What you are looking for is a Class E amplifier. The essence of the design is to use a switching transistor to generate a square wave output and then shape the output using an RLC circuit. The math is fun... honest !!! Look at the following web sites for more information: http://argesim.tuwien.ac.at/comparisons/c3/c3.html http://www.its.caltech.edu/~mmic/reshpubindex/davisAmp/class-e.htm http://www.its.caltech.edu/~mmic/reshpubindex/papers/ClassE.pdf There was an article on this with in the past year in one of the engineering publications (IEEE Spectrum, I think) which highlighted the benefits of Class E amplifiers in laptop computers where energy management (battery life) was a critical issue. Tom -----Original Message----- From: pic microcontroller discussion list [mailto:PICLIST@MITVMA.MIT.EDU]On Behalf Of roger Sent: Sunday, 02 January 2000 11:09 To: PICLIST@MITVMA.MIT.EDU Subject: digital sound: part deux excuse me for failing to mention i understand a transistor as a switch. any further discussion of analog electronics will be over my head. now, if you are still with me, what i want is a direct digital sound. chirp the processor out at the desired frequency(ies) direct to the sound as perceived. a square wave is produced by the processor. i can count until i'm blue but the output of any logic chip is a square wave. (no conversion please.) a square wave makes a buzz. a tone or note, (and forgive my misuse of terminology), is a complex form made of several different frequencies, and presumably different wave shapes. an analog speaker is used to create an audible note, pleasing to the ear. it is controlled through a magnet which switches on/off. (perhaps the secret is in its switching partially on, partially off, but i think that is volume). by switching a magnet attached to a speaker on/off at a particular frequency i will have a tone, but a buzzing tone. to make it more pleasing i have to ramp the magnet up and ramp it down at some sub frequency, is that correct? using digital it seems that i am stuck here. if i use multiple loops to control the various sub frequencies then, it seems to me, they will add and create some other, lower frequency. what i had begun to think of was some sort of direct ear canal implant using MEMS technology that might use a hammer striking directly onto a membrane next to the ear drum. or a hammer striking tuned rods next to the eardrum. digital to digital until inside the earcanal where the analog conversion takes place without any outside ambient noise to interfere. any thoughts on this?