Further experiments on local QRM

 Below are two pictures showing the spectrum from zero to 5MHz. The first has Radio 4 on 198KHz tuned and the second has the receiver tuned to 600KHz. The displays are roughly centred on zero Hz with the LHS almost a mirror image of the RHS. The broadband noise is quite apparent with strong broadcasts shown as spikes sticking out of the noise. The second picture has different amplitude settings but has much the same noise pattern as the first.

Click here to listen to a wav file for the second picture



 Clearly listening to broadcast stations and anything below 7MHz is fraught with difficulty at my QTH. The answer maybe is a loop aerial mounted away from sources of noise. For long and medium wave reception the loop will need to be tuned but for 80m perhaps this is not too important. The design of the loop aerial in my previous page is of course entirely unsuitable for transmit, but this would anyway be done using my dipole aerials.

The following pictures show the next phase of the loop aerial experiments. I've fitted three coils viz. Long Wave, Medium Wave and 80m. Control signals and DC power are fed via a cable whilst RF is carried via 50 ohm coax. The small diecast control box is fitted at the centre of the loops. This carries the single transistor amplifier, two relays and some wiring. The loop is temporarily mounted close to the radio shack and the mast is supported by a Workmate (just outside my workshop door) allowing it to be rotated. The angle of the loop is adjustable and is set at roughly the setting for best nulling of noise.




 This picture shows the power supply and the varactor tuning potentiometer. The relays for selecting the appropriate coils are operated by switches which will be fitted later next to the pot. Click to see the first experiments.

 Below are two pictures depicting reception of Radio 4 on 198KHz. The first using a long wire that's picking up interference and the second using my new loop aerial tuned to 198KHz. You can see the signal strength differs by only a couple of dB despite the fact that the wire is around 250 feet long and reaches 30 feet in height and the loop only 3 feet across and 10 feet high. What isn't immediately apparent are unwanted products generated in the receiver front end. The second picture shows the half dozen or so broadcast stations operating on long waves and the first shows not only these but a mass of other signals. Interestingly, using the tunable loop, one can see these spurious signals rise and fall as the medium or short wave bands are tuned whilst the receiver is set to 198KHz. I suspect this is due to effects of instantaneous power supply changes within the SDR together with harmonics and spurious responses. As an example one of the signals to the left of the group of four broadcast stations around Radio 4 peaks by about 20dB when the loop is tuned to about 900KHz. Lots of the signals which appear to be in the long wave band fed by the long wire are really images and other mixer by-products.

Because the loop is still situated close to the house mains wiring and my CAT5 cables it picks up some interference (basically some of the hump (or mountain of noise!) you can see on the first picture between 250KHz and 1MHz) but when I move it 50 feet away this and the associated background noise level will drop sharply.

SDR fed with Long Wire 

SDR fed with tuned loop aerial 

 If you're interested in Long Wave reception, the largest spikes in the lower picture around the selected one are, from left to right France Inter 162KHz, Europe1 which broadcasts in French on 183KHz but is located in Germany, BBC Radio 4 198KHz, then a weaker station Radio Monte Carlo (Monaco) 216KHz, RTL (Luxembourg) 234KHz and RTE Radio (Ireland) 252KHz. Over on the left you can see some smaller spikes which are outside the tuning range of my loop aerial. These are ELF and VLF broadcasts used for a variety of purposes including communications with submerged submarines, time signals and frequency standards.

 The next phase of my experiments was to tidy up the loop aerial, extend the coax and control cables, and position the thing on a pole as far away from mains wiring and telephone lines as possible. The receiver under test was the SDR Play. I had to add something like 100 feet of cables to enable me to place the aerial in the centre of the garden.

This done I listened on long waves initially. I found that the weaker 216KHz French station could now be heard clearly although there was still a little buzzing background noise. I also found that I could operate with more gain because broad band interference was much lower. The receiver is very good but is critical in operation requiring just enough front-end attenuation to provide a good output without overloading. I think, at least for frequencies up to say 30MHz, an aerial tuner is very desirable. Without some sort of front end tuning you'll see a huge number of spurious signals including images and cross modulation products, interfering signals from computers, low energy lamps and loads of other stuff.

Using my tunable loop it's easy to see spurious signals. For example there's a strong AM broadcast station between the two French stations at the low side of Radio 4. It's about the same strength as its companions but as the loop is tuned the signal fails to peak as the loop is tuned. The two French stations rise and fall in amplitude as the loop is tuned but the spurious station peaks at a medium wave loop setting.

