Original AM Micropower Transmitter

The picture to the left is a high quality radio transmitter for the A.M.
broadcast band. The transmitter legally operates with "micro-power"
and will not set any distance records but, unlike simpler designs, the
frequency stays put and the fidelity is excellent. Although the
schematic looks somewhat complex, the circuitry is easy to build and
adjust for experimenters with a little "tweaking" experience. A simple
output meter confirms proper signal level and checks antenna tuning
while "on the air". Add an audio mixer, tape recorder, and perhaps a CD
player and have a near-professional micro-power station.

1
Most values are not critical but a few choices must be made
carefully for best results. The output tank is tuned to the
crystal frequency by selecting the values from the chart above.
For example, for a 1 MHz transmitter, the chart indicates 500
pf and 35 uh. A 33uH and 550pF (470 + 82, perhaps) would be
a good start. This chart assumes that a 220 pf capacitor is
already connected between the collector and base of the
output transistor as indicated in the schematic so the indicated
capacitance is in addition to the 220 pf. A variable inductor or
capacitor will allow the tank to be fine-tuned for the maximum
meter reading with no antenna connected (a few volts with a
10 megohm voltmeter or about 50 microamps with a current
meter). After the antenna is connected, the loading inductor in
series with the antenna is selected for the minimum meter
reading (best antenna loading). (A 3 foot antenna will need
about 820 uH for a 1.6 MHz output frequency.) Longer

2
antennas or higher frequencies need less inductance and
shorter antennas or lower frequencies will need more. The
meter reading should drop by more than half with a reasonably
good antenna but the reading can be ignored if sufficient
transmit range is achieved. The antenna, which is short
relative to the wavelength, is hard to match well because it has
a very low radiation resistance in series with a very small
capacitor. (The power dissipated in the radiation resistance is
the power that is transmitted.) The loading coil helps to
resonate out some of the series capacity resulting in more
antenna current and thus more radiated power. Some retuning
of the tank may be desirable when the loading coil value is
changed. A remote radio playing back through a baby monitor
or walkie-talkie makes a good signal quality monitor for
antenna tuning and positioning.

Note: The antenna in the picture above is just a short metal rod
from an old fireplace screen stuck through an important-
looking insulator strictly for appearance. It's really too short for
optimum range.

The crystal can be practically any surplus crystal with a

fundamental frequency between 530 kHz and 1.7 MHz in 10
kHz increments but the higher frequencies work best. Choose a
crystal frequency away from strong local stations at or above
800 kHz for best transmit range. Proper operation of the
oscillator may be verified by probing the junction of the two
1000 pf capacitors with a high impedance oscilloscope probe
connected to a scope or frequency counter. Full modulation is
achieved by applying about 2 volts peak-to-peak to the base of
the current source transistor in the differential amplifier. The
modulation voltage varies the current in the diff. amp. away
from the nominal 20 ma. setpoint and this modulated current is
converted to a clean, high voltage sinewave by the output
tuning circuit. The modulated signal may be observed with an
oscilloscope connected to the antenna terminal if desired.

3
The photo above shows a prototype built with metal transistors
(just for looks!) and with a few additions like the variable
capacitor in series with the crystal for fine tuning and the
variable inductor in the collector of the output transistor.
Circuit construction is mostly non-critical but a few points
should be observed. Ground-plane is not mandatory but it
helps control parasitic feedback elements when less than
perfect layout techniques are used. The two capacitors across
the base-collector leads of the diff-amp transistors should have
short leads. Bypass the 15 volt supply well, perhaps with
additional 1 uF capacitors not shown in the schematic. The 100
ohm emitter resistor in the modulator may be bypassed with a
22 ohm resistor in series with a 470 uf capacitor to increase
the modulation sensitivity to about 1 volt peak-to-peak which is
typical of many sources. Eliminating the 22 ohm resistor will
increase sensitivity to under 100 mv but the linearity will suffer
somewhat.

