Deciphering an FM transmitter

[Gott_ist_tot]

Hi, I am attempting to figure out how an FM transmitter I'm building works. I was wondering if anyone had some suggestions on a resource for transistor circuits or just has some ideas of how to go about figuring it out. If anyone is interested, here is the schematic.

http://www.uoguelph.ca/~antoon/circ/fmt5.html

I believe that if I understand the 4 mini circuits built around each transmitter that would serve my purposes. I tried building the first stage (the 3904 on the left) by itself and I do not understand its output. It actually seems to be reducing the gain of the input signal. I don't know if its purpose is for some sort of oscillation. The second one looks like a common emitter amplifier. The 7001 stages I am completely unfamiliar with. Thanks for any suggestions. I really just want to understand what I'm building and I do not have much experience with circuits.

 

[chroot]

Q1 is a common-emitter amplifier stage, biased with R1 and R3. The microphone input is ac-coupled into the base of Q1 via C9.

The output of that stage is ac-coupled via C4 into the second stage, which is also a common-emitter amplifier. The RC formed by R6 and C10 is used for emitter degeneration, to selectively decrease gain at audio frequencies (time constant ~20 Hz) and improve stability and reduce distortion.

These first two stages are together just a basic discrete audio amplifier; Q2 is the "output stage" of the audio pre-amp.

Q3 is configured as a Colpitts oscillator. The frequency is determined by C7, C8, and L2 by the formula

$\frac{1}{2 \pi \sqrt{L_2 \cdot \frac{C_7 \cdot C_8}{C_7 + C_8}} }$

Or, about 100 MHz

The audio signal is ac-coupled via C11 into the collector of the oscillator (and biased up by diode D1). As the audio signal tugs on the collector of Q3, the oscillation frequency is modulated. Viola, FM modulation.

Q4 is the power-amp stage, also in common-emitter configuration, again with emitter degeneration to improve stability. The load of this stage is antenna itself.

- Warren

 

[Gott_ist_tot]

Wow, thanks. That was really helpful. One other question came up. You may not be able to answer this. It may not even be a question to ask here. I plan on emailing the drawer of the circuit. But there seems to be a RC circuit missing (R5 and C5). From what I've seen on the circuit, my preamp (Q1 and Q2) works. But Q3 doesn't seem to be getting any signal. The output on an oscilloscope is around 100 MHz but I have no evidence that it is the signal I am feeding in. In fact, I am quite sure it isn't. Could an RC circuit be needed on the other side of the diode, D1, separated by the output of C11? Thanks.

 

[chroot]

Yes, I see that he lists R5 and C5 on the bill of materials, but they are not present on the schematic. This may be a "missing" subcircuit, but it's more likely that he just made a mistake while renumbering his designators before publishing the circuit.

I don't know why you say that "Q3 doesn't seem to be getting any signal." Clear as day, there's a 5p capacitor there, C11, coupling the output of Q2 into Q3. How exactly are you determining that Q3 is not "getting any signal?"

Keep in mind that the actual modulation of the 100 MHz carrier may only be a tiny deviation -- say, 99.5 MHz to 100.5 MHz -- which would be virtually impossible to see visually on an oscilloscope. Do you have a spectrum analzyer?

As it appears to me, the circuit is correct as drawn. I have not built it or simulated it, though, so I cannot be entirely sure. Either way, I see no reason to start tossing in RC's anywhere.

- Warren

 

[Gott_ist_tot]

Yeah, I might just have a short somewhere that I can't find. The reason I thought it wasn't working is that the output signal (at the antenna for the complete circuit) is the exact same with or without an input from a function generator. I can definitely see the correct input signal just before C11. After that I see no hint of an input signal, even when I adjust the frequency and amplitude of the input signal. The oscilloscope probably is just a bad way of viewing it. I do not have a spectrum analyzer but that is a good idea. I'll play around with it some more and try it on a radio again. Thanks.

 

[chroot]

Well, the standard channel bandwidth for commercial FM radio transmission is only 200 kHz. That means the frequency of the output should only vary by at most 200 kHz, or about a quarter of a percent. There's pretty much no question about it -- that will be invisible on a 'scope. If you see the correct signal on one side of C11, well, there's also no question that it's getting fed into the oscillator correctly. Once you've assembled the whole thing, let us know how it goes with an actual radio receiver.

- Warren