Magnet Binary Adding Machine
Our efforts to design a magnetically based binary adding machine have resulted in this system. Compromised of several modular parts, this whole system is actually a number of individual pieces that when used in conjunction, create the workings of a complete half adder. When run through the system multiple times, one can continuously add binary digits, therefore working as a full adder.
Its important to note how we coded these pods. A vertical pod can be read as a “1″ and a horizontal can be read as a “0″. In prior experiments, we had coded magnetic polarity as either a one or a zero (North side being “1″ and South side being “0″), but labeling them as such gave us limited options. Namely, this only resulted in a natural XOR gate. This made it extremely difficult to develop any other system than this.
In order to accomplish a new method of inputs and outputs, we used two forms of magnet pods. The first one (on the left side) functions as the initial input. These are fixed values, incapable of rotating around within their housing. The second (on the right) can spin, thus rotating itself into the proper orientation.
When the rotating output pod leaves the gate, it is lock in place using piece like the one pictured. This allows the output pod to work in the same fashion as the initial fixed inputs.
The AND gate is our simplest mechanism. Two inputs placed on the outside slides control the orientation of a single magnet. Every placement of the input except a “1″ and “1″ result in a “0″. The is due to the proximity of the horizontal “0″ input. Because it is closer to the rotating output pod, the output pod rotates closer to the horizontal.
Our XOR and OR gates function as identical mechanisms with very small differences in their locking mechanisms.
Our gates would sometimes result in undesirable outputs, namely 45 degree angles. This was an issue, since we can only work in horizontal and vertical variables. In order to solve this, each gate was given a different locking mechanism for its output pod.
These output mechanisms had extending arms that, when entered into the second stage of the gate, would push out either a fixed “1″ or “0″ pod.
In essence, if the output pod is rotated at a 45 degree angle, the arm is as well, therefore bypassing the adjacent output arm and directly coming into contact with the opposing output.
Our system did not come with out issues. Although we were able to work out some of the larger problems of polarity (and magnetism in general). Our system lacks an infrastructure to clearly denote and mark the path of the outputs. Perhaps a railing system could be of some use in the next draft.
Its hard to avoid the fact that our locking arm solution was more of a fix for an existing problem rather than a cohesive part of the system. It was, in fact, a method proposed to resolve an undesirable outcome that happened to be in inherent issue of our machine.
E+R