Wednesday 25 September, 2013, 11:35 - Radio Randomness, Spectrum Management, Equipment ReviewsIn the past, Wireless Waffle has discussed various things that cause radio interference but which are not supposed to including, for example Power Line Telecommunications devices. This time around it's the turn of a Class T audio amplifier to come under the spotlight.
Posted by Administrator
Posted by Administrator
Class T amplifiers are really Class D amplifiers but are supposedly more efficient. Any clearer? No, probably not. The idea behind these types of audio amplifiers (noting that the Class D principal is also used in some radio transmitters too) is that instead of amplifying an analogue signal in an analogue way, such that the output voltage is just an amplified version of the input voltage, they switch the output voltage on and off at a frequency higher than the audio signal, and then use a filter on the output to smooth the square wave that they produce back into a nice analogue signal. This method is known as pulse width modulation.
This switching technique is exactly the same one that is used in the majority of modern power supplies (SMPS) and has the prime advantage that as the transistors that do the switching are either turned on or off, they are never in some intermediate state where they would have to act as a resistor and in doing so dissipate power and heat. So they are highly efficient and it is possible to generate audio with over 90% efficiency meaning that more of the power is converted to sound and less is lost as heat, which is, after all, a very admirable quality.
As with switch mode power supplies a good filter is critical in ensuring that none of the original square waves find their way to the output. Square waves are very good at producing harmonics and therefore are equally good at generating radio signals and, of course, radio interference. There have been many cases of switch mode power supplies causing such radio interference and their use in, for example, LED lighting, means that the number of possible sources of interference is ever increasing.
The main problem is that, in many cases, the device will work without the filter fitted - if (and only if) the device that it is powering is not too fussy about all those square waves (e.g. an LED) or has a method of smoothing them out itself (e.g. a loudspeaker). A loudspeaker is basically a large inductor, which is what the filters in switching amplifiers also comprise. Feeding the nasty square waves on the output of the switcher directly into a loudspeaker will not result in a noticeable loss in fidelity (assuming the switching frequency is well above the audible frequency range), nor any particular loss of efficiency. So why fit the filter? To stop radio interference, that's why.
So step up to the examination table, the Topping TP20-MK2 Class T Digital Mini Amplifier. One of these was recently purchased for the Wireless Waffle office, so that we could listen to the oidar through a bigger set of speakers. Being compact, and efficient, and coming in a shiny silver case, it ticked all the right boxes. But ouch, what noise from yonder shiny case breaks? As soon as the amplifier was turned on, reception of radio signals on just about any frequency was wiped out by noise. Even an FM tuner sat receiving a strong local transmission which was previously a perfectly quieting signal, was sent into oblivion by the amplifier. Obviously, the filter on the output of the Topping amplifier is completely inadequate for the purposes of curbing radio interference.
In cases such as these, there is little that can be done. Other than taking the device apart and replacing the filter components with better ones (an idea that is not as daft as it sounds), the solution is to junk the device and use a traditional linear amplifier instead. Which is what has been done. Bye bye trendy, offendy Topping, hello dusty, trusty Sony.
Wednesday 4 September, 2013, 09:46 - Spectrum ManagementThe new 700 MHz mobile band (703 - 748 paired with 758 - 803 MHz) is a hot topic amongst spectrum aficionados around the world. It raises a number of technical and political issues which are far from being fully resolved. On the political side, the main battle is between broadcasters (who currently occupy the band) and the mobile community who are keen to brush the broadcasters aside to clear the spectrum for more mobile broadband services. Broadcasters argue that they need more spectrum to cope with high definition and other developments whilst the mobile operators believe that the spectrum has greater value if used for extending broadband capacity and especially, given the good propagation characteristics at 700 MHz, the coverage, of their networks.
Posted by Administrator
Posted by Administrator
Whilst the political issues are like a pile of tinder dry kindling just waiting for a spark to set them alight, there are technical problems to address as well: The way in which the band is paired means that mobile handsets will be transmitting in the frequencies immediately adjacent to television broadcasts. Television broadcasts using channel 48 (in the Region 1 8 MHz raster) will use frequencies up to 694 MHz. Mobile handsets will use frequencies as low as 703 MHz. This leaves a 9 MHz gap between the two.