4
An amplifying audio mixer may be added as shown in fig. 2 if
more than one audio source is to be used. The gain resistor
might be near 2.8k for typical 300 mv sources or considerably
higher for lower level sources. If the signal level is different for
each source then vary the 600 ohm resistors to compensate. A
larger resistor will reduce the gain. Set the main gain resistor
for the weakest source then increase the 600 ohm resistors in
the other channels for the proper balance. A fancy mixer panel
could be constructed with potentiometers in place of the
resistors. Remember that some op-amps are not sufficiently
fast to amplify high fidelity audio. For simplicity, choose an
internally-compensated audio op-amp such as the LM833.
Since the LM833 is a dual op-amp the second amp could be
used as a separate pre-amp for a microphone or other low-
level sources using the same schematic as the mixer. The
output of this amp simply feeds one of the mixer source inputs.

Applications:

A continuous-loop tape could give sales information to passing cars.

Place a sign that says, "tune to xxxAM for information," next to the
house or car that is for sale.
Transmit special seasonal music at Christmas or Halloween to
enhance your decorations. (Use a similar sign.)
Transmit a cassette player or other audio source to the car radio for
better sound.
Make a pair of toy AM band two-way radios by adding inexpensive

5
 AM radios. Or talk between cars on a trip using the car radio for
reception.
Make a baby monitor that works with any AM receiver.
Transmit control tones to a number of cheap AM receivers for
unusual remote control applications.
Build a fully functional radio station for the kids - complete with vu
meters, slide faders, and an "on the air" light.

Besides making a nice general purpose radio transmitter the

Personal Radio Station is suitable for some nice practical jokes:

Hide the transmitter with a cassette tape player in your

personal effects as you ride in the back seat of a friend's car.
(Leave out the meter circuit to keep the size down.) Ask your
friend to tune in that new radio station - since your transmitter
is crystal controlled it will be at the right place on the dial.
What your victim hears is up to you. The circuit will work
reasonably well with a single 9 volt battery instead of 15 volts.
How about a less than desirable school lunch menu for the
kids. Or, if you are younger, an unexpected school closing for
the day. (I didn't really suggest that one, did I?) A news
announcement of your marriage proposal will get results. Local
news personalities will probably be delighted to help make a
tape.

The Law

Part 15 of Title 47 of the Federal Code of Regulations

addresses the construction of homemade AM band
transmitters. The three most germane paragraphs follow:
§ 15.5 (General conditions of operation)

(a) Persons operating intentional or unintentional radiators shall not be deemed to

have any vested or recognizable right to continued use of any given frequency by
virtue of prior registration or certification of equipment, or for power line carrier
systems, on the basis of prior notification of use pursuant to § 90- 63(g) of this
chapter.

(b) Operation of an intentional, unintentional, or incidental radiator is subject to the

conditions that no harmful interference is caused and that interference must be
accepted that may be caused by the operation of an authorized radio station, by an
other intentional or unintentional radiators by industrial, scientific and medical

6
(ISM) equipment, or by an incidental radiator.

(c) The operator of a radio frequency device shall be required to cease operating the
device upon notification by a Commission representative that the device is causing
harmful interference. Operation shall not resume until the condition causing the
harmful interference has been corrected.

(d) Intentional radiators that produce Class B emissions (damped wave) are
prohibited.

§ 15.23 Home-built devices.

(a) Equipment authorization is not required for devices that are not marketed, are not
constructed from a kit, and are built in quantities of five or less for personal use.
(b) It is recognized that the individual builder of home-built equipment may not
possess the means to perform the measurements for determining compliance with
the regulations. In this case, the builder is expected to employ good engineering
practices to meet the specified technical standards to the greatest extent
practicable. The provisions of § 15.5 apply to this equipment.

§ 15.219 Operation in the band 510-1705 kHz.

