All the circuits described in this article must be used for
experimental purposes only
and must not be used for any covert activity.
These circuits are very powerful, in that they are very hard to detect
and will pick-up the slightest whisper and transmit the information 3km
and more. Because of this, government laws are very strict and will
come down very heavily on any mis-use.
The purpose of this article is to compare different circuits and
explain the features that contribute to a good design. Some of these
circuits (the best of each design) are available from Talking
Electronics in kit-form and/or ready assembled. See the
Electronics website for the list of kits and devices.
Get this handy data
Just click HERE
The first group of circuits we will discuss are FM TRANSMITTERS. They
can be called SPY TRANSMITTERS, FM BUGS, or a number of
other interesting names. They all do the same thing. They transmit on
the FM band in the range 88MHz to 108MHz.
Most of them can be adjusted to transmit above or below the band and you
need a radio that will pick up these frequencies, to detect them.
Since the FM band is almost entirely filled with radio stations, we will
be providing details on how to adjust a radio so it will scan EITHER
above the band or below the band. Most radios can only be adjusted 10MHz
above or below but this will be enough to provide a "blank" space for
FM transmission provides perfect quality and when one of these
transmitters is used in a house and received on a good quality radio,
you cannot tell if the person is actually talking in the next room or
via an FM link, through a radio.
This means all the FM bugs have the same perfect audio, but some
circuits will detect fainter sounds and others will transmit further.
Some circuits can be handled without drifting off-frequency and others
are designed to be very small or fit on top of a 9v battery.
By building these circuits you will learn an enormous amount about high
frequency, audio and getting the maximum output with the least current.
This is the main aim of this article. It will add a number of "building
blocks" to your understanding of electronics.
Before I start, there are two things that particularly annoy me. The
first is a circuit diagram with C1, R1 etc and a parts list identifying
the values. Circuit diagrams like this are obviously drawn by a non-electronics
person. The whole concept of looking at a circuit diagram and seeing the
values gives the reader an indication of how each section will work. A
section may be operating very lightly with high value components or it
may be working very hard with low value components. The whole idea of
providing a circuit diagram with marked components is to give the reader an immediate
understanding of how the circuit is operating.
The second thing that annoys me is the labelling of parts
on a PC board as R1, C1 etc.
Again, the board has been designed by a non-technical person.
Why design a board without component values? Do they think the
values will change? How can you assemble a board without referring to a
The whole purpose of well-designed PC board is to build it without referring to any other data.
Keep this in mind when designed your own boards.
Also, don't name your boards "A51/834-2." Give then a name you can
remember or one that refers to the application it will perform.
1 TRANSISTOR CIRCUITS
There are a number of 1 transistor FM transistors on the market in
kit-form and already assembled.
These circuits are interesting to look at but do not really perform very
1. They do not have a good transmitting range.
2. They do not detect low-level sound, and
3. They do not operate very well on 1.5v. No transmitter can be expected
to operate very well on 1.5v. If you want to use a single cell, use
a lithium cell as it produces 3v.
4. Some have a coil etched on the PC board. No FM transmitter will
perform very well with a coil etched on the board.
Why use 1.5v?????
Transistors do not operate very well below 0.9v and the collector load
resistor needs a small voltage so it can perform its task (the same
applies to an emitter resistor) and thus the lowest voltage for a circuit
is 1.5v. If only a single cells is used, there is no allowance for a
voltage-drop as the cell becomes depleted. Always use 3v as the lowest
THE SIMPLEST CIRCUIT
The following circuit is the simplest FM circuit you can get. It has no
microphone but the coil is so MICROPHONIC that it will pick up noises in
the room via vibrations on a table.
The circuit does not have any section that determines the frequency. In
the next circuit and all those that follow, the section that determines
the frequency of operation is called the TUNED CIRCUIT or TANK CIRCUIT
and consists of a coil and capacitor. The transistor and components
surrounding the tuned circuit simply keep the tuned circuit operating at
its RESONANT FREQUENCY. This circuit does not have this feature. The
transistor turns on via the 47k and this puts a pulse through the 15
turn winding. The magnetic flux from this winding passes through the 6
turn winding and into the base of the transistor via the 22n capacitor.
This pulse is amplified by the transistor and the circuit is kept
The frequency is determined by the 6 turn coil. By moving
the turns together, the frequency will decrease. The circuit
transmits at 90MHz. It has a very poor range and consumes 16mA.
THE SIMPLEST BUG
The components soldered to the 2 cells
Rear view of the simplest bug
After making a transmitter, you will want to know if it is
transmitting. In the case above, the circuit will only produce a carrier
and this will be heard on the radio as a "quiet spot."
