Listening in on Very Low Frequencies

 Since I wrote this I've bought an SDR which is better than the soundcard method described below.

click to read more

 Now that I've almost finished refurbishing my R206 I remembered my low frequency adaptor, now long gone, cannibalised for an HF transmitter. In those days the bands were nice and quiet without deaded "digital revolution", switching power supplies, broadband and mains-bourne networking interference, and several very long wave transmissions could be heard. Wondering what those low bands sounded like today I thought about building a converter but when I looked on the Internet for ideas I saw that many people use their computer's soundcard as a VLF receiver.

Intrigued, I looked for the requisite software and initially found some which after installing left my brain numb. Not to be defeated so easily, I tried a package called "SAQ Panoramic VLF Receiver" and installed that. It's ever so much more user friendly and after a few minutes I could hear some transmissions. My aerial is a long wire which is threaded through various trees and connects to a phono plug at the end of a lead plugged into the computers microphone jack socket (in doing this I'm aware of the potential damage to the sound chips from high levels of static).

The picture below shows a green trace for the radio spectrum from zero to 96KHz and, inset on the left, the audio spectrum in blue. This latter extends from about 0.2 to 1.2 on the AF scale and represents the CW bandwidth. Switching to SSB and AM stretches this out. The peak on the left of the audio is the bleep from the timing signal selected by the cursor on 60KHz. There isn't a tuning knob alas because tuning is by cursor keys or mouse.

The hardware used for the VLF receiver is the Realtek sound chip on my Asus motherboard. Although this is intended for audio, the fact that it's a digital processing circuit, making use of a sampling technique, means that one can use it for processing RF signals. The first point to consider is the sampling technique. A typical sound card or built-in sound handling hardware will now have a selectable sampling rate. Mathematical theory predicts that, given a fair wind, one can reconstruct a complex waveform given a sampling rate of just double the maximum frequency in the waveform. You can see below that the frequency coverage provided by my sound chips is 96KHz because the sample rate I selected was 192KHZ. In practice I guess there's a compromise between fidelity (the reproduction of precisely what is fed in) and the errors met in working close to the sampling limit, for example in reducing processing time, but listening to the results shows that my sound chips seem to make excellent RF chips. Of course, all credit for this excellent VLF receiver must go to the original software designer who I understand is Johan Bodin, SM6LKM.

 

 The strongest signal is that at 16KHz but, tuning upwards in frequency, I found the 60KHz signal used for radio clocks. With all the hash, burbling noises and buzz saws you must listen out for clean, steady carriers carrying identifiable modulation or having regular pulsing to identify proper radio transmissions although some of the weird noises may well be emanating from aerials because the ELF and VLF bands are used for communicating with submariners and providing direction-finding information.

Results can be very variable and, to hear signals most cleanly, I set my microphone input to maximum volume and used the maximum boost of +30dB then selected the best sampling rate of 192,000Hz (Studio Quality) which lets me tune to about 96KHz. At this setting and with the software AF gain at -6dB, selecting L (I'm using the left channel for my aerial), the amplitude of the transmission is indicated as -40dB. The audio level baseline is -80dB with a 600Hz tone rising to about -50dB. The mode setting is CW, the tuning cursor carrier reads 60,390Hz and the VFO 59,640 Hz. There are loads of signals across the band but I guess many are just squeals and groans from local equipment, lighting, and other stuff. I'll investigate later maybe... below is a list of what I could hear. I wonder if there are any low frequency crystals in my computer, radio clocks, telephones, printer etc etc ? For example most digital watches and clocks use 32.768KHz.

Don't pay too much attention to the frequency I've entered as my tuning was only approximate. To identify some signals I might connect a small pickup coil to the Mic input and investigate some items around here...

Later, around midnight rather than daylight when propagation is different I checked and found at least some of the wobbly carriers may have been teletype and a weak carrier about 64.4KHz had 1 second pulses present.

 FREQUENCY KHz

 SIGNAL

 SOURCE

 95.5
 S3 rapidly fading in and out  

 80.6
 S2 carrier  

 79
 S7 buzz  

 77.2
 S5 1 second pulses  DCF77 Germany

 69.4
 S2 carrier  

 66.1
 S8 buzz  

 64
 S2 carrier  1 sec pulses

 62.6
 S8 buzz  

 60.1
 S7 1 second pulses  MSF UK

 57.1
 S2 carrier  

 52.2
 S2 carrier  pos TTY

 37.1
 S3 wobbly carrier  

 31.8
 S5 buzz  

 24.1
 S4 buzz  

 22.5
 S3 wobbly carrier  pos TTY

 21.3
 S3 wobbly carrier  pos TTY

 20
 S2 carrier  

 16.4
 S9 buzz  

 Return to Reception