FM Transmitter

[Hero999]

I have seen many of these little FM transmitter circuits but have never bother to sit down and analyse them in any; detail until now that it. I simply entered the audioguru's circuit into LT Spice and the sinewave was only 10mVp, I tweaked a few things and now it's 3Vp, which will deliver 60mW into a 75hm: load. I also decided to add a diode varactor to get rid of the crappy AM output.

I'm thinking about building this little transmitter. I will obviously need to add an amplifier but I don't know what gain I require. The only thing is I don't plan to use the 1N914, I'll use the 1N4001 as the varactor.

One thing I've figured out is that the tank capacitance only has to vary slightly to make the frequency deviate by 75kHz.

With a 100nH inductor assuming C is set to the appropriate capacitance:
@108MHz
+4.8fF for -75kHz
-4.801fF for +75kHz.

@88MHz
+8.872fF -75kHz
-8.875fF +75kHz

Another thing is I hear people referring to it as a Hartley or colpitts oscillator; I've done some research and it's neither of the aforementioned.

A real hartley oscillator
**broken link removed****broken link removed**

A real colpitts oscillator
**broken link removed****broken link removed**

I can't see any similarities between this FM oscillator and a either of the above.

I think it's more likely to be an oscillator based around the common base amplifier.
**broken link removed**

In my drawing the tank replaces Rc and C3 replaces Cb. C2 provides feedback from the output to the input, as the gain is Rc/Re, it's no surprise it oscillates when the collector is at the highest impedance.

What does everyone here think?

 

[audioguru]

Hi Hero,
Why do you think my circuit has much less power than yours?
1) The different biasing for the oscillator? Yes, a little.
2) The different transistors? Mine is better.
3) The much bigger RF coupling capacitor on yours? Yes, a little.
4) The emitter resistor on your amplifier? Mine is much better.
5) Because the simulation is screwed up? Probably.

My sim with part of my circuit in yours has an output the same as yours.
Your value of C2 is double the value of mine which boosts the output voltage by almost double.

 

[Hero999]

To be honest I only copied it from memory.

I wouldn't say your circuit is better as it produces a distorted waveform.

I agree simulations aren't perfect but they can at least give clues as to which configuration gives the best performance.

I'll have another play with it and find the optimum configuration.

Anyway doesn't anyone know what gain I require for my circuit which uses a varactor>

 

[TheVictim]

I wasn't aware that a plain 'ol 1N4001 could be used as a varactor. What capacitance range can you get out of it? Do you just put it in parallel with a fixed capacitance you want to add it to and reverse bias it with a control voltage?

 

[mneary]

The 1N4001 family is about 15 pF at a reverse bias of 4V, and decreases to about 5 pF at 20V. The best datasheet I've found is from OnSemi.

Other diodes have variable capacitance, but this one is so common that people like it. Some diodes (like 1N4148) don't have enough capacitance, and others are too lossy.

 

[audioguru]

This FM transistor uses a 2N4401 transistor as an FM modulator and two in parallel to tune the frequency:

 

[RadioRon]

The diode being used as a varactor is very tightly coupled into the oscillator resonant network via that 10 pF capacitor. This could be reduced if the modulator is simply too sensitive, as you mentioned.

 

[Hero999]

Most diodes can be used as varators and it helps if you're using a cfapacitor as the main tuning element and only relying on the diode for modulation because it doesn't take a great change in capacitance to chift the frequency by more than 75kHz.

I have posted a copy of the capacitance graph from the datasheet in the first post in this thread.