Rather than chasing up and down the dial, Talking Electronics has
produced a piece of test equipment to let you know the bug is
transmitting and the approx frequency of transmission.
It is called
FIELD STRENGTH METER MkII.
The photo below shows the Field Strength Meter near the bug. The plastic
knob on the trimmer allows adjustment without affecting the detecting
circuit. Simply turn the knob (with the two antennas
near each other) and the 3 LEDs on the project will illuminate.
Field Strength Meter and Bug
So far we have seen an unstable circuit in action. Placing a finger
near the bug will change the frequency. This is totally unsuitable.
A ONE TRANSISTOR CIRCUIT
The next circuit uses a TUNED CIRCUIT or TANK CIRCUIT to create the
operating frequency. This is clearly shown in the diagram. For best
performance the circuit should be built on a PC board with all
components fitted close to each other. The photo below shows the circuit
using a coil etched on the board. This type of coil is totally
unsuitable. It does not have a high "Q" and the range is very
poor. The board cannot be touched as the capacitance of your body
causes the circuit to drift. A wound coil will improve the stability
considerably. See photos below for the details of a wound coil.
A one transistor
Do not use an etched coil
Here is the 1-transistor circuit produced by Darren
Dazaro on a home-made PC board and heat-shrunk so the air-trimmer is
adjustable via a small hole.
The PC board drilled ready for fitting the components
The 10 components (plus heatshrink, battery and wire)
The components mounted on the board
The board ready for fitting into heatshrink
The finished bug with "studs" for the battery
and a cut-out for the air trimmer
a circuit with a very bad layout. The circuit may work but it relies on
a "Q-factor" from the coil and capacitor in the tank circuit to produce
a high voltage. This high voltage gives the circuit a good range.
Firstly the coil and capacitor should be near each other. The coil
should not have long leads. and a 22n capacitor should be across the
supply to give the circuit better performance. The value of C2 is too
high. It should be 10p. The coil should be 5 turns. The electret mic
should not be connected directly to the base of the transistor. As you
can see, the circuit is full of faults.
You can learn a lot from other people's mistakes.
Before we go to an improved design, here is an unusual circuit using
a PNP (BC 557) transistor. Firstly, PNP transistors do not work as well
as NPN transistors. I would reverse the 4k7
and electret mic as the voltage between base and 0v rail is very small
and the 4k7 is not biasing the transistor - it is not needed! The range will be 50 to 100
metres and the current is about 3mA.
Note: A reader built this circuit. It did not work!!
Try putting the 47p across the coil. The 33p may need to be reduced to
It is just a BAD design, but a good challenge to see if you can get this
type of design to work.
Simple 1 transistor FM
The 22n is not shown. This
is a later addition.
AN IMPROVED DESIGN
This design uses a "slug tuned coil" to
set the frequency. This means the slug can be screwed in and out of the
coil. This type of circuit does not offer any improvement in stability
over the previous circuit. (In later circuits we will show how to
improve stability. The main way to improve stability is to add a
"buffer" stage. This separates the oscillator stage from the output.)
The antenna is connected to the collector of the transistor and
this "loads" the circuit and will cause drift if the bug is touched. The
range of this circuit is about 200 metres and current consumption is
about 7mA. The microphone has been separated from the oscillator and
this allows the gain of the microphone to be set via the 22k resistor.
Lowering the resistor will make the microphone more sensitive. This
circuit is the best you can get with one transistor.
If you want more stability, the antenna can be tapped off the top of the
tank circuit. This actually does two things. It keeps the antenna away
from the highly active collector and turns the coil into an
auto-transformer where the energy from the 8 turns is passed to a single
turn. This effectively increases the current into the antenna. And that
is exactly what we want.
The range is not as far but the stability is better. The frequency will
not drift as much when the bug is held. As the tap is taken towards the
collector, the output increase but the stability deceases.
The next circuit has been picked out for its low emitter resistor on the
This resistor does not have to be a very low value as the transistor is
working at its maximum potential, due to the high frequency and a low emitter resistor will
simply consume more current without improving the output.
The emitter resistor is too low
Two photos of the bug
STEREO TO MONO
To combine two
channels to a mono output, the following circuit can be used:
2 TRANSISTOR CIRCUITS
The next progressive step is to add a transistor to give the
electret microphone more sensitivity. The electret microphone contains a
Field Effect Transistor and you can consider it to be a stage of
amplification. That's why the electret microphone has a very good
A further stage of amplification will give the bug extremely good
sensitivity and you will be able to pick up the sound of a pin dropping
on a wooden floor.
Many of the 1 transistor circuits over-drive the microphone and this
will create a noise like bacon-and-eggs frying. The microphone's used by
Talking Electronics require a load resistor of 47k for a 6v supply and
22k for a 3v supply. The voltage across the microphone is about 300mV to
600mV. It will produce an audio waveform of about 2-20mV.
