First of all, have a look at the circuit diagram. It’s basically a common base collpit oscillator transmitter.

FM transmitter part list
R1 – 100 Ohm, carbon film 1/4 watt
R2 – 10 kOhm, carbon film 1/4 watt
R3 – 22 kOhm, carbon film 1/4 watt
C1 – 68 pF, ceramic disc
C2 – 10 pF, ceramic disc
C3 – 68 pF, ceramic disc
C4 – 1 nF, ceramic disc
C5 – 1 nF, ceramic disc
C6 – 100 nF, ceramic disc
C7 – 470 uF, 10V electrolytic
C8 – 150 pF, ceramic disc
C9 – 150 pF, ceramic disc
L1 – 11.5 turn on 6 mm diameter, see text below
L2 – 3 turn on 3×2 mm ferrite bead, see text below
Q1 – BC548 transistor
U1 – AMS1117-3.3 LDO regulator

So, it uses 16 components excluding the wires, connectors and circuit board. As always, you can replace all the components with their nearest value counterparts. The antenna is just a peace of 75 cm single stranded wire, technically a quarter wave whip antenna. The coil L2 is specially important, it’s winded on a ferrite bead. The bead MUST be a RF ferrite bead, else the circuit wont work. You can salvage them from old TV balun, TV tuner box, DVD RF box, radios and so on. The coil L1 is winded over 6 mm diameter, 11.5 turns of single stranded hookup wire. It’s basically a RF choke coil. You can use any 3.3 volt LDO regulator instead of the AMS1117-3.3, but definitely not a Zener diode and resistor combo to get the regulated voltage.


Construction

You’ve to pay a little attention while constructing RF related circuits.
- Component leads should be trimmed to minimal.
- There should be no or lowest possible capacitance between two PCB tracks, anyway this can’t be avoided.
- Use as tittle solder as possible.
- Clean the solder flux thoroughly after soldering, preferably with alcohol.
- All components should be soldered tightly.
- Finally, enclose it inside a little metal case if possible.

I had constructed the circuit on a strip board, which is not fit for this purpose. Also messed up all the above rules, due to an extensive trial end error to find the right component values. But finally it worked, now the FM transmitter circuit is reasonably stable, without any frequency drifting. You can see the little ferrite bead in the above picture.

So why this miniature FM transmitter circuit is reasonably stable? I did some research and calculations before making the final circuit and choosing the components. I’m not going to extensive details, but these are main reasons why this transmitter is stable.

- A simple common base Collpits oscillator is a voltage controlled linear harmonic oscillator. In fact the frequency modulation is achieved by the varying voltage at the transistor’s base, due to varying P-N junction capacitance. So for a stable operation I’ve to keep the supply voltage as stable as possible.
- All oscillators generate some order harmonics along with the fundamental frequency. So I’ve to choose such a transistor which can operate at the fundamental frequency of the LC tank circuit, but not at the 2nd harmonic. BC548 is a good candidate for this purpose, which has a transition frequency of around 100 MHz, but also a larger noise figure.
- Ferrite cores tends to absorb higher frequencies much more than lower frequencies. Thus by using a ferrite core inductor at the LC tank circuit, it it minimizes the 2nd order and 3rd order harmonics further.
- Ferrite core inductors also tends to have higher Q factor than air core inductors, so the coil also improves the transmitter’s quality.


Range test and future improvement plans

Undoubtedly it’s a very low power FM transmitter circuit, it’s intended for stability, not range. In my tests it’s transmitting not more than 20 meters, but the audio quality is good enough, and no frequency drifts observed. But increasing the range is rather simple than increasing stability, it just needs a buffer stage and RF amplifier. It’s also transmitting AM waves, as the modulating signal is directly fed into the base of the transistor, though the AM modulation index is quite low. I’m planning to use a varactor diode for FM modulation in the future design.