Drawing Board

Sudden Ionospheric Disturbance Detector

Most people are familiar with FM radio broadcasts, which are at relatively high frequencies (88-108 Mhz) and can only be picked up in the local area around the transmitter. However if you switch to Long Wave  (1500m), these lower frequency (200Khz) transmissions can be picked up at far greater distances (100’s kms) from the transmitter e.g. BBC world service. This is possible because the radio waves bounce off the upper layers of the atmosphere known as the ionosphere. This ‘mirror’ in the sky allows radio signals to be picked up from transmitters that are way over the horizon and would normally be out of range.

As the frequency drops further to around 20Khz (VLF), the distance over which the signal will propagate increases further. As does the signal’s ability to penetrate deep into water, which makes it an ideal way to communicate with things that spend a lot of time far away and underwater, submarines! To this end  various Military organisations around  the world operate VLF transmitters to communicate with their submarine fleets.

Now, back to the Radio Astronomy. The fact that the reception of these VLF transmissions is dependant on the ionosphere or rather the level of ionisation within it, is very useful. This is because the level of ionisation is largely dependant on solar radiation striking the upper atmosphere, so variations in received signal strength can indicate variations in solar activity, specifically x-ray emissions associated with solar flares. Why are solar flares of interest? Because of the effects they have on power distribution and communications equipment. The potential problems are considered  to be so great that there are organisations such as the National Oceanic and Atmospheric Administration (NOAA) that operate satellites to monitor ‘Space Weather’ . The data is used to provide warnings of the impending arrival of spikes in solar radiation and allow power and telecommunications companies to prepare for possible outages.

Traditionally, radio signals are received using a receiver tuned in to the desired signal; so why not just listen to LW radio and monitor the signal strength? Well, although this is a possibility there are a few problems. Firstly, standard radio receivers tend to include an Automatic Gain Control (AGC) which levels out the exact variations in signal strength that we are looking for. Second, we would only be able to monitor one station per radio receiver and since not all stations transmit 24/7 it would be useful to monitor a number of stations simultaneously, without needing multiple receivers.

The usual weapon of choice for anyone wanting to look at a range of radio signals at the same time is a Spectrum Analyser. This device plots signal strength as a function of frequency and is widely used in the professional environment. At high frequencies these devices are expensive bits of kit, full of dedicated electronics. However, in the VLF range (approx 20Khz)  the function of a Spectrum Analyser can be carried out quite adequately by your PC, using mathematics in the form of the Fast Fourier Transform or FFT. This is a mathematical algorithm that converts time-domain signals into the frequency-domain and back again (inverse transform IFFT). The ‘Fast’ part refers to the fact that the algorithm has been optimised to run on a computer, subject to the restriction that the number of time-domain samples is a power of two e.g. 2,4,8,16,32,....512,1024,4096..... etc.

Above is an example of 1,2 and 4khz carrier signals at -20,-23 and -26dB amplitude levels. Left is the time-domain representation, this is  similar to what will come in from the antenna. Right is the same signal in the frequency-domain, the 3 peaks corresponding to each of the three signals present at the input. The example was setup using the free ‘Spectrum Lab’ software that we will use for this project.

OK, so we can use our PC as an FFT Spectrum Analyser, what about actually picking up the radio signals and getting them into the computer? This is where the Very Low Frequency of around 20Khz is of real benefit since it is within the audio range of a PC’s sound card . Most PC sound cards are quite capable of capturing signals up to 24Khz (Half the maximum sample rate of 48Khz), so the radio signal can be captured directly. This is really handy because it means that the only hardware needed is an antenna to pick up the signals (see below), the rest can be done using software. Even better, is that a very nice gentleman called Wolfgang Buescher has written a program called ‘Spectrum Lab’. This software has everything you need to capture the data from the sound card, perform the FFT and log the results in the form of a strip chart.

