
John's homebrew pages
Microwave directional coupler - and an RF detector
More deviations from the main project.
I was concerned that, as I manage to generate a bit more power
(hopefully!) at microwave frequencies, a simple RF probe would not be
sufficient for power measurement, and the diode could be over-voltaged.
A directional coupler allows a small amound of the forward signal to be
sampled and measured. A very simple design is to use microstrip, as
shown in the photo. A BAT85 Schottky diode is used as the detector,
with 100pF and 10nF decoupling. The other end of the coupling loop is
terminated with two 100 ohm resistors in parallel.
Here it is in use, with the signal being dissipated in the dummy
load. A 100 microamp meter is used as the detector (100-0-100 microamps
here since that's what I had available). It seems to work well, and is
very directional indeed - very little signal is measured if it's
connected the other way round. I don't know what the coupling is with
any accuracy yet, but it will be fine for relative measurements for now.
Having proved its success, I will probably make one with forward and
reverse detectors that can be put in a box and used as a "through line"
power meter for forward and reverse measurements.
I also wanted something a bit more reliable than my hand held RF probe
for making measurements of power in test rigs. With that, since it
really needs to be mounted in a screened tube, the signal can vary a
bit depending on how the probe is set up. I thought that if I could put
the RF detector into a BNC plug, then I might get rather more
consistent measurements. Here are the components before assembly into a
BNC compression plug. The BAT85 diode fits right down near the probe;
two SMD 1nF capacitors are soldered to a brass washer, and to the other
end of the diode, to give the decoupling; and for good measure, the
output is taken through a 2mm ferrite bead. When I get round to
scanning some circuit diagrams I'll try to remember to add a sketch of
the arrangement here.
This new detector seems to be much more consistent than the RF
probe, though I'm not yet happy that I can be sure of power
measurements calculated from the detected DC voltage!