Toward 3D printed Circuit Boards

[LesterCHall]

I have thought up three or four methods for making circuit boards with 3D printers, each with it's particular advantages and disadvantages. The one I will describe here is my chosen method, one that might be called the Guided Wire method. See the images below for reference.

WireEmbedding2.jpgwires3tries.jpgCubism.jpg

in the first image we see the method of guiding the wires, one of creating a round trough with a smaller channel above it. The channel is just wide enough to accept the wire, and the wire once forced into the trough with your fingernail or other appropriate tool resides in the cylindrical region below the surface. The second image shows the attempts that I made in doing this. I found that a multiplier of 2.75 times the radius of the wire works best for ABS on my ROBO3D printer's glass bed. Note that the diameter of 22 gauge wire is 0.644mm. In the third photo we see the "breadcrumbs" that I posted to Thingiverse a while back, basically 3D printed perfboard.

Now that I have wires (with some labor involved) and parts holders (perf board or part footprint), I'll need to make an example for soldering. Then after that some trace crossings perhaps and that will be all we need for a 3D printed circuit board!

Les

[LesterCHall]

Here are the results in attempting to add a chip:

chip8wiresPCBtop.jpgchip8wiresPCBbottom.jpg
Note that the photo on the right shows the pins soldered to the wires. It's a bit tight but not too tough to solder without touching the plastic substrate, even with my shaky 47 year old ex-alcoholic hands I can do it so you can too (most can). I found the scale zoom of 2.75 to be too large, i will try 2.25 next.

Les

[LesterCHall]

LunettaRingModulator.jpgproto1.jpg
OK so I'm happy with the notion that I can now make circuit boards, having tested the most basic features and imagined the practicality of others (like crossover wires and multiple layers, 3dism, etc). Next Up I'd like to make a demonstration project that illustrates 3d circuitry in action, preferably with some actual 3D involved rather than just flat circuit boards. To that end I have decided to reach into my bag of tricks and pull out: "Electronic Music Circuits!".

As it turns out I have somewhat of a hobbyist background in electronic music circuitry, both in software and in hardware development. One circuit I've been really wanting to build someday is a one-chip-wonder, where you get whatever krazy sounds (and they may be harsh) from a single chip. That chip, of my choice is the ATtiny85, a miniature Arduino-compatible beauty that sells in quanity one for only $2.00 at Jameco Electronics (www.jameco.com). Jameco also sells everything else I need for this little gem so it's a one stop shopping experience which saves on shipping.

Anyway back to 3D-ism and wouldn't it be nice if i could make a hand-mounted sort of a curved band that fits in the palm snugly and holds all the circuitry on the periphery? That way the instrument could be operated with one hand and i could use the other hand for another of it's kind and do performance art at amateur music festivals such as the electro-music festival which I attended in 2009. OK, TMI but you get the idea, the author is going to make a music box thingie all three-dee-ified on a form that fits his hand.

Now to the photos. Form follows function and the main function here is to fit the hand so I thought I'd take advantage of the rapid prototyping nature of 3D printing by printing up a hand band of sorts. Well as it turns out I got it sized wrong at about 70% of the full size needed for my mighty meat hooks of a pair of hands! Rather than resize and print again, I took this proto and marked it up to approximately designate where the circuitry is to go. Next I'll go back to CAD (in OpenSCAD) and put in approximations of the circuitry but without the wire containment channels (that's a lot of work that is to be done later). As a matter of fact, I'm off to go do that right now so pardon me for making my escape right now!

Les

[LesterCHall]

proto2.jpg
It was a two and a half hour print, but worth the time as the photo shows, a perfect fit to my hands. I can mount potentiometers and switches and maybe CAD up a battery holder and some circuit sections all built in at various locations that make sense on the structure. The freedom to morph circuitry with shape and form of the device, to use thru-hole parts mixed with solder tab parts, and to morph or grow some parts into the structure itself is what 3D printing is all about! Now to think of what to buy and where to mount it and how to get good sound from it. I think delay lines are in order and perhaps a few ATtiny85's not just one, well we shall see as we say right?

Les

[old man emu]

I think I understand what you are doing, but not having seen the layout of the circuit you are making, it is a bit hard to get up to speed.

Would you have a look at this simple circuit, http://homebrewedrobotics.blogspot.c...e-circuit.html , and see if you can print the circuit board? (The circuit is just one I grabbed at random following a Google search for PCBs.)

Old Man Emu

[LesterCHall]

Yes, Old Man Emu, I understand after seeing your circuit that it will be much clearer to the readers what this technique is and is not capable of doing if I first create your proposed circuit, then make the kooky sproingy artistically inspired thing I am planning to make. I will begin work on your proposed circuit immediately.

Les

[LesterCHall]

RobotPCB.jpg
Here is the layout after a fast effort of 3 hours of CAD in OpenSCAD. I'm printing it now...

[Mjolinor]

It should be possible to use the gerber to gcode utility for PCBs. That's what I use for my PCB mill, not sure how you would change it to an extraction gcode rather than a cut gcode though. Maybe Skienforge can do it.

[LesterCHall]

That's an astute observation, Mjolinor, I wasn't even aware that a gerber to gcode utility existed. That would be just the thing for this. As I have it now, I used basic trigonometry and pythagoras' theorem to make it so wires can be specified as grid coordinates [x, y] at begin and end. Same for placement of electronic parts, which also take a rotation, pin count, and spacing parameter. Here is the result:

robotProto1.jpg

[Mjolinor]

I was also thinking that for the most part, assuming modern high impedance electronics it should be possible to use graphite powder and PVA to paint it then sand off back to the plastic. I have no idea how high the resistance of the tracks would be but I reckon for the most part with CMOS stuff it should be fine with the exception of power rails which you would perhaps have to lay with copper.

I have a strange mind.