(a) The total input power to the final radio frequency stage (exclusive of filament or
heater power) shall not exceed 100\milliwatts.
(b) The total length of the transmission line, antenna and ground lead (if used) shall
not exceed 3 meters.
(c) All emissions below 510 kHz or above 1705 kHz shall be attenuated at least 20 dB
below the level of the unmodulated carrier. Determination of compliance with the 20
dB attenuation specification may be based on measurements at the intentional
radiator's antenna output terminal unless the intentional radiator uses a permanently
attached antenna, in which case compliance shall be demonstrated by measuring the
radiated emissions.

In this circuit, the final radio frequency stage is the transistor

connected to the output tank. This transistor conducts one-half
of the bias current flowing through the modulator transistor
which is set to 20 ma in the circuit as shown. This current may
be determined by measuring the voltage across the 100 ohm
resistor. The output transistor drops about two-thirds of the
power supply voltage which is 10 volts with the 15 volt supply.
The power dissipated in the output stage is therefore 10 ma
times 10 volts which is the legal limit of 100 mw. An antenna
9.8 feet long is the legal limit and is more than adequate if a
proper loading choke is selected. In fact, an antenna only a few
feet long is more manageable and may be adequate in many
applications. Harmonic content of the circuit as shown was
measured at the output terminal to be 27 dB below the carrier
when tested at 1.6 MHz. If the tank values are selected near
the values suggested by the chart, similar performance should

7
be achieved. The connection of a properly loaded antenna will
further filter the radiated signal so the device should be well
inside the technical requirements.

Copyright, 1995-2002

Charles Wenzel

Improved Circuit
The following circuit is an improved version of the transmitter
above. It features a high Q pot core autotransformer that
provides a very high voltage to the antenna, greatly improving
the range and an improved crystal oscillator section. (Also
seephono oscillator for a tunable version.)

8
The pot core is an 1811 size high-Q ferrite core with an AL of
250nH/T2 . Two complete turns are wound on the bobbin, a loop
is brought out for the tap, and 28 more turns are wound to
complete the coil. Notice the knot in one end of the wire to
help identify the ends after assembly. The bobbin is inserted in
the core halves, the core halves are held together with a weak
clamp (a strong clamp can break the core), and a couple of
drops of epoxy or hot melt glue are applied to the outside of
the core halves. Do not get glue on the faces of the core, the
halves must be held tightly together BEFORE glue is applied.

9
The transformer Q combined with the turns ratio is selected to
give high antenna voltage without clipping on the peaks of the
modulation and without excessively limiting the bandwidth of
the transmitted signal. The output tuning capacitor is adjusted
for the maximum field strength and will be near 20 pF at the
top end of the band and near 80pF at 1 MHz. The diode/meter
circuit in the first design may be connected to the collector of
this new design but the tuner is adjusted for the maximum
meter reading.

The prototype is built on a piece of Vectorboard with a nice

multi-turn trimmer for tuning the antenna. Longer antennas or
antennas that have more capacitance may require less
inductance and a few turns may be removed from the pot core.

10
If you prefer to have a tunable transmitter, consider
the oscillator circuit below (second schematic). The AM radio
oscillator coil will give excellent stability. Simply connect the
output directly to the base of the leftmost NPN in the
differential amplifier and adjust the 500 ohm potentiometer to
get about 1 or 2 volts on the collector of the oscillator. The
circuit will work fine as-is at 15 volts but raising the emitter
resistor from 470 ohms to 1k will save a little power.

Field Strength Meter

Here is a simple field strength meter that is helpful when
tuning the output stage:

The circuit draws less than 10 uA with no signal so no switch is

required. The variable capacitor is adjusted to tune the meter
to the desired frequency by adjusting for the highest meter
reading when held near the transmitter's antenna. If the meter
reaches full scale during tuning or use, move the meter further

11
from the antenna. (As you peak up a transmitter, you may
need to move the meter several times to keep it on scale.) The
MPSA18 may be replaced by other high gain NPN transistors, if
desired. This meter has an "expanded scale" in that it goes
from a zero reading to full scale over a fairly small signal level
change making fine tuning easy.

