8.7 Mhz Oscillator

Project Briefing
This is a crude and simple RF oscillator which generates a semi-sinusoidal signal at 8.7 Mhz. It is merely a prototype implementation of an existing circiut entitled "Simplest RF Transmitter" located at http://www.uoguelph.ca/~antoon/circ/rft1.htm. The whole project was merely a proof of concept for filtered feedback oscillation and was certainly not intended to be a clean or stable source of carrier RF.

Technical Overview
The general idea of this device is filtered feedback. As you may know, when you hold a microphone up to a P.A. speaker, there is an obnoxious screeching noise which is reffered to as feed back. All electronic systems have noise. Noise is transient oscillations induced or put through the line by background radiation and ambient radio waves. This noise is so subtle you would not normally recognize it (if the device is well grounded and filtered) but the P.A. speaker is still emitting this constant, inaudible noise. When you put the microphone up to the speaker, that noise is picked up by the microphone, amplified, and then ouput again through the speaker louder. This noise would normally be an infinite number of frequencies causing the sound of static. However, due to minor delays induced by the speed of sound and the reaction time of the circuitry, some frequencies will regenerate out of phase and interfere with themselves thus destroying those noise signals. But, there will be one frequecy (and harmonics thereof) with a phase that will line up and thus amplify itself. This is the resonant frequency. When a noise signal of the resonant frequency is fed back through a P.A. system, it continues to be amplified thousands of times per second thus resulting in the deafening screech of feedback.

The principles in this simple RF oscillator are the same. Transient noise in the lines are fed back through the amplifier to find resonance. The only difference is the feed back of the oscillator is fed through a filter which allows (in perfect theory of operation) only one desired noise frequency to be reamplified. The cunning is in the feedback filter. When AC passes through a capacitor, the phase of the current actually lags behind the that of the voltage by 90 degrees. When AC passes through an inductor, voltage lags behind 90 degrees in phase. When these two components, a capacitor and an inductor, are placed in series, it creates a filter. Only when a signal has a certain frequency will its phase lag line up in such a way as to not be destroyed by the filter.

Construction
Breadboarded from scraps in about ten minutes, it was certainly quick gratification as it worked on nearly the first revision.

Component List
 * R1 47K
 * R2 Photocell (unknown value)
 * C1a .1uf, C1b .02uf
 * Q1 2N2222
 * L1 (30 turns around 1/4" plastic form, center tapped 1/3 from hot end)

I made only minor revisions to the original schematic. I used C1a and C1b in parallel for .12uf as a value for C1 on the original. This pulled the frequency just into the easy viewing range of my oscilloscope with the coil I was using. The coil is obviously homemade consisting of about 30 turns about a piece of 1/4" (approximately) diameter plastic and centertapped 1/3 of the way from the hot end. In fact, I came into some problems when trying to use a twist tie to lock one side of the coil.  The tiny wire in the tie was inducing the magnetic field at the end of the coil and was bent in such a way that it was feeding the signal back over the coil to cause some bizzare multiphase interference.  This was remedied, however.  Finally, I added a photocell in series with R1 to observe how the varying resistance affects the output frequency.