I have just had this idea rolling around for some time now. I understand that microcircuits operate on DC voltages, and not AC. Direct Current can only be expressed in terms of "On" and "Off" and therein lies the inherent problem, as far as I can see. I also know that logic circuits are basically arrays of Transistors, which are the offspring of good old Vacuum Tubes. I know the way machines can process data is through crunching strings of binary voltages together along programmed logical lines. This is about ALL I know about the processes involved. Make no mistakes about it!
I am trying to determine if there is a way to use AC voltages and trinary numbers in any type of logic circuit. Is it possible? How might it be done, if at all? I am really not concerned with the size or speed of such a new processor, and neither should anyone else considering the possibility. Just look at the size and speed differences between the ones we are working with now, and ones just 10 years old. Size and Speed are things that can be resolved later. Not one of those individuals who conceived of and built ENIAC, or its kin, was the least bit concerned with how big and slow it might be.
I know as well, that AC current is not as easy to keep track of as DC, as these are frequencies that are actually waves and part of the Electromagnetic Spectrum. But that opens up possibilities in itself. Consider logic circuits that might not have any physical conductors of current through their communication ports at all. Advanced radio technology searches for, responds to, and processes only certain frequencies if so programmed. Wireless and Conductorless microprocessors? There is also the possibilty that input devices might even function BETTER with speech, than a keyboard or mouse. And what of the potential for Modulated Waves in conveying even more information than a simple Alternating Current? AC may well be far worse at conveying clean and accurate information without much work and thought. Radio waves are usually subject to more interference than any line signal. Perhaps this may well rule out standard sine-wave AC, but there may be some other way to use an alternating voltage to convey data of three types rather than only two. This is where it must begin.
I am doing my own studies, and will continue to update this site as more ideas are generated and more information found. I am not writing this so that some expert in the field can say "what a dumbass that guy is!" What I want is for someone to apply what they know, and ideas and knowledge from within the computer science community and start working on it. It WILL be done, if it is not already so. I am quite certain of this. In all probability it will not be me. If I can get what I think out to others who may be searching for the answers along different lines than I have been, I may just live to see it happen in my lifetime.