At VLF frequencies a loop type antenna is the most efficient way to pick up the signals in a limited space, a square loop being the easiest to construct. The cross-support can be made out of pretty much any non-conducting material onto which is wound about 125 turns of insulated copper wire for the loop. I used some PVC tube electrical trunking, joined with a piece of wooden curtain rail and some 0.2mm enamelled copper wire . The copper wire is supported by Nylon screws  in at right-angles at ends of the PVC tubes. The DC resistance of my loop was about 100ohms, the actual value is not that critical, it is more to do with the practicalities of winding the loop. Too thin and the wire will keep breaking, too thick and the whole thing gets rather chunky and heavy.

At its very simplest the loop could be connected directly across the sound card’s input. I have added a couple of back-to-back diodes (IN4007) and an old Neon lamp across the loop to protect the sound card in case of nearby lightening strikes, which could induce quite high voltages at the loop output. The diodes clamp the output to their forward voltage drop of ~0.7v. The Neon is there to take the sting out of any really big spike. The phono sockets were recovered from a dead video player, as were the diodes. The Neon was from an old multi-socket extension lead.

SID Event 09/06/2012

50cm

125 Turns Copper Wire

(0.2 to 0.5mm diameter, length 180m approx).

SID Loop Antenna

Protection Circuit

Phono / 3.5mm Jack Lead

To Sound Card Input

Next we need a computer capable of running the Spectrum Lab software. Something around 500Mhz-1Ghz with onboard sound and running XP is quite adequate and fairly old-hat by 2012 standards . My setup was built entirely from bits recovered from the local tip. Here in France they seem to have a far better attitude to recycling electrical goods I.e. Take what you like, but also take some responsibility for your own health and safety!  The Spectrum Lab software will run on a variety of OS’s so check Mr Buescher’s website for the latest version and compatibility.

Download and install the software on your PC and don’t be put off by the complexity of it, it’s quite a beast if you’re not into Digital Signal Processing (DSP) as a hobby. However, the great thing about this software is that ‘configurations’  for specific applications can be saved and numerous examples are provided. Using these to set up the basics makes setting up your own experiment much less daunting. I have included a ‘configuration’ file for my setup in the download (see end of page). Also included are the details about setting up the ‘watch list’ for transmitters you are able to pick up in your area. It isn’t difficult but would interrupt the flow a bit if I included it here.

If all goes well you should end up with a screen that looks something like the one below. The smooth humps are the different transmitters  each at their own frequency (remember the FFT example). The reason they are not sharp spikes is explained in the download  section.

If you then go into the View/Windows menu and select the Watch List & Plot Window, you will get the Watch List that details which frequencies to monitor. Selecting the Plotter will bring up a strip chart like the one below. This particular chart shows what you will be looking for, a SID event! The overlaid Geostationary Operational Environmental Satellite (GOES) x-ray data from NOAA shows good correlation between the x-ray spike and the brief (30 mins) boost in received VLF signal strength. Notice the very characteristic shape of the graph. A rapid rise, followed by a gradual decay over about half an hour. There are other possible graph profiles that could indicate a SID event and many other variations due transmitter powers, day/night and noise etc. These are explained in more detail in the download. Note the NOAA data gives the x-ray spike at 16:50 while the VLF  data is at 18:50. This is just the 2hour time difference between Universal Time Coordinated (UTC) used for NOAA data and French Summer Time, which is what the PC clock is set to. Click on individual graphs to see the original full size plots.

If you Google ‘SID Monitoring’  you will find that there are lots of other sites and useful information. So, although you won’t be the first to do the experiment, the data will be unique to your location. It’s also pretty cool to be able to detect a solar flare from a star 93 million miles  away using some old junk, some loops of wire and your PC, the sort of thing Dr Who would be proud of !

Drawing Board  |  Top of page

The ‘configuration’ file for Spectrum Lab together with other graphs and explanations can be downloaded here.

This is the link to Mr Buescher’s Spectrum Lab site.

This is the link to the NOAA GOES data.

My Station

European VLF Stations, below 24Khz

If  you’ve ever fancied  dabbling in Radio Astronomy, this is definitely one of the easier projects to get you started. Basically it involves monitoring Very Low Frequency (VLF) radio transmissions intended for submarines, using nothing more than your PC and some loops of wire. The project comes from the excellent book “The Radio Sky and how to observe it” by Jeff Lashley.

This web page and the associated download documents  my take on this simple yet intriguing project.