle. [Baz] accomplished this with a slew ring turned by a stepper motor connected to a 5th axis on the control board.
There are a few things this prototype doesn’t cover – cutting the wire, connecting the wire to components, fine-tuning, and a host of other things that prevent [Bas]’ machine from building real functional circuits. Despite these limitations, the machine could probably fabricate the secondary for a tesla coil right now, something that’s really annoying to make unless you have a lathe.
Video demo below.
“Despite these limitations, the machine could probably fabricate the secondary for a tesla coil right now, something that’s really annoying to make unless you have a lathe.”
Not likely. You need significantly larger wire gauge, otherwise the resistance of the overly thin wire limits the current and you get shorter arcs. (it’s a tradeoff). I seem to remember that something in the 20 – 22 GA range is best.
Also, you don’t need a lathe – just the right kind of jig and patience. ;)
I wonder if anyone has considered either modifying the extruder head to include the wire in the process, or using filament with wire already embedded (if that is possible)?
Not sure why this ended up as a reply to the previous comment…
No idea, but it also just occurred to me that most solid copper wire is usually enameled, so I’m still not entirely sure I see the benefit of this approach.
Enameled is not the same as insulated. The thin enamel doesn’t hold very much voltage.
Which is why you lacquer the hell out of the form after winding, preferably with transformer grade varnish.
Likewise, in any secondary coil, the voltage per turn isn’t terribly large, however if the dielectric strength at any one point is too low, flashover occurs and usually (not always) destroys all of your hard work. This is where the lacquer helps (and sometimes hurts, as in the case of included air bubbles).
fabricate the secondary for a tesla coil right now, something that’s really annoying to make unless you have a lathe.
No lathe needed, just some scrap and a hand drill with an autotransformer to adjust speed. Unless you have a variable speed drill.
20 to 22 gauge wire? I used 32 gauge for the secondary and had about 5 inch “brush” discharge, driven by a pair of 811A tubes in parallel.
Indeed. Don’t even need a drill…unless maybe you’re planning on making a lot of them? I wound a couple of secondaries, found the best way was to wind it while I was watching a movie with the family at home…I was done with each coil in less than an hour. Don’t shortchange the power of patience for one-offs or two-offs.
The secondary was about 19 inches long and 2 inches in diameter, and took a few hours with the drill. I wouldn’t try it without the mechanical assist. With 32 gauge wire you have to pay attention to avoid overwrap.
I don’t think you’re going to be running high voltage through that without some serious fine tuning. In the example video you can see parts where the wire isn’t in the plastic at all.
I have been trying to do this to lay an antenna for 433 MHz inside the box that holds the PCB. Early days yet, maybe it will lead to something viable.
Makerbot will be filing the patent for this next week.
nice
Firstly, you wouldn’t want to use anything but bare copper wire for this, with the plastic as the insulation. (did you guys miss the bit about the wire being insulated by plastic after being laid down?)
Second, this thing might not be particularly accurate at the moment, but that’s just a Saturday afternoon of tinkering away from the case.
(@[Bas]) :
Third, in order to cut the wire, have you considered mounting a razor blade next to the wire guide ring? it the wire’s thin enough, and been buried in the stream of plastic well enough, you could probably break it off predictably without pulling it all up. (failing that, have the printer run the wire to a loop outside the print that can be cut after the print is done -a little wasteful, but a simple solution)
Fourth, what gauges of wire have you tried so far? What do you think the maximum thickness might be?
Fifth, an idea for how to connect components: You might be able to get the printer to wire-wrap the legs of components that have been inserted into an already printed hole upside-down.
Simply print the “board” or whatever surface the components will be mounted in, then pause to insert the components into the holes (legs apart), before having the machine run over them with wire, wrapping tight circles (possibly even with the servo ring?) as it darts between connections.
Just have it either cut the wire after a connection is made using a mounted razor blade, or have it run to the outside and make a loop that can be cut later it isolate separate runs. (you would then have to re-anchor the wire with a bit of plastic, which brings me to an interesting idea, have you considered a knob, and tube kind of wire wrap, with wire run around plastic binding posts?)
Hope all this helps,
-Matt
P.S. Screw you makerbot and your patenting of ideas that have already been published as open-source!
Hi Matt,
You are right, This device needs some tweaking. But before spending a lot of time detailing the basics have to be right. I think they are, and everything that gets added will improve. The white tube is not rigid enough and needs to guide the wire better under the nozzle.
