"Vertical" circuits

The problem

When designing a circuit, one of the coordinate planes must be “constrained”: the X/Y plane, or the Z one. Almost every single circuit chooses to constrain the X/Y while leaving the Z free to move. For example, in a PCB, when soldering a resistor in, there are specific holes that the resistor leads need to enter which exist at specific X/Y coordinates that connect it to the traces. The resistor can be placed as “deep” or as “high” as desired as long as the leads enter the correct holes (again, almost every circuit places the item as low to the board as possible but the choice is there). Even on breadboards, while the multiple holes provide a flexible arrangement, the conducting item must still be placed within specific (small) areas across the surface. This creates the constraint that every item must be placed in parallel, but that was fine for my use case.

The problem I was trying to solve is how to create a working circuit where an item would receive it’s necessary power and ground connections no matter the orientation it was placed in. This means any element, even single-direction ones like LEDs, could be placed with diodes forward or backward, with the leads directed north/south, east/west, or any weird combination of directions. The only way I could come up with to achieve this “freeing” of the X/Y dimension was to do the opposite of conventional boards and restrict the Z. This means that instead of a circuit element receiving it’s power from one bus and ground from another ~5mm to the left, it would receive that same ground from ~5mm down. Instead of each connection having a specific coordinate, it now would have a layer on the Z axis.

Once again though, to be able to achieve contact with different surfaces at differing heights, how do you enable the element to reach the lower layers? If you use some sort of breadboard-like grid that has a single connection for all insertion points that allow connections to pass through its bottom, it defeats the purpose. You’re still restricting the X/Y. Even if some sort of fine mesh is used, that still creates a finite amount of points to pass through. Therefore, I determined the most flexible strategy would be an unbroken sheet of some material that allowed each element to penetrate it.

So at this point, the idea is that there will be two sheets of something separated by a small distance that the element will penetrate, and in doing so, make a connection with each. The first obvious problem here is that you can’t have both leads of an element going through both layers - it will short. An easy fix is to clip whatever lead needs to connect to the top layer much shorter than the longer one, so even when the element is fully inserted, it doesn’t reach the bottom conductive layer. The second issue is that if you have one lead of an element connected to the top layer, and the second lead going through the top layer to connect to the bottom, you’re creating another short. Why would any voltage flow through your element when it can just go straight from the top layer to the bottom one?

These two problems, the conducting surface to use and how to insulate one element lead, are the problems I’ve tried to solve a variety of ways, and can still be done better than what I’ve arrived at.

Solutions for insulating one lead

1. Tape

I was doing my first prototyping with a breadboard sitting in my living room, and the quickest and easiest attempt I could think of was to use tape that was on hand. I figured it’d be relatively slim, cheap, and easy to do in bulk.

This was not the case unfortunately. The leads of an LED are too small to get the tape wrapped around one - it just wants to encircle it and then tape back to itself, sticking out like a little flag. I tried painters and masking tape and I couldn’t get either to reliably stay on without having a tag sticking horizontally out from the side.

2. Heat shrink

I moved down to the Pier 9 lab after trying the tape as I knew any other materials I needed wouldn’t be in my apartment. I figured heatshrink was a good choice since insulating wires is exactly what its job is! I got the slimmest diameter heatshrink tube I could from the lab and wrapped it onto a single lead of the LED. The result was…okay. Even the smallest diameter was still a little big, and could slip up and down the if I forced it. It was also a little thicker than I’d hoped, and made poking the leads through the insulating surfaces a bit difficult. Nevertheless, it was definitely both the quickest to implement (which is important for making a lot of these) and also the most effective out of all the options I tried, so after trying the other options I decided it’s what I’m sticking with for now.

3. Hot glue

I figured that maybe I could make hot glue act as kind of a malleable heatshrink by daubing it on and smearing it down the length of the LED end. However, I didn’t feel like getting out the hot glue gun and waiting for it to heat up. This option also felt like a little involved of a process and not one I could churn out easily. I relegated it to the back burner and will return if I can’t get anything else to work.

4. Super glue

This was what I landed on right after coming up with hot glue. I thought that the lower viscosity would help keep a thinner layer on the metal while also giving me a harder dried product that would be less prone to pulling off when pushed through the conductors. This didn’t work out for multiple reasons. First was that by the time I had applied a layer to one LED and waited for it to fully dry, I had run through all the other prototypes I wanted to work on. This made it unappealing to produce many in a reasonable time. Also, this just seemed like a short waiting to happen, as if the lead bent too much I felt like it was liable to crack the glue and expose the metal underneath. I never even ended up testing this one.

5. Sharpie / paint

So I’m a little embarrassed about this one. I thought that maybe I could get a thick enough layer of sharpie to form a thin layer between the metal and the conductor. Yeah absolutely not, it’s just ink (idiot). This failed immediately. It did lead me to think more about different surfacing liquids though. I didn’t try it, but I think paint might have a better chance. It would have to be globbed on well enough to form a non-zero thick boundary, which might take multiple layers. It also brings the bending/cracking problem back into the equation that I considered for the glues, and more so this time since paint would be less malleable. I left this one untested for now.

