Technical Topics, September 1997
Digital SW Broadcasting
(Part 2)
This month we continue our look at some of the issues associated with making digital short-wave broadcasting a reality. I’ll address some frequently asked questions. One problem, during the transition to digital short-wave broadcasting, is how to keep analog listeners. The folks at Telefunken may have the answer.
Recently Telefunken installed four new 500 kW transmitters at Deutsche Telekom’s transmitter site at Nauen, Germany. As a part of the inaugural broadcast, Deutsche Telekom and Telefunken conducted a digital demonstration.
Kim Elliott of Communication’s World on the VOA recently interviewed two Telefunken engineers about the test. (You can still listen to the interview on the internet at http://www.wrn.org/voa.html if you have Real Audio on your computer. Check the May 24, 1997, program. The program also contains some sounds of a digital signal transmitted from Delano CA to Washington DC. This VOA test allowed evaluation of ionospheric distortion. The Jet Propulsion Laboratory helped with the VOA test.)
The Telefunken engineers called the test system the T2M. It stands for the (Deutsche) Telekom and Telefunken Multicast system. Here is the structure of the signal spectrum.
Let’s assume an AM transmitter rated for 500 kW. To provide an analog signal for existing receivers, engineers allocate most of the power to the analog signal. For these tests they reduced the analog signal power to 400 kW. That is only a reduction of about 1 dB. The listener would never be able to detect the difference even if the signal was right at the noise level. Of course we are talking program listeners here so the actual levels will be much higher than the noise floor. The digital information uses the 100 kW freed by this reduction.
Engineers offset the digital information from the analog carrier to minimize interference to the analog signal. Two 50 kW signals take advantage of the available 100 kW. One stream was centered 10 kHz above the center frequency; the other was 10 kHz below the analog carrier. Both signals can be modulated with the same information for monaural broadcasting. Alternatively, each can carry one channel of a stereo broadcast. The two digital channels can also send two separate programs. Only one of the digital signals needs to be transmitted if spectrum conservation is a primary concern. The spectrum is shown in Figure 1.
![[ Figure 1, The carrier and two sidebands ]](/journal/images/tech0997fig1.gif)
The digital signals are more efficient than analog. Broadcasters can reduce power. Error correction techniques developed for weak-signal space communications can actually detect and repair noise hits. Your CD player uses similar technology to correct errors. So the technology is available for adaptation to the unique problems of short-wave transmission.
How well does it work? The Telefunken engineers claimed that quality equal to today’s analog FM radio will certainly be possible. They even held out hope of CD quality audio.
How soon could a digital system be on the air? About one year after broadcasters adopt a standard according to the Telefunken Engineers. There is currently an international committee, headed by Dr. Bob Everett of the VOA, working to develop the standard. After defining the standard, the major stations will begin broadcasting using the new mode. Then manufacturers can begin cranking (apology to Baygen) out radios for a new and eager customer base.
What is the downside of this system? Clearly the frequency planning will be a nightmare. It is hard enough for broadcasters to find a clear 10 kHz wide channel. Finding three channels in a row will be really hard. International broadcasters will need wider spectrum allocations. The bands will get bigger. Broadcasters may have to choose an out-of-band frequency. Limited coverage analog radios may no longer be able to tune in the analog signal.
Will digital broadcasting eliminate whistles and interference? Yes, the digital signals have an enhanced immunity to interference. The top dog on the frequency will be the station that will get through. Digital signals will exhibit a capture effect similar to that which occurs on FM radio. Jammers will either dominate the frequency or not be heard at all.
Why must the channels be adjacent on the dial? Existing transmitter high power amplifiers and antenna tuning networks must pass these signals. The bandwidth of the tuned circuits is a function of the efficiency of the tuned circuits. Highly efficient, low-loss circuits have a narrower bandwidth than circuits that are lossy. When running power levels in the hundreds of kilowatts, a lossy coil quickly turns into a blob of molten copper. So the tuned circuits are necessarily efficient and narrow. The digital information must be sent on frequencies close to the analog signal. This allows the transmitter to serve simultaneously for both analog and digital services.
There is no guarantee that the final system will look like the T2M system described here. The folks at Thomcast have a system called Skywave 2000. A check of their web site revealed only a short reference to the Skywave 2000 system and no technical details.
In succeeding months I will talk about some concepts unique to digital transmission. We will look at sampling techniques, quantization noise, and bit error rate. Until next time, stay tuned.