12
Point-to-point wiring is fine for this low frequency circuit.
Remember, the meter DOES draw current when there is a

13
meter reading above zero so don't leave it near the transmitter
for long periods of time if you value you battery!

Note about part 15 AM band transmitters:

The above circuits do not represent the absolute optimum

setup for maximum legal signal strength but they are closer
than some think! For example, some people object to the 470
ohm resistor across the output coil in the first circuit on the
grounds that it "wastes" power that could have been
transmitted. Although another type of matching might give a
better output, virtually all of the power is going to be wasted,
anyway. A "matching network" that can provide the required
loading when connected to a legal antenna without any
resistors and gives a good quality signal must be lossy. You
just can't see the resistor. I can state this with confidence due
to the tiny "radiation resistance" and capacitance that a legal
antenna has. If you somehow manage to efficiently match to
these values, the Q of the resulting network will be tens of
thousands and the resulting bandwidth would be a few tens of
hertz. All the listener would hear is a low rumble!

In order to get the bandwidth high enough for music, the

natural Q of the short antenna must be lowered by a factor of
several hundred (at least) and there goes all that precious
power! Basically, you want the highest possible voltage on the
antenna but that value is limited by the maximum acceptable
Q (lowest tolerable bandwidth) and the allowed power. One
way or the other, you must include resistive losses. These
circuits (especially the first) "over do it" a bit to make them
more forgiving. (A poorly tuned matching network may not
exhibit the necessary loss causing the modulation to be
distorted.)

Having stated that, you can do somewhat better than the

above circuits. If the transmitter were 100% efficient, you
should be able to get antenna voltages above about 200 volts
RMS and still have good audio bandwidth. More realistically,
100 volts RMS is practical and the second circuit does fall a bit
short of this value. Playing with the transformer might yield

14
better output but it is hard to pick up significant range
improvements when only a few dB are a stake and the tuning
becomes more and more critical; major improvements just
aren't possible. I have read suggestions of different circuit
topologies and higher quality reactances to reduce circuit
losses but when you are done you will just have to add a
resistor to get the Q back down!

Now, if you don't care about the bandwidth, that's a different

matter! I wonder how far low speed data could be
transmitted...

I should point out that starting with a higher supply voltage will
directly give more antenna voltage for a fixed Q. And, it may
be practical to make a low-loss non-resonant transformer to
get a higher antenna voltage with a given Q. I would bump up
the power supply voltage first, the transformer would be a
challenge! A high voltage mosfet or even a vacuum tube would
make an interesting output device for a high voltage version.
The same circuit might work fine simply by connecting the
output inductor to the higher voltage, leaving the rest of the
circuit connected to 15 volts and reducing the current by
increasing the 100 ohm in the emitter of the bottom transistor.
Adjust the current down to keep the transmitter legal.

Richard de los Santos , Jack Dofelmire and Phong Nguyen

teamed up on their ECT final project entitled " Micro Power
Transmitting Station" using this AM transmitter. Nice looking
report!

Tunable Phono Oscillator

A "phono oscillator" is a simple, short-range AM band
transmitter that was typically used to send the signal from a
phonograph to a nearby radio, eliminating the need for an
amplifier and speaker. This version uses only one transistor

15
and can be tuned to any desired frequency near the top of the
AM band. Fidelity is surprisingly good considering the simplicity
and is suitable for transmitting "Golden Age of Radio" type
shows to a restored tube set.

Instead of a crystal, this transmitter uses an oscillator coil

intended for AM radios. These coils usually have a red adjustor
and the winding with the highest resistance is used, leaving
the other pins unconnected. The 33 pF places the optimum
frequency near the top of the dial but it may be increased to
68 pF for operation near 1 MHz. The input and collector chokes
are millihenry values (not microhenry). The transistor isn't
critical and just about any NPN small-signal type will work fine.
The collector output may be used to drive other circuits (like
the differential amplifier in the previous circuit in place of the
crystal oscillator) but it may be desirable to add a few
thousand ohms in series to prevent excessive loading of the
oscillator (see circuit below). The Ant. output has a loading coil

16
in series to achieve a much higher voltage on a few meter
length of wire but you won't get much range from this circuit.