Whilst it might seem that the likelihood of a low power (50 milliWatt) transmission from a mobile phone causing interference to reception of a high power (100 kiloWatt) television service is small given the vast (2 million times) difference in transmitter power, the reality is that the mobile handset could be within only a few metres of the television receiver whereas the television transmitter might be tens of kilometers away. The strength of interference from the handset could therefore be orders of magnitude higher than the incoming television signal. Using free space path loss:
- A 50 mW (23 dBm) mobile transmitter 3 metres away produces a field strength of 91 dBuV/m
- A 100 kW (80 dBm) television transmitter 30 kilometers away produces a field strength of 68 dBuV/m
The 9 MHz gap between the mobile transmissions and the television reception is known as a guard-band and is there to give a chance for the television receiver to filter out the unwanted mobile transmissions. There is plenty of work going on to check that this is the case but it will depend on a number of factors that are outside the control of the mobile operators or television broadcasters such as:
- The quality of the television receiver. Those made to a low price may not perform as well in this regard as more expensive receivers.
- The quality of the receiver installation. An old antenna may receive less signal and poor coax will allow more mobile signal to leak into the receiver exacerbating the problems.
- The proximity of the mobile transmitter to the television antenna. In the case where television reception is through a 'rabbit ears' antenna on top of the TV, the distance between the antenna and the mobile could be far less than 3 metres, or even 1 metre.
- The distance of the receiver from the television transmitter. Those close to the transmitter are less likely to suffer interference but those in areas of fringe reception are at much greater risk.
- The use of television signal amplifiers. Such amplifiers can easily overload and stop working when presented with a strong nearby mobile signal.
The rules of spectrum use state that new users should implement their transmitters in such a way as to protect existing, incumbent users from interference and in that respect, the work to ensure that the new mobile services do not cause harmful interference to television services is totally appropriate. But there is another technical issue that needs to be considered, that of interference from the television transmitter into the mobile network.
Consider a mobile base station that is 1 kilometer away from the television transmitter, trying to receive a signal from a mobile handset that is 500 metres away. Let's run the free space path loss equations again:
- A 50 mW (23 dBm) mobile transmitter 500 metres away produces a field strength of 47 dBuV/m
- A 100 kW (80 dBm) television transmitter 1 kilometer away produces a field strength of 98 dBuV/m
Whilst it is possible to develop filters that can provide 50 dB, 60 dB or even more rejection of the television signals, they are costly. Of course not every site requires such an expensive filter: only those close to television transmitters on channel 48 will need them. But whilst television frequency use is not normally very dynamic, as re-planning of networks takes a lot of co-ordination, they do change channel from time-to-time and so knowing which site to fit the filters to cannot be done with complete certainty.
So if the political issues form tinder dry kindling just waiting to be lit, the problems of interference from mobile handsets into television receivers are a bucket of petrol poured on that kindling. It may therefore be that the problems of interference from television transmitters into the mobile network are the spark that gets the fire burning!
Saturday 31 August, 2013, 15:07 - Spectrum ManagementAccording to an article in the Daily Mail, a funeral taking place in Windsor, direcly under the flightpath for aircraft landing at London Heathrow airport was interrupted by the voice of a stewardess coming over the church's public address PA system saying, "fasten your seatbelts", and, "prepare the doors for landing".
Posted by Administrator
Posted by Administrator
The article claims that this occured because the 'church's wireless microphone and the plane's radio were on the same frequency'. Radio microphones in the UK operate in a variety of frequency bands. Most commonly they operated between 173.1 to 175 MHz and 863 to 865 MHz, both of which are available on a licence-exempt basis for such purposes. There are other frequencies available but these are subject to the need for a licence and are only normally used for professional programme making and special event (PMSE) uses. Most domestic and everyday business use (including local DJs, schools, hotels and churches) use the licence-exempt channels.
So how could the church's radio microphone be operating on the same frequency as an aircraft? If this is true, there are two possibilities:
- The systems on-board the aircraft used the same licence-exempt frequencies that the church's radio microphone used. Or
- The radio microphone system being used by the church was not on the officially sanctioned frequencies but was operating on frequencies reserved for aeronautical communications.
The CEPT has considered the use of short-range devices on board aircraft and it has concluded that, from the regulatory perspective, such use is allowed under the same conditions provided in the relevant Annex of Recommendation 70-03. For aviation safety aspects, the CEPT is not the right body to address this matter which remains the responsibility of aircraft manufacturers or aircraft owners who should consult with the relevant national or regional aviation bodies before the installation and use of such devices on board aircraft.
So it would not be illegal for an aircraft manufacturer to use licence-exempt wireless microphones on-board their planes. There is also increasing interest in using wireless technology to control the actual flight of the aircraft (the ailerons, flap, engines and so on). Planes using 'fly-by-wireless' technology have already been tested and there are now moves to try and find dedicated spectrum for them to operate in. This negates the need for the usual wiring that is required and saves weight which in turn saves fuel and cost.
