For the objectives of this project please refer to our Project Proposal . What we present in this document is the basic theory of FM transmissions and the results of our examination of the FM transmitter used in this experiment.
Frequency Modulation (FM) is the method of varying a carrier wave's frequency proportionally to the frequency of another signal, in our case the human voice. This compares to the other most common transmission method, Amplitude Modulation (AM). AM broadcasts vary the amplitude of the carrier wave according to an input signal. Standard FM broadcasts are based in the 88 - 108 MHz range; otherwise known as the RF or Radio Frequency range. However, they can be in any range, as long as a receiver has been tuned to demodulate them.
Thus the RF carrier wave and the input signal can't do much by themselves, they must be modulated. That is the basis of our transmitter. An example is useful to illustrate what is actually going on. If we were to broadcast a 100MHz signal and tune a radio into that frequency, we would hear nothing. That 100MHz signal has locked or captured that spot and simply produces a DC value. Now if we were to move the incoming signal +/-100KHz in either direction at a frequency of 1000Hz, then we would hear a 1000Hz signal on the radio. If we only moved +/-10Khz then the sound from the radio would be 1/10th the original in loudness. Thus the rate or frequency at which we change the RF carrier produces the audible frequency that we hear, and the further from the main RF carrier we move, the louder the output will be. This is the basis of all FM transmitters. We will now look at how this is achieved by examining the basic circuit.
It is also important to note that the Federal Communications Commission (FCC) have very strict rules regarding broadcasting in these ranges. It is important that you check FCC regulations before attempting to build this circuit or any similar circuits.
To find out how our basic circuit functions, follow this link. However we have made some modifications to our circuit so that it is easier to use. We discuss these in the following section.
Our goal was to produce a working bug that was easy to install in a telephone line. We found that the initial setup of just using wires to connect it to the telephone lines was inefficient. Once we received the initial kit, it was assembled and tested. We however, wanted to see if it was possible to reproduce these result ourselves. We copied and modified the original artwork of the circuit and produced new PC boards. With common parts from Radio Shack, we were able to make another functioning device. We decided to use one percent precision components to prevent the effects of thermal drift. Some hole positions were moved since we used slightly different sized components. The most severe modification was the removal of one hole used for the variable capacitor. We used a nice shielded variable capacitor (5-90pF) with only two legs instead of three (thus the need for only two holes). This slightly larger value capacitor allows us to compensate for some modifications to be discussed later.
The component we worried about the most was the inductor. Since this was made by winding a piece of wire around a pencil (6 turns), it was not possible for us to measure the inductance. Also inductance varies with frequency, and most inductance meters only test in the kilohertz range. What we did instead was to use a straw similar in diameter to a pencil. Once the inductor was soldered into the circuit, the straw was left in place. The circuit was then tested, and once the desired carrier frequency was found, the inductor was glued to the straw using rubber cement to prevent the separation of coil to change. This fixed the inductance and thus made the operation of the oscillator more stable.
The next thing we did was to mount the circuit in an RF shielding box. This metal box prevents radio frequencies from interfering with the circuit operation. This interference can be caused by emission from other appliances, or from the closeness of the person to the circuit. One effect is to make it extremely difficult to tune the variable capacitor since we must hold the circuit board in order to do so. However, placing it in the box changes the capacitance of the circuit, this results in a shift of carrier frequency. This shift can be compensated for by the variable capacitor, and the advantage is that holding the box will no longer adversely affect the circuit inside. A small hole in the box lid allows the users to insert a small tuning screwdriver, yet it will not let RF signals pass.
The final modification was to wire in a telephone socket and plug so that no permanent modification needed to be made to the telephone or the house wiring in order to use the device. This was done to facilitate testing.
Our results and analysis Of course with all FM broadcasts, we are interested in how far our transmission can travel. In our case, the FM transmitter could broadcast up to 86 ScottFt[1] = 77 Ft.
This section contains actual, empirical data that we gathered from a 500MHz oscilloscope. We will explain our results and provide a link to the actual scope plot(s).
The FM transmitter is powered off of the telephone line. It only activates when the receiver is off hook. When the telephone receiver is on hook, we measured the DC voltage to be 58Volts. This is high enough to keep the PNP transistor, Q2, from turning on. When the telephone receiver is off hook, we measured the DC voltage of 6.8 Volts which will turn on the PNP transistor as explained in Basic Circuit Theory. These values are what we expected.
On hook DC voltage
Off hook DC Voltage
Here we examined the audio signal of a standard dial tone by probing the green wire going into the telephone. It shows frequency modulation of the original dial tone signal.
Dial Tone frequency modulation
The amplifier takes the audio signal and boosts it's amplitude by a factor of 1.21. We measured the original signal to be approximately 660mV and the amplified signal to be 800mV. The following is an oscilloscope plot of the two signals.
Audio Signals(CH1: original signal; CH2: amplified signal)
We tuned our oscillator to resonate at about 94.6 Mhz checked it with a Philips universal frequency counter. This is our carrier frequency. On our oscilloscope we examined the carrier wave while the dial tone was present. This resulted in our seeing a small shift in the frequency to 93.992 Mhz.
Frequency shift due to dial tone
We then pressed a touch tone button and the dial tone was removed. This resulted in a second shift to 94.452 Mhz. This is closer to our carrier frequency but not exact due to ambient room noise being picked up by the telephone receiver. The fact that the amplitude of the noise is extremely small, means that we will have an extremely small shift away from the carrier frequency.
Frequency shift without dial tone.
We wanted to examine the oscillator circuit in more detail, therefore we decided to run SPICE simulations to examine various aspects of the FM transmitter circuit.
FM Simulation
We used a special mode of SPICE which simulates a Frequency Modulated Signal, since it is extremely difficult to view the modulation of an RF carrier with a low frequency signal. Here is our SPICE file for this simulation. We set up the simulation with a much higher modulating signal so that the effect of the modulation could be seen with the high carrier frequency. The following plot is the result.
Simulated Frequency modulated signal
This plot also shows that the frequency does change about a natural carrier frequency of about 94 Mhz.
Frequency spectrum of simulated FM signal
The following two are close-ups of the previous plots. The first shows the high frequency portion of the modulated carrier wave from the left of the plot. The second shows the low frequency portion of the modulated carrier wave from the right of the plot.
Modulated frequency of about 90.9 Mhz
Modulated frequency of about 91.7 Mhz
Basic Oscillator
We simulated the actual oscillator to resonate at about 94 Mhz. The following plot shows a simulation of the oscillator of the circuit. The parameters are shown in the following SPICE circuit file.
Basic oscillator circuit result
Change in Inductance
We found that our circuit is very sensitive to changes in inductance and capacitance. We reinforced this idea by changing the inductor by 0.05 uH in our SPICE circuit file. We found that the frequency changed by almost 20 Mhz as shown in the following plot.
Result of changing the inductance
Based on the accuracy of our models, we concluded that our variable capacitor, C2, was approximately equal to 20pF and our inductor, L1, was about 0.15uH.
As always, one must think about the future. This circuit, being so elementary, begs to be improved. The following is a list of easily implemented improvements, followed by more challenging design changes.
Basic Improvements
Challenging Improvements
We hope you have enjoyed learning about FM transmissions. We hope you will use this information responsibly.