One of the first components we made was a resistor. Our first couple were VERY simple: 14-gauge solid copper wire soldered to the legs of a standard 1/4W resistor with a glob of polymer clay molded around it.
It worked, sort of. The solder joints were fragile, and there was nothing in the design to keep rotational force on the legs from being transferred directly to the solder joints, which caused them to fail pretty much as soon as they started being used.
We also experimented with a slightly reinforced model, with the resistor being inside a piece of plastic tubing (I'm sorry, I don't know what kind of tube it was, other than to say NOT PVC). We tried to build up a structure around the outside with polymer clay to provide a more convincing physical shape but the plastic softened too much during the bake cycle (polymer clay must be baked at about 275F/150C), deforming and warping and looking in the end not at all like a resistor (or even a cylinder, for that matter).
Team members Karin and Jon finally hit upon what became the best solution: an armature of wood with the heavy gauge wires wrapped around and threaded through holes in the wood. An extruded sheet of polymer clay was wrapped around this and molded into the right shape; smaller strips of polymer clay of the appropriate color were made into the bands to be wrapped around the outside.
Obviously, that's not a resistor- it's an electrolytic capacitor. The principle is the same, though- the main difference is that, with the capacitor, we bring both leads out one end and the resistor they depart in opposite directions.
This technology can be used to form the inside of any small, leaded component: diodes, capacitors, resistors, transistors, etc. Below are some images of various completed and uncompleted components:
 |  |
A mostly completed electrolytic capacitor. The armature was inserted into a piece of black plastic tubing, wrapped in black polymer clay, and baked. By bending the leads to support the component off of the surface it sits on in the oven, deformation of the tubing was avoided. | First-generation ceramic capacitors. Formed by slicing apart the flat edge and inserting the capacitor with its pre-attached larger leads. |
 |  |
Clear-cast diode. The diode was attached to the wooden lead frame in the manner previously described; that was then inserted into a piece of plastic tube which was filled with "mass-cast" brand clear resin. The result was quite spectacular; this may be a good method for providing students with more visibility to what is inside the large-scale components. The resin is fairly expensive, however, and noxious to work with. | Unfinished third-generation ceramic capacitors. The body is a piece of insulating foam purchased from a home improvement center and cut to form on one of our home-brew CNC machines. These were subsequently covered in papier-mache and spraypainted. |
 |  |
Finished resistors. The substrate polymer clay is the lighter weight, less expensive sort used for internal volume filling applications. | Pre-cut "bands" of polymer clay for marking up the resistors according to value. Wrapped in plastic food wrap to protect them as the small oven made this a multi-day project. Note the can-type transistor off to the side. |
So, here we have demonstrated a number of methods for making these components. Naturally, other methods exist: we discussed (and discarded) silicone mold resin casting as too expensive and complex, but that would be a good solution for making a lot of components quickly (perhaps as a maker-based business). Wrapped layers of taped might be a cheap and easy solution, or paper with a single layer of tape on the outside. Pieces of pipe with the component inserted, and subsequently backfilled with plaster of paris, would be a less portable but still very durable and inexpensive option.
In short, anywhere one can find (or bring, or scavenger, or order) the innards of the components, the larger-scale parts can be constructed cheaply and easily.