AmateurRadio.com

Yesterday I completed the construction of the crystal-controlled tone generator which will be used to modulate my lightwave transmitter during future clear-air / cloudbounce tests.




It was installed on the lightbox, right beside the original 556 CW beacon / tone generator.




The crystal-controlled oscillator uses a CD4060 IC as an oscillator-divider and produces a ~550Hz or a ~1098Hz squarewave from the 4.5MHz crystal.

4500KHz xtal divided by 8192 showing 549Hz output

The original 556 tone generator will be kept for aural CW and beacon modes as it provides a stable enough signal for this purpose but for the very narrow bandwidths that I plan to use when digging into the noise with Argo, I reasoned that the signal needed to be more stable.

As can be seen by comparing the two oscillators (crystal on the left and 556 on the right), the 556 has a lot of drift (although it looks like it might eventually stabilize) and, as well, produces several spurious signals ... probably robbing power from the main tone. The crystal-controlled signal is rock solid and doesn't appear to generate any parasitic signals in the process. The trace below the crystal signal is unrelated to the oscillator.




When I first wired the unit up, I found an unstable low frequency oscillation from the 4060 during key-up conditions, due no doubt, to the lengthy leads inside the box. This was cured by adding a pull-up resistor to the keying line as shown in the final schematic below.












Now it's on to building another fresnel-lens receiver box which will be needed for any field work here on the island.

It's always interesting to use one of the many 'audio viewing' software programs, such as 'Argo', 'Spectran' or Spectrum Lab, driven with receiver audio, to dig deep into a section of the spectrum using very narrow bandwidths. The narrower the bandwidth becomes, the greater the signal-to-noise ratio, increasing sensitivity in effect. The use of digital signal processing software can create the extremely narrow milliHertz  filtering needed to view signals buried deep in the noise...the deeper the search (20db, 30db or more into the noise), the longer the time needed to build up the visual display of signals that would be far too weak to detect aurally.

Recently, rather than using Argo to view a slice of spectrum I set up the Perseus waterfall display to have a deep-look at 1240kHz. This is one of the broadcast-band's 'local' channels and one that BCB DXers refer to as a 'graveyard channel'. Almost all stations on the various graveyard frequencies run a maximum of 1kW day and night. According to the Medium Wave List, there are presently 166 stations in North America operating on 1240kHz...one of the reasons that 'DXing the graveyards' is so interesting.

With this in mind, I recently took a mid-afternoon listen. Just one station was audible to my ear, likely one of the stations to my south in Washington state. Centering Perseus on 1240kHz, the waterfall was set to display a ~50Hz slice of the spectrum...1240kHz +/- 25Hz. The screen below shows the display, ranging from 1239.976kHz to 1240.024kHz. After letting the waterfall visual display slowly build up, it revealed 26 separate carriers. The remaining signals, although propagating to my location, were too weak to be detected aurally.

26 midday carriers visible on 1240kHz

Checking the MWList once again, it shows only 11 stations to be within expected daylight groundwave distance from here...Washington, Oregon and B.C. Evidently there is some mid-afternoon skywave involved or I am seeing some extended groundwave from Idaho, Montana and California. MW List shows a total of 27 possible candidates, assuming extended groundwave, so it would seem that most of them are making it here in daylight.


The screen above was made approximately one hour after sunset (looping E-W) and shows the same 25Hz slice centered on 1240kHz. There are ~70 carriers visible by now, with most of them fading. Some transmitters appear to be rock-solid while most exhibit a cyclical drift, no doubt the result of some form of crystal temperature stability attempt. I'm guessing that the rock-solid ones are using more modern synthesizers for frequency generation.

Listening on 1240 during this period reveals a boiling cauldron of audio, mostly unreadable until one station fades-up and becomes intelligible for a short period before fading to be replaced by another. Sitting on this frequency at the top-of-hour identification time can often catch a few idents with careful listening.


Even comparatively empty 540kHz reveals 22 different carriers, the only audible one being CBK in Watrous, Saskatchewan, the brighter trace at 540.002kHz. The one at the right looks as if it may have just come on the air for the evening. It's possible that many of these are low powered traffic information stations (TIS) running at 10 watts.