How to Start a Very-Low-Power FM Station

How to Start a Very-Low-Power FM Station 
to Serve your Immediate Neighborhood

based on information published in the Alternative Radio Handbook ($8 from
R. K. Harrison, Box 547014, Orlando FL 32854) and in the forthcoming
Free Radio Handbook, Volume 2.

For starters, a little knowledge of electronics wouldn't hurt.  You'll
also need audio sources such as microphones, tape decks, CD players etc., 
and a low-cost audio mixer, such as Radio Shack #32-1100.

Next, obtain the stereo transmitter kit (catalog # FM-10) from Ramsey
Electronics, 793 Canning Parkway, Victor NY 14564.  Their phone number is
716-924-4560.  As of late 1990, the price of the kit was $29.95 plus $2
shipping and handling.

Similar kits are advertised by DC Electronics, P O Box 3203, Scottsdale AZ
85271, and WJD Electronics, RD#2 Box 80A, Middleburg PA 17842.

The Ramsey FM-10 is a straight-forward stereo FM transmitter kit that comes
with a very clear set of instructions and all necessary parts (except for
an enclosure and a 9V battery).  It is designed around a BA-1404 integrated
circuit.  When assembled according to instructions, it complies with FCC
Part 15 rules which allow unlicensed, low-power broadcasting on a non-
interfering basis.  The rules are explained very clearly in the manual.
Realistic expectations for this kit are discussed later in this file.

Several experimenters have assembled FM-10's and have come up with the
following modifications to improve its performance.

(1) Increase the value of R1 and R4 to 10K if you experience a problem
with over-modulation.  A value of 10K provides a better match to most
mixers, CD players, and other household-type audio sources.

(2) Resistors R3 and R6 govern the pre-emphasis curve.  It is suggested that
replacing them with 75K resistors (available from such sources as Mouser
Electronics) would more closely match the pre-emphasis curve used in the
USA, thus providing even better sound quality.  68K resistors (avaiable
from Radio Shack) would suffice if 75K resistors are not available to you.

(3) Frequency drift is often experienced with the FM-10 and is partly
caused by the low quality of cermaic disc capacitor C16.  Replace C16 with
a silver-mica capacitor of the same value, or a temperature-compensated
disc capacitor rated from N150 to N750.

(4) Any change in the power supply will also cause frequency drift.  Use
of a filtered and regulated power supply is recommended.  DC Electronics
sells a reasonably-priced variable-voltage power supply kit.  A 12-Volt
supply will work okay and will produce a stronger output signal than a
9-Volt battery.

(5) Insert an RF choke in the positive power supply lead.  Any value from
3.3 microhenry to 1 millihenry can be used.

(6) A home-made dipole antenna will provide better results than the built-in
telescoping whip antenna.  If you won't be using the whip, remove C21 (the
capacitor which feeds RF to the whip).

The output stage of the FM-10 matches the impedance of either 50-ohm or
75-ohm coaxial cable, which is why it works well with a half-wave dipole.
To use this kit with a commercially manufactured outdoor TV/FM antenna,
it might be necessary to construct a 4-to-1 matching transformer, as
described in the ARRL Antenna Handbook or the Alternative Radio Handbook.

For best results your transmitting antenna should be outdoors and should
be at least as high as the rooftops in your neighborhood.

(7) It is often difficult to get the stereo subcarrier to work properly.
One way to deal with this is to replace C7 with a 100 pF capacitor, and
replace C8 with a 6 to 50 pF trimmer cap (Radio Shack # 272-1340).  The
RS trimmer won't fit the holes in the PC board; cut the leads off a spare
resistor and solder them onto the legs of the trimmer to mount it on the
component side of the PC board.

(8) The power output can be increased by reducing the resistance of R9 to
180 or even 150 ohms.  Two warnings are in order if you make this modifi-
cation: a) it might violate the power limit established by FCC Part 15
rules, depending on what kind of antenna you're using, and b) feel 
transistor Q1 occasionally to make sure it doesn't overheat; if it gets 
scorchingly hot it will burn out (croak).  Experience indicates that it 
is working near maximum efficiency when it is just a few degrees warmer 
than room temperature.  This transisitor has a "maximum device dissipation"
of 600 mW, so don't expect to get more than half a Watt of RF out of it
(best case scenario).

(9) For the advanced experimenter: Try replacing R8 with a 1K trim-pot
and carefully adjust for the best balance between output power and sound
quality.  If you find that beefing up the power supply makes it harder to
get the stereo subcarrier to work, try increasing the resistance of R12
to 1 Megohm, 10 Megohms, or even remove R12 altogether.  If you have a VHF
field strength meter and lots of patience, you can experiment with using
different values for C13; see what capacitance provides the best match to
your particular antenna.

realistic expectations:

The sound quality of the FM-10 is great, considering that it only costs
$30, and overall the kit is a much better value than the transmitter kits
currently available from Panaxis.
 
As mentioned in the manual, the signal from an unmodified FM-10 (or any
FM transmitter that complies with FCC Part 15 rules) can be heard with
a good receiver at a 1200-foot radius, assuming there are no obstructions
between transmitting and receiving antennas. High quality stereo reception
with a reasonable signal-to-noise ratio will be limited to a smaller
area, depending on the quality of the receiver and its antenna.  

Although it's easy to increase the range of the FM-10, for example by
connecting it to a directional "Yagi" antenna that provides some "gain",
such a modification can easily violate the FCC's rules.  If your broadcasts
cause any kind of interference, or attract attention from the wrong people, 
it is likely that you will eventually be inspected by FCC agents.  Rule-
breakers are usually given a fine ranging from $750 to $1000 on their 
first offense.

acknowledgements:

Thanks to Rick Harrison, Rob Peebles, Dr. Laszlo Xasczkanuski, John
Arthur, and Al Gorhythm for contributing to this research effort.
                                                                          
23 Feb 1991