I found an old smoke detector - and when I noticed that the transistor used was an S9014 - I brought it upon myself to make a transmitter using only the parts inside.

The only "outside thing" I used is solder, because I'm not *completely* insane.

Supplies

The supplies I salvaged from the smoke detector that were actually of use were as follows:

1x S9014 NPN Transistor (The heart of the circuit)

1x Piezo Transducer (Used as the microphone)

1x 2.2 M Resistor & 1x 920 K Resistor (Used as the voltage divider, explained later)

1x 150 K Resistor

1x 470 K Resistor

1x 10 nF Capacitor

1x 10 pF Capacitor

1x 820 pF Capacitor

1x Ionization chamber metal casing (Used as the circuit board - like Manhattan Style)

1x Spring (Used for the inductor. I was going to use wire but the spring was very stiff so it's more stable.)

Step 1: Soldering the Piezo

This is a tedious step! Too much heat and you will destroy the transducer, they are very fragile.

For the centre, place a piece of wire and drop some solder onto it. The iron should never touch the centre of the piezo as this will probably destroy it.

For the outer rim, scratch away some of the metal to expose the fresh metal. Use a good amount of flux for this part, and place some solder on the scratched part. You can hold the iron here for a couple of seconds max, and then stick the wire into the melted solder.

Maybe you're lucky and the piezo already has wires - who knows!

Step 2: Preparing the "board"

I used the casing for the ionization chamber. This will be the common ground for your circuit.

Scratch away some of the metal to expose the fresh shininess. With flux, hold the soldering iron on the scratched away area and apply some solder. Leave it for a few seconds and it should flatten out a bit, indicating it's ready to take an electronic component. You will only need to do this a few times - not many parts will connect directly to ground.

Step 3: The Most Basic, Terrible Circuit

This is the initial, bare-bone circuit which we will improve upon soon.

The battery or power supply (9 volts DC) is connected by adding the positive lead to the spot between the 820 pF capacitor and the spring inductor. The negative lead is connected to the bottom rail of the circuit (the emitter is grounded through the 470 ohm resistor)

Solder everything into place and there should be some output. Chances are, the fundamental frequency, or one of its harmonics, will fall into the 3m FM broadcast band (88-108 MHz). If you do not hear anything on an FM radio, I would suggest you get an RF power meter to ensure that it's transmitting at all.

If you don't have an RF power meter, compress/expand the coil, and see if you can hear anything on your radio. If you still haven't heard anything, it's likely something is wrong in the circuit.

If you get the circuit to work, you may notice you have to scream into the piezo for any response, or that even going near it will make it drift off frequency. This is for three reasons - there is no voltage regulation on the base, and hand capacitance (literally capacitance between you, ground, and the circuit) changing the frequency. The piezo is weak because, well, the piezo is weak as a microphone. We fix this in the next few steps.

Step 4: Voltage Stabilization on the Base

The voltage between the base and ground of the transistor controls the frequency of the oscillator. A lot of different variables can change the voltage and therefore the frequency - so we will add a simple voltage divider.

By measuring the voltage across the microphone while the circuit is powered, I found it was about 2.5 volts.

We want to keep this voltage exact, so we will make a power supply to keep it at 2.5.

Step 5: Adding the Voltage Divider, Solving Hand Capacitance and Increasing Mic Gain

This is the new circuit - it now has a voltage divider to make it slightly more stable - as long as you never touch it that is.

To make sure I don't touch the circuit, I added a paper cone. I also shoved the piezo inside of the ionization chamber cavity so that it doesn't flop around (and change the frequency slightly whenever it does). The cone, of course, increases the microphone gain as it concentrates the audio.

Now, the frequency stability is only slightly terrible (:

Step 6: Some Tests

Here are some SSTV images I transmitted (and decoded using MMSSTV) by sticking my phone's speaker into the cone. The audio file attached is me saying "victor alpha one alpha lima oscar" (VA1ALO, my callsign) into the microphone, and once again but in a higher tone of voice. The piezo reacts to higher frequencies besas the resonant frequency is about 3 kHz.