My circuit board was designed to plug into those white

prototyping boards. If you count parts and find one is missing,
it's on the back due to a slight layout boo-boo.

Tune up is easiest with an oscilloscope. First tune the oscillator

to the desired frequency with no modulation. Then apply a 1
kHz, 100 mV p-p sinewave to the input and adjust the 500 ohm
potentiometer for the most symmetrical waveform. Looking at
the Ant terminal with a few feet of wire may give the best view

17
of the actual waveform since the audio signal can make the RF
look a bit distorted at the collector even though it isn't. Simply
placing a scope probe near the antenna is usually sufficient.
Play with the input level and 500 ohm pot to get the best
waveform but keep the input a little short of 100% modulation,
perhaps 90%. The1N914 (or similar diode), 0.1uF capacitor and
1 megohm resistor allow a digital voltmeter to monitor the
oscillator level. The best operating point will produce a meter
reading near 0.7 volts with no modulation. These parts may be
left out if the circuit is to be adjusted with a 'scope.

A simple oscillator to replace the crystal oscillator in the first

project above may be constructed as follows:

The circuit will operate on 15 volts without modification but the

470 ohm emitter resistor may be increased to save a little
power.

18
 Unfair Radio Transmitter

It may come to pass that AM radio stations begin to disappear,

switch to a music format, or, gag, switch to local programming.
The process will probably be slow and those interesting talk
programs will be available for some time over the Internet. For
those caught in such a state of Limbo, the Unfair Radio
Transmitter comes to the rescue. The Unfair Radio Transmitter
is a personal radio station that broadcasts an Internet audio
feed throughout the house so that those AM radios scattered
about remain useful.

Those who have experimented with electric field transmitters

like those above have probably discovered that they only work
well when the antenna is free and clear and there is little solid
matter between the transmitter and receiver. The electric field
is easily "shorted out" by even slightly conductive materials. In
contrast, this loop transmitter generates a magnetic field that
can cut right through the thickest walls. I live in an old rock
home filled with concrete and reinforcing wire mesh and this
transmitter's signal can be easily picked up in the bomb shelter

19
at the opposite end of the long house. I'm sure it can be
received in the cave next door!

The circuit is nearly identical to the ones above with the

exception of the loop components and a few of the bias
resistors. The dimensions of the loop were chosen to adhere to
the most strict interpretation of the FCC rules regarding
antenna length. The simple square design yields an inductance
of 2 uH when constructed with 1/2" copper tubing. That
inductance determines all the rest of the antenna values. First,

20
the series combination of the two capacitors (shown as 8000
pF) must resonate with the inductance at the selected
frequency. The Q must be kept down to about 225 for
adequate bandwidth for typical talk radio programming (even
less for music). That Q value means that the losses in the
antenna circuit must be near 4,000 ohms, shunt. The 10 k
resistor shown is representative of a value that might be added
to bring the Q down to the right value. Don't worry about the
"loss" of signal power; it is unavoidable if the proper bandwidth
is to be maintained. Tuning and selection of loading
components will be discussed below. It turns out that the
maximum voltage swing that can be achieved at this Q with a
class-A driver is about 70 volts p-p at full modulation, so a
capacitive tap is used to cut that voltage in half, just right for a
24 volt power supply. 35 volts p-p swing on the collector will
bring the collector voltage down to about 6.5 volts, leaving
about 4.5 volts from collector to emitter.

The 180 ohm emitter resistor sets the current in the output
stage to about 10 mA or 5 mA per side. The stead-state
voltage across the output transistor is about 20 volts so the
power in that last transistor is 20 V x 5 mA = 100 mW, the
legal limit. You can't simply lower the 180 ohm resistor for
more power, the voltage swing will be too big.

