Dear Dennis... analogZONE's Dennis Feucht answers your design queries in his new Circuit Design Clinic! December 2006

analogZONE presents a new, interactive analog design service to readers! Send us your design questions (with relevant data; schematics in JPEG or GIF, please) for some free engineering advice from analogZONE's circuit consultant, Dennis Feucht, on how you might solve a design problem or improve circuit performance. Submissions may be edited for clarity or brevity, and submitters and their email addresses will remain anonymous (unless otherwise indicated). Please send your questions to Dennis here.

Brainstorm Session: High-Efficiency Linear Amplifier

The way the Circuit Design Clinic usually works is that you send me an electronics circuit (or circuit-related) problem and I try to provide you a somewhat decent solution to it. The alternative format is that problems encountered on a particular project are surveyed and solutions discussed. This scheme assumes a category of problems that I call standard engineering, problems of a kind that have known solutions. The parameters and other features of your particular problem might be unique, but the kind of problem posed is not novel. Solving it will not break into new, unexplored territory on the electronics knowledge landscape.

The other category of problems is those that have no known solutions, or for which alternative, possibly better, solutions are needed. If a reader were to ask: "How do I solve the problem of communicating through the earth to the other side of the planet?" I could not provide any known, elegant solution other than the existing long-way-around methods. (And, not being well-oriented to the frequency domain, I wouldn't even be good at that.) However, I could instead go into brainstorming mode and propose some ideas that might lead to something novel -- maybe neutrino-based communications. (But how, you ask, without using too much carbon tetrachloride in the detector?!) This category of problem is, in contrast to standard engineering, a state-of-the-art, or leading edge - or simply an engineering research problem - a problem not ever really solved (or solved well) before.

Read the entire response here (52 kb Adobe Acrobat PDF file)