LOW POWER FM TRANSMITTER
I did this project to my friend Hebert that was requesting a low power fm pirate radio. The output power is around +35 dBm ( 3.16 watts) over 50 ohms load and +24 volts power supply.
The project consists in a hartley oscillator ( modulated VCO), and tree stages ( Class A , common emitter) of RF amplifiers. The first one , you can use a MAV-11 (from MiniCircuits) that have +12 dB of gain and a good 1 dB compression point of +17.5 dBm, at my project I used a MWA-130 ( from Motorola) as driver stage.
At second stage I used a 2N5109 (Motorola) , but you can also use a 2N3866, and at final stage a 2N3553. With proper biasing and setting up the trimmers to maximum output power you can achieve around 2.8 watts of EIRP using a 0 dBi antenna.
A good irradiant system is also important, as well the coax cable that I recommend low loss cables if you are using a long cable distances. RG-58A haves a 67% of propagation velocity and 21.65 dB/100m of insertion loss at 146 MHz.
I also recommend you use a low pass filter at final stage to prevent harmonics products cause harmful interferences on the MARS/CAP systems. Here in Brazil, many pirate radios don't use such filter causing many trouble at MARS service. At final you can found many schematics, and of couse a suggested low pass chebyshev filter and the respective response frequency curve.You will experience a low insertion loss within this filter, but not so expressive.
Understanding the project
The VCO ( Voltage Control Oscillator)
The VCO can operate as free ( open loop) or synthesized (PLL) oscillator, you just must insert a jumper to close the loop for PLL or release it to free oscillator operation. The frequency deviation is controled by varactor diode that change the instrinsic capacitance within the voltage audio input causing a frequency deviation. Note: set it to ± 75 KHz to FM pattern broadcast operation.
Another varactor BB809 control the operation frequency , 200 KHz step increments is allowed at PLL schematic circuit at final of this page, the oscillator was constructed to operate in FM broadcast (88 to 108 MHZ) without any change in components.
As hartley oscillator, if you wanna change the operation frequency just change the numbers of turns in the inductors at LC tank circuit. As you know, decreasing the numbers of turns, you get upper frequencies, and vice-versa.
Pay attention to start up the oscillator if you wanna change it to operate in high frequency than FM braodcast. Following the Barkhausen Criteria, we have a feedback in a simple amplifier and also sustain the oscillations, so the K fator( Loop gain) must be equal to one. The MPF 102 oscillator transistor operates quite good until 150 MHz, a high Ft transistor will be required to high frequency operation.
If you could use a network of varactors diode ( paralel association) to control the frequency deviation would be interesting because the BB809 have hyper abrupt curves. Maybe MV209 in paralel produces less distortion and also ensure more linearity.
At input ( audio signal) , I did a pre-emphasis network to garantee that modulation index at high audio frequencies ( up 1KHz) . At FM radio, the relation S/N depends of modulation index employed. The index decrease as modulation frequency increases, to a constant frequency deviation. Pre-emphasis is a high pass filter, that the cutoff frequency is slight upper of a maximum modulated frequency. The filter must provide 6dB/octave, that can be stablished by a series capacitor with the patch audio signal.
Between the oscillator and driver, a buffer stage was inserted to prevent oscillations and increase isolation, high impedance at gate ( oscillator side) and low impedance at source terminal.
At MAV-11 output there are a sample frequency output to the pre-scaler ( if using PLL), and under the conditions of a high power, I added a resistive PI attenuator of 8 dB, it prevents not overload the pre-scaler input.
RF section ( driver, PA)
The driver stage is the own MAV-11 with 12 dB of gain, I choose these part because it haves a high 1 dB compression point, considering that our oscillator drives around +3 dBm, and the MAV-11 haves +17.5 dBm of 1 dB comp., we have a margin of 2.5 dB that is enough to assure against IMD's at one tone.
The following stages are class A configuration, each transistor after the MMIC provides 10 dB of gain in cascade configuration.
In frequency modulation, we don't need to concern about high linearity transistor response, I could choose another biasing class like class C at final stage, of course in driver stage I not recommend you do it because the base-emitter junction is reversed biased and you need at least 0,65 volts (Vbe) to transistor start to conduct any signal.
Between the driver and PA stages, there are L-C matching networks to ensure that the maxiumum signal (S21) is being drived, and minimum of reflection is returning (S11).
At final stage there are a matching network to 50 ohms antenna system, and obvious you must add a low pass filter to reduce spurious harmonics and adjacents interference. Remember that eletronic devices that is not of agreement with FCC part 15 may cause harmful interference. The project is of your total responsability.
To indicate that the RF energy signal is "on the air", I mean transmitted, a LED brights more when you turn the trimmers indicating that more power is being delivered to the load.
I hope you have a good entertainement with this bug.