6. Liquid electrical tape

This idea came about from some googling around insulation and liquids; I had never heard of it before. I think this might be a decent option depending on the dry time. Unlike the other liquid solutions, this one’s job is to insulate, so I wouldn’t have to worry about multiple coats or material thickness. I think it would also be an improvement over the heatshrink since it (I’m assuming this, I don’t know) wouldn’t have the hard corner edges of a piece of heatshrink and could smoothly slope up and down on its edges. Definitely keeping this idea in my back pocket in case I can’t get the heat shrink to work, I just didn’t feel like going out and buying any at the time.

Solutions for the conductive surface

1. Copper foil

What does every circuit board use to connect different elements with the lowest internal resistance? Copper!

I figured something like foil would make the most sense; it’s basically big copper tape without the stickiness, and (going off my knowledge of common aluminum foil) very thin so it’d be easy to puncture. I googled around and there are several options on Amazon, but I wanted to try it ASAP so I stopped by my local craft store and picked up a roll. To prototype its viability, I cut out two ~2 two square inch pieces of the foil and two ~1cm thick layers of foam in the same dimensions. I soldered a single wire to each square of foil and taped everything up in alternating layers. From the bottom up it went foam -> copper -> foam -> copper.

The good news - it worked. The bad news - it was difficult to make work.

The actual connection of the element to the foil was great. There were some flickers if I was moving or jiggling it around, but for the most part, the foil held it’s shape well enough that a constant connection held between the copper and the leads. The issue was how hard it was to puncture the foil. I noticed this when unrolling it a bit after purchasing it, but it is not as thin as regular foil. With the thin, flexible leads of the LED, it was really hard to pierce the top layer without bending the leads at the base and splaying them everywhere.

I made it work better by first puncturing my desired spot with a pushpin and then feeding the leads through that hole. This removed the difficult initial hole, and the hole in the lower layer was easier because it was guided vertically by the top layer + foam and was just one single, skinny lead versus the two grouped together that went through the top.

I didn’t really consider this a solution though, since this was a pretty arduous process just to insert an LED.

2. Copper tape

This was the first thing I tried and what I used to prototype the different LED lead insulating options. There were some rolls of tape available at Pier 9 and $free is better than any other $ so I gave them a shot. The only downside is that they were about 5mm wide, so I didn’t consider them a viable option for the final product. I’m not trying to lay down 300 perfectly-overlapping layers of tape per conductive-layer, thanks.

The positive for tape was that it was thinner than the foil I had tried (!). This meant it was relatively easy to puncture with the LED leads, and removed the issue I encountered with the foil. For this reason I’m tentatively calling tape my winner, but I’m not sure if the wider rolls of tape will come with the same gauge as the thin ones, and if they don’t then I’m back to the drawing board.

3. Conductive paint

This is one that I considered for a bit but eventually decided not to try.

At first I figured that paint would be a good option: easy to cheaply scale to larger surfaces, and easily puncturable since it’s just thing layer on the lower surface it’s applied to.

However after thinking about the implementation some more, I don’t think it would work as well as I first thought. For one, dry paint is decidedly not a thin layer that’s easily puncturable. Have you ever seen painted canvas/paper/cardboard? The paint is a solid layer that cracks when bent. On top of that, even if I managed to make it so LEDs could penetrate it easily, I’m not so sure it’d maintain a consistent connection. With the copper foil and tape, the material itself was conducting throughout it. With paint, once the lead was past the conductive surface, it’d just be in contact with whatever non-conductive material that was painted on. This means the only connection between the paint and the element would be the super-thin point between the LED lead and the layer of paint.

4. Conductive marker

This option was borne from the paint idea. Since it was just a marker, this would solve the issue of the paint being hard to push thin leads through. I could even double up on the connection area since I could color the top and bottom of the material.

This might work, but I still think the issue of an extremely small area of conductive connection like the paint would bite me, and it would be hard to maintain a sturdy connection between the leads and the marker when LEDs continued to be inserted.

5. Wire “mesh” just from hundreds of thin wires strung together that would be flexible for XY

Holding the title for my most “out-there” idea is a spaghetti mesh of wires bundled in a thin flat layer. This would need hundreds of super fine gauge wires all connected to a single power source. It would probably be a few mm thick with tons and tons of wires arranged so that there were no ways to get through without touching one. This would be just like using a fine mesh, but one created with multiple vertical layers. It’d be horizontally flexible enough for the wires to move sideways, so any position of the LED would go through it with no resistance, but stiff enough that it’d hold the LED in place and maintain a connection.

I think this would work the best out of all of the options, but it seems like quite of bit of effort that I’m not sure I want to go into when I can get tape or foil to work well enough.

Where I’m at and next steps

Currently, my best prototype with the materials I have two sheets of (thick!!) copper foil separated by light craft foam. The LEDs would have one lead clipped and the other with a premeasured bit of heatshrink attached.

In terms of continuous connection, this works pretty well. Some of the LEDs will blink out when I’m jostling in another pin, but they’ll they pop right back on if I give them a little flick. I think the fact that I’m having to push these in on a little foam block that’s moving around the table as I hold it doesn’t help either - being mounted on a wall would improve that aspect.

I think the best thing for me to do next is to suck it up and buy a roll of copper tape. If it’s the same thickness as the foil then I’ll have to deal with what I’ve got. Maybe I can reinforce the LED leads from the base somehow so they have less tendency to splay from their connection point to the plastic? We’ll see.