Next I looked at 80m. The band was in poor shape but using the SDR Console (V2) I was able to see an AM QSO at 3.615MHz. It was very noisy and readability was hit and miss but I could hear a Welsh ham telling a G4 that he was putting out about a watt from a 62 Set. Best readability was using synchronous AM. My aim is to use an AM transmitter feeding an 80m dipole and if that picks up too much noise I'll use the loop aerial for reception.

 The next day I continued my experiments. After listening to Radio 4 for a short time using two different software receive applications I realised that the rasping interference which sounded like an old fashioned video signal was indeed a video signal. Recently I'd bought two new network cameras and although I'd turned off one during tests I'd overlooked the other and left it turned on. I unplugged the power supplies for both and magically the rasping interference disappeared. Not quite all, but enough to allow me to hear Radio 4 without the background interference. Tuning to 80m I could now hear several sideband signals and medium waves suddenly came to life. The curtain of noise had been lifted and an array of stations were there, all in the clear.

Now that I can see a clear spectrum of signals it's apparent that the software receivers have a number of basic problems. These problems are a combination of software glitches and weaknesses in the design of the hardware. The former will no doubt be sorted out when improvements are made to the code, but hardware weaknesses are best sorted out by the user. For example, because the receiver is wide open to radio signals from zero to 2GHz a particularly strong signal might influence the reception of others. When one is listening to signals in the spectrum uo to 30MHz a low pass filter operating at 30MHz will reduce interfering VHF/UHF signals and vice versa. Ideally the use of a tunable front end is even better. If this includes measure of gain, even better.

 Today I connected up my discone aerial to my new SDR Play using the superb SDR Console (V3). I spotted a strange signal amidst lots of noise spikes. It was a white blob, a bit like a spiders nest so I altered the span to open it up. It had a semi-circular shape and comprised around 60 signals about 3.5KHz apart. Each was amplitude modulated by a fixed low frequency tone. Ignore the frequency in the box, the S-meter reading and the green line etc and as I moved this away to take a clearer picture.

The large number of spikes which spread right across the radio spectrum represent interference from electrical equipment using chopper power supplies.

What is the semi-circular signal? Is it a real broadcasted signal or something emanating from our washing machine which plays musical notes into a portable radio brought near to it.

The details are as follows.. starting at 74.146MHz and finishing at 74.357MHz there are about 60 discrete signals spaced at 3.5KHz. Each signal carries what appears to be a low frequency AM tone.

 Summing up my first experiences with a software defined receiver I should say they are an excellent proposition for the short-wave listener and a useful tool for the professional RF engineer. There are probably a dozen different types, some exceedingly sensitive but some that are only useful for monitoring only strong signals. Many are designed for use in the VHF bands and above but a few will cover down to a few KHz. Decent quality receivers use an analogue to digital converter whilst really cheap models use the computers sound card. As a spectrum analyser a receiver will really need a bandwidth of at least 8MHz. Of course a professional spectrum analyser may have a span of more than 1GHz so a span of only 8MHz feels very limiting.

After experimenting with an SDR it quickly becomes apparent that it can easily be overloaded. As the gain is increased you'll suddenly see a whole mass of spikes appear on the screen, however if you merely want to listen to a specific broadcast with best signal to noise ratio, then overloading may not be troublesome if it doesn't affect the signal to which you are listening.

A tunable RF stage is extremely useful because you can then readily determine the source of spikes. For example, like any superhet receiver you will see images and mixer byproducts. A tunable RF stage will let you figure out the true frequency of any spike. Of course, in a typical urban environment, you'll see loads of spikes that are within the range of frequencies in which you're interested that are emating from numerous household appliances.

 An interesting use for an SDR is checking ones car key, a task frequently presented to me by the proprietor of our local garage in Burley. Frequently a key will apparently stop working but as most transmit a signal in the UHF region, specifically 433.92MHz here in Europe you can check whether its operating. Setting the SDR to this frequency give or take 100KHz or so spread will enable you to see transmissions from keys. Surprisingly, some transmissions travel quite a long distance so in an urban area or even in a rural area like mine there are loads of signals popping up as car doors are locked and unlocked. To hear them you may need to increase the AM bandwidth setting to a custom 20KHz because, like aircraft they pop up for a very short time. Usually a key problem can be diagnosed to a faulty sub-miniature switch.

to be continued

Return to Reception