Only a very simple self-biasing common-emitter stage is needed for the
audio amplifier. This
will give a gain of approx 70 for a 3v supply. The next circuit shows
this audio amplifier, added to the previous transmitter circuit. This
circuit is the best design using 2 transistors on a 3v supply. The
circuit takes about 7mA and produces a range of about 200 - 400metres.
2 Transistor FM Transmitter
Five points to note in the circuit above:
1. The tank circuit has a fixed 39p and is adjusted by a 2-10p trimmer.
The coil is stretched to get the desired position on the band and the
trimmer fine tunes the location.
2. The microphone coupling is a 22n ceramic. This value is sufficient
as its capacitive reactance at 3-4kHz is about 4k and
the input to the audio stage is fairly high, as noted by the 1M on the
3. The 1u between the audio stage and oscillator is needed as the base
has a lower impedance as noted by the 47k base-bias resistor.
4. The 22n across the power rails is needed to keep the rails "tight."
Its impedance at 100MHz is much less than one ohm and it improves the
performance of the oscillator enormously.
5. The coil in the tank circuit is 5 turns of enameled wire with air
core. This is much better than a coil made on a PC board and is cheaper
than a bought inductor. The secret to long range is high activity in the
oscillator stage. The tank circuit (made up of the coil and capacitors
across it) will produce a voltage higher than the supply voltage due to
the effect known as "collapsing magnetic field" and this occurs
when the coil collapses and passes its reverse voltage to the capacitor.
The antenna is also connected to this point and it receives this high
waveform and passes the energy to the atmosphere as electromagnetic
When the circuit is tightly constructed on a PC board, the frequency
will not drift very much if the antenna is touched. This is due to the
circuit design and layout as well as the use of large-value capacitors
in the oscillator. If low value capacitors are used, the effect of your
body has a greater effect on changing the frequency.
The only way to get a higher output from two transistors is to
increase the supply voltage.
The following circuit is available from Talking Electronics as a
surface-mount kit, with some components through-hole. The
project is called THE VOYAGER.
Voyager on a 9v Battery
All the elements of good design have been achieved in this project.
The circuit has a higher output than the 3v circuit above, but
most of the voltage is lost across the emitter resistor and not
converted to RF. The main advantage of this design is being able to
connect to a 9v battery. In a technical sense, about half the energy is
wasted as the stages actually require about 4v - 5v for maximum output.
The Voyager has been copied by many kit-makers but none has surpassed
Here is a "knock-off" of our older design. It is mounted flat on a 9v battery:
Here are 2 more two-transistor circuits using a 9v
supply. We have also included the technical limitations of the
Faults with this circuit:
1. Very low microphone load resistor.
2. 4u7 not needed from microphone. 22n is sufficient.
3. 3p3 is very low for BC 547. May need to be higher. - 10p preferred.
4. Bridge biasing of audio stage is not needed. Simple biasing is
5. Base biasing of oscillator is very wasteful in current.
6. 22n is very high for base bias of oscillator - restricts incoming
The circuit above
has a 56p on the emitter. This capacitor has no effect on improving the
operation of the circuit and merely requires the capacitor connected
between the collector and emitter to be increased to counter the losses
delivered to the 56p.
The two 56p on the output do not improve the performance as they are
equivalent to 28p across L2 and this forms a tuned circuit. The aim of
the output stage is to be untuned so that any frequency can be selected
for the oscillator. If the output is "tuned" it will only have gain at
one particular frequency and will have to be re-tuned every time the
main frequency is altered.
A BIRD'S NEST DESIGN
Faults with this circuit:
1. Load resistor for microphone is very low - should be 47k
2.10u on output of microphone is not needed - 22n is
3. Current through audio state is very high. Load resistor
should be 47k
4. Base biasing of oscillator very wasteful.
5. Load resistor for oscillator (emitter resistor) is very low
6. Feedback capacitor for oscillator should be 10p.
7. No ceramic in tank circuit. Adding a ceramic makes it
easier to adjust
the trimmer capacitor.
8. No capacitor across the battery.
See the layout below:
Birds-nest of above circuit showing how tight the circuit can be
There is nothing wrong with a
bird's nest. It is very easy to experiment with
components and wiring that can be seen and changed without having to
work on a PC board. The only problem with the bird's nest above is
the lack of an earth plane. When you have an earth plane, the signal
can push against the large mass of an earth rail (or battery) so that it
can push the signal out the antenna.
The circuit on proto-typing board - a quick way to build a
The oscillator components must be kept near each other,
otherwise the circuit will not oscillate!