2
RF DIRECTIONAL COUPLER
A directional coupler is based on the principles of inductive (magnetic) coupling and capacitive coupling.
In the LARCAN quad directional coupler implementation as shown in Figure 3 (schematic equivalent) and Figure
5 (assembly), the RF to be sampled passes through a microstrip transmission line that is connected between the
transmitter output filter at J3 and the antenna system at J4. The magnetic field surrounding the hot conductor of
this transmission line induces a small RF current flow in other conductors situated parallel to it. One end of each
sampling conductor is terminated by a resistor to ground. Sometimes small capacitors are connected across
these resistors to provide a termination that remains resistive over the band. The other end of each sampling
conductor connects to an external load, usually a 50
Ω
input of something such as an RF detector for AGC, the
station demodulator, or an RF detector for VSWR sensing.
If the sampling system as described in the forgoing paragraph were dependent only on magnetic coupling and
absolutely no capacitance were present, the external loads would be driven with RF samples regardless of the
direction they came from. Omnidirectionality is not wanted; our objective is that the system should be directional,
that is, a signal coming from the transmitter should be seen by the "forward" ports, and a signal reflected back
from the antenna should be seen by the "reflected" ports, but at the same time as little as possible of the forward
signal from the transmitter should be seen on these reflected ports.
The desired directivity is achieved by the capacitance between the main line and each sampling line. The
presence of this capacitance changes the relative phase of the RF signal seen in the sampling line such that
the
capacitively coupled signal adds to the inductively coupled signal at the end of the line nearest the signal source,
and subtracts from it at the other end, thus the sample becomes directive.
This capacitance is trimmed by small "gimmick" capacitors designated L1 through L4. They are in reality short
pieces of Teflon sleeved magnet wire which, although they may possess a fraction of a nanohenry of inductance,
are mainly small capacitors which are factory adjusted by bending the wire to control the amount of coupling
capacitance between the transmission line and the sampling loop concerned. The position of the capacitor along
the loop does not seem to matter.
Terminations are provided at the subtractive ends of each of the four sampling lines.
In the enclosure shown in Figure 5, J3 and J4 are the filter and antenna ports respectively, and J1, J5 are
"forward" samples which are maximum amplitude for signals incident on J3; while J2, J6 are "reflected" samples
which are maximum amplitude for signals incident on J4.
Different coupling values are obtained from the spacing of conductors; the nearer the spacing, the greater the
coupling. Coupling is also greater according to frequency, and rises at a rate of about 6dB per octave. In the
boards shown in Figure 5, the J1 and J2 signals will be about 10dB greater amplitude (about 36dB below the
generator level at 70 MHz on low band or 200 MHz on high band) than the signals sampled from J5 and J6
(about -46dB). Generally for system purposes the reflected signal sample to the VSWR supervisory system
should be taken from the J2 connector because it has greater coupling and we need to measure a much smaller
signal in a detector having finite small-signal sensitivity. System forward signals can be taken from J1 for the
AGC detector, and J5 for the system monitoring demodulator.
A network analyzer and extremely accurate terminations are required for setting up the directional coupler. The
adjustments are made to the trimming capacitances L1 through L4, and the capacitors in parallel with resistors
R1 through R4. Our target is directivity of 30dB or better on each sampling port, and coupling (forward direction)
for J1 and J2 about 36dB down, J5 and J6 about 46dB down.
No user adjustments are possible or recommended. Very little can go wrong with the directional coupler other
than from the antenna being hit by lightning, and inspection is all that is recommended, nothing more.
PUB96-26 Rev 1 September 13, 2005
26-4
RF Output: BP Filter & Directional Coupler