I’ve done 0.05 mm, 0.10mm and 0.15mm so far. Im no electrical guy so you’d have to translate that :) My feeling is that 0.3mm would be a max thickness in these dimensions. I did this with enamelled wire because that’s what I had, but in reality it needs bare copper wire.
The thing needs a scissor and a clamp, mounted on a second slew ring. The clamp picking the wire just outside the tube (or other), rotate this slew ring 180 degrees so the wire is then beneath the nozzle. The nozzle being in the middle of the length of wire. Then when starting the clamp unclamps and the wire-laying commences.
Maybe that second ring can do the wrapping around legs, but i think the other legs of the component would be in the way. I’d rather leave a series of holes where the legs of the component can be pushed into. But before pushing the legs into those holes, there could be an overhanging loose end of the wire (start/end point). When inserting the component, the pins bend the wire, effectively bending it around the edge and into the hole. Then the wire is pressed between the wall of the hole and the material of the leg.
Could you go a little bit into detail of your knob and tube wire wrap suggestion?
Regards,
Bas
Really cool! Can’t wait to see some ingenious uses for this technique. Seems to work well enough to lay down an antenna or small winding.
Very exciting! Mad props to this guy!
So, getting close to being able to 3d print a stepper motor…
If a wire cutter were added, and this trick works with fine mild steel wire, then it could be possible to embed the metal cores necessary in 3d printed parts, as a build up of parallel magnetic-circuit conformal but insulated wires.
This would seem to be especially well suited to smashing down the eddy losses, since the eddy loops would be even smaller than for eg a laminated core, and the resulting motor should work quite effectively even at very high speeds… but maybe such a technique would be even better employed just to print a toroidal form magnetic core, which should then work very well as a very high switching speed transformer.
Although it would need to be manually wound, or wound with a toroid winding machine.
(Protip: wind “back and forth” around the toroid ideally starting and ending each winding at the same place around the circumference. Don’t lap it around the circumference or you will accidentally form a couple of “slow windings” around the big radius of the toroid, which won’t help the transformer, and will just lead to increased leakage flux – which is especially undesirable at high frequencies.)
Since transformer power handling is related to operating frequency, and reducing the losses – both eddy and leakage inductance – helps increase usable operating frequency… Well, it’d be interesting to see if a toroid printed this way in fine soft Si-steel wire has a high enough fill factor to be useful…
well, look into planar transformers. the windings are laid out in PCB traces, it’s pretty cool. I don’t expect a 3d printer to be making pcb boards, different tool set. but fun to keep the ideas flowing, who knows what you might hit on.
Could this method be used to lay glass thread inside the plastic for reinforcement?
Never mind the Tesla coils – I want a quadrifilar helix antenna. http://lea.hamradio.si/~s53mv/navsats/qhelix.jpg
No comments about homemade PCBs? I’m surprised. This kind of improvement could allow a move away from the layered approach to pcb design entirely.
I’m assuming this would be one of two (or more) filaments, so not every part of your print is wires. It would also need some kind of cutting attachment, so it could add breaks in the wires.
The hard part is getting the wires to attach together. The plastic, while conveniently acting as insulation, would also prevent interconnects. Hand soldering all the joints would be a little painful. Maybe a solder paste extruder could provide pockets to be melted once a print is finished.
@Willrandship I’m a little surprised too about nobody thinking about their own PCB, imo It’s not that difficult to do from a physical point of view. I agree that people should think further than 2D layers, think in 3D multi-multi-multilayer and components at different planes or positions in the product. Need another crossover, add another layer.
Because it’s not yet possible now, doesn’t mean it isn’t possible.
The good thing about not having the wire inside the nozzle is that the ends are clean.
When you’ve cut a wire you can print “only plastic” again because the wire is not in front of the nozzle anymore. If the cutting and picking of the wire for the next trace proves too difficult you could use 2 extruders, one for the wires, one for the rest of the plastic.
again, interconnecting, don’t think in 2D, when you have a pin, pressed in a hole, end of the wires gathering (in different height) inside that hole, you can connect different levels together with that pin.
Wouldn’t it be easier to extrude the wire through the nozzle along with the filament? It could affect the heat transfer a bit but you wouldn’t have to worry about laying it in front of the plastic all the time
Why not “printing” the copper as well and not using a solid wire? I’ve seen that somewhere in the internet maybe sombody can help me finding it? Copper powder was also melted, however the plastic as well. Afterwards a chip was put into the part manually. Hey you, if you are out there and know what I’m talking about, please send the link of the start-up/ video doing this; can’t find it.