One nice thing about this circuit is that it operates at a fairly

low impedance. A 10X scope probe may be directly connected
to the output collector to directly observe the voltage:

21
This image was taken with an oscilloscope probe connected
directly to the collector of the output transistor. The 22 uH
choke that provides bias to the collector of the output
transistor may be a higher value, typically 100 uH; I just
happened to have a large 22 uH at hand and the value isn't
critical. The low 22 uH value just reduces the effective
capacitance of the capacitors slightly, perhaps 5 %. The two
series 8000 pF capacitors are selected to resonate the loop at
the desired frequency. Tack in candidate values then add a
trimmer capacitor across the coil to see how you did. You can
also leave a trimmer capacitor in the circuit, if desired. I
discovered that my capacitors were slightly too big so I simply
paralleled a small 100 uH molded choke to lower the
inductance a bit. It turned out that this inductor had just the
right amount of loss to eliminate the need for a shunt resistor.
But, earlier, I used a trimmer capacitor and about 10 k to get

22
the proper Q. If the loop is not sufficiently loaded, the sinewave
will begin to flatten on the top and bottom at full modulation,
at about 70 volts p-p. You can increase the Q (by reducing the
loading of the coil) to get more swing but your signal will sound
muffled. Just add a resistor across the coil until the waveform
looks like the one above. The voltage above suggests that the
circulating current in the coil is over 3 amps!

I found it handy to have a trimmer capacitor across the coil for

peaking with my field strength meter when prototyping. Once
everything is working perfectly, the capacitor may be replaced
with a fixed value. I found that sticking my hand in the loop
didn't detune the circuit much so a trimmer capacitor may be
mounted right across the coil and tweaked by hand at the very
end. The scope probe does detune the loop slightly but a light
adjustment of the trimmer brings the amplitude back to the
peak.

23
The circuit is mounted in a standard plastic conduit box but the
interface to the loop was more difficult than might be obvious.
The copper tubing doesn't really fit inside the PVC tubing. I
heated the PVC until it was soft, then shoved it on the precise
distance I needed. I then had to sand down the end of the PVC
pipe so that it would fit into the box again. The end result is a
very snug fit, but it would be a lot easier to just use all PVC
tubing and run some heavy wire or braid through the inside.
Perhaps a piece of RG58 inside the PVC would be about right.
The smaller diameter of the conductor will increase the

24
inductance a bit so expect lower value capacitors for
resonance.

In order to solder to the ends of the pipe, I cut little tabs with
an high-speed abrasive wheel. They're bent toward the bottom
of the photo. Notice the 100 uH choke that tuned my final

25
design perfectly. That was replaced by a resistor in parallel
with a 365 pF trimmer capacitor at one time. The two
capacitors are made by paralleling two old-style mica
capacitors for a value near 9000 pF. Stay with high-quality
capacitors like mica, glass, porcelain, Teflon, and NPO ceramic.
Ordinary ceramic is a bit lossy but those losses might be just
about what you need to set the Q. If your capacitors are too
lossy, you will be able to get a peak but the voltage will be low.
The large 22 uH could have been another 100 uH or higher. I
used some old 2N718A metal transistors instead of the 2N4401
but most small-signal transistors should work fine. Using two
pin connectors for the power, audio, and antenna make it a lot
easier to make changes to the board. Notice the piece of wire
soldered to the ground plane in the top-right corner of the
board. That is for connecting the scope and voltmeters. Not
shown is the 24 volt molded power supply and the cable that
runs to the computer speaker jack.

The audio gain is a bit high to accommodate wimpy sources

but the 22 ohm in the emitter of the bottom transistor is low
enough to start causing distortion. If your computer can drive
the circuit readily, increase that value until you are running the
volume on the laptop near full. There are freeware programs
for generating a sine wave that are great for looking at the
modulation level.

This was a fairly trouble-free project and it works great but let
me know if you have any problems. (charles@wenzel.com)

FYI: The name "Unfair Radio Transmitter" is a play on the

"Fairness Doctrine", intended as humor. Well, I think it's funny.

26