S Band Signal Source
Based on a design by Bert, VE2ZAZ as posted on amsat-bb. See his version at http://www3.sympatico.ca/b.zauhar/
Jim White, WD0E. 1/6/2002
This signal source was built with a TTL clock oscillator, MX045T-40.000, DigiKey
part number CTX215-ND. The rest of the parts were pretty much out of my junk box.

It's built in a Bud Minibox, about 2" x 1.5 " x 3". The switch turns it on/off.
See more on the antenna below.

With the cover removed. This is the first try. I found the oscillator drifted quite a bit
with temperature. Just put a warm finger on it or a bit of warm breath and it would
move out of the pass band of the receiver. As long as the temperature stays stable
so does the 60th harmonic of the 40 MHz fundamental (2400 MHz).

Closer look at the inside before the change to stabilize the frequency.
From this particular device the signal is on 2400.030 at room temperature and
moves about 5 kHz either way with a 10 to 15 degree C change. But when the
temperature is stable it moves very little.

To keep it from drifting quite so much a chunk of aluminum was added with
a brass strip to hold it down and some heat sink compound under the oscillator
and the brass strip. The strip is soldered to the oscillator case.

The top has just the SMA with a simple little ground plane for 2400 stuck in it.

The little antenna is a real hack, made from a chuck of semi-rigid hard-line with some
brass rod soldered on. Just for the fun of it I put it on the HP 8753 to see how well it
matched. SWR 1.16 to 1 at 2.4 Ghz, and 49.97 ohm impedance. Not bad for a 20
minute hack. Initially I cut the center conductor radiator and brass rods to 1.17 inches
based on calculation. It was a bit high in frequency so a small lump of solder was added
to the tip, which brought it pretty much right to 2.4 Ghz as you can see in the SWR plot
above.
Point being, it doesn't have to be pretty and you don't need at $25k Network Analyzer
to get close enough to make it work for this purpose. The original cut would have been
fine.