Looking at the second possibility, there are radio microphones on the market which operate in the aeronautical communications (VHF) band from 108 to 137 MHz. These devices, commonly imported from China, are available on internet outlets at very low cost (check the frequency range on this microphone for example). They are, of course, completely illegal to use and not only are they subject to the kind of interference from aircraft that would have caused the problem in our church above, but they also have the potential to cause interference to air-to-ground aircraft communications which is stupid and potentially life-threatening. However despite such devices being available, this scenario is much less likely because the stewardess's voice would not be transmitted on the air-to-ground communication frequencies which are for pilot and air traffic controller communications only.
So it seems that it was probably the use of licence-excempt frequencies for short-range audio communications on-board the aircraft that was the culprit. It's worth considering that interference is usually bi-directional, especially where the systems are using similar powers and modulation. So it is just as likely that passengers on-board the aircraft could have been subject to the eulogy being given from the church as vice versa.
Which opens up a completely legal but very naughty set of ruses that anyone with suitable equipment could carry out if particularly bored on, say, a Sunday afternoon. Just stand near the end of the runway at an airport with a completely legal radio microphone operating, most likely, on the 863 to 865 MHz band and pick yourself a channel. As each aircraft comes overhead shout 'Brace, brace, brace' and if the aircraft's on-board wireless audio system happens to be on the same channel as your microphone the passengers are going to go ape. Of course, Wireless Waffle would never condone such activity, but if aircraft manufacturers are going to use licence-exempt frequencies, which are subject to no protection from interference, for on-board communications they are opening themselves up to all manner of prankery. If they choose to use bluetooth to actually control the aircraft itself, heavens only knows what might happen. Perhaps it would be best to stick to using wires or even optic fibre for controlling aircraft.
Tuesday 7 May, 2013, 08:14 - Spectrum ManagementThe Aussie auction of 700 MHz spectrum (one of the first of it's kind in the world) has, on the one hand, left the Australian government a few cents short of a brass razoo but at the same time left operators paying big bikkies for the spectrum. Wireless Waffle has previously discussed who the winner of the UK spectrum auction actually was. Whilst the UK auction (and most others around the world) are for 800 MHz spectrum, the Australian auction is relatively unique in that it is for 700 MHz spectrum. The auction has left a third of the available 700 MHz unsold, which is not a good place to be as: (a) the spectrum is not being used, which is clearly inefficient, and (b) it means it will have to be re-auctioned, which means a bit more hard yakka for the Australian regulator ACMA.
Posted by Administrator
Posted by Administrator
The fact that spectrum could remain unsold had been predicted by some, as a result of having set the reserve price too high. So just how does the price paid in the Australian auction compare to that paid elsewhere? The table below sets out the prices paid for 800 MHz spectrum in Germany, France, Italy and the UK and compares this to the price paid for 700 MHz spectrum in Australia. The comparison made is on a 'price per MHz per population' basis and for fun, GDP (PPP) has been factored in to see whether this makes any odds. Prices are all converted to Euro.
|Country||Band||Euro per MHz per pop||Euro per MHz per pop
|United Kingdom||800 MHz||0.51||0.52|
What does this tell us? Other than the fact that it demonstrates an ability of Wireless Waffle to do some simple maths, it shows that the price paid per MHz per person in Australia was almost 40% higher than the average of all these countries put together. Similarly, when scaled for GDP, the price paid is still 20% over the odds. The only more expensive spectrum was that sold in Italy and this is only because Italy's GDP per capita is somewhat smaller.
So why would operators down under be willing to shell out more moolah for spectrum than elsewhere? Perhaps the secret lies in the rapid population growth taking place. Just a few days ago, it was predicted that the Australian population had reached 23 million. But more importantly, the population is predicted to rise to 40 million in around 40 years time representing growth of around 1.5% per year. This is much faster than the population growth in other countries and means that in just 10 years time, the population of Australia will be 15% larger. If you do the price per person calculation now, the €1.06 becomes €0.92 and the €0.92 becomes €0.80. This puts the prices on a GDP adjusted basis just 5% ahead of average. In 15 years time the prices would be 5% below average. So over the life of the licence maybe these prices are not too high. Of course the population of other countries is growing too, and this would have to be factored in to any more detailed comparisons (such as a competent economist might make).
There are statistics to show that the number of bananas imported into the UK is closely correlated with the growth in population (as the above graph clearly demonstrates). It therefore follows, using Spock-strength logic, that there is a connection between bananas and the price differentials paid at auction for spectrum. Australia grows lots of bananas and it's population is rising, and lots was paid at auction. Q.E.D. as they used to say in ancient Rome.
What does this prove? Nothing at all, but having called the European Commission 'nuts' last month, the time was right to make another foodstuff related comment about a regulator somewhere or other. And so it seems (as so obviously can be deduced) ACMA is clearly bananas!