Digital Radio Mondaile

Digital Radio Mondaile is a new technology for broadcasting on medium wave (AM broadcast band) and short wave frequencies that looks interesting. It uses a form of modulation called Coded Orthogonal Frequency Division Multiplex which is claimed to be resilient to common forms of interference.

Click on the link above to read the details. The description of this modulation scheme sounds to me very much like the modulation scheme used by the old Telebit Worldblazer and Trailblazer modems except at RF rather than audio frequencies.

I wonder how this will work in practice with distant stations at night where you usually have several stations coming in at once and interfering with each other.

I’m not excited about the encoding scheme used for the audio, AAC compression followed by Spectral Band Replication.

I know a lot of people have good things to say about AAC, but I’ve compared to AAC to OGG and MP3, and of the three usually OGG is better even at substantially lower bit rates. There are rare instances where AAC does a better job, particularly things like picked acoustic guitar. For some reason OGG seems to suppress the fundamental relative to the harmonics leaving the instrument sounding hollow. But this only happens if there aren’t other simultaneous sounds. The beginning of the Beatles Here Comes the Sun is a good example of something OGG chokes on. On the whole though OGG does better than the other two even at much lower data rates.

Then the other part of that is pretty ugly too. Spectral Band Replication, that’s a buzz word for not transmitting the high frequencies and then trying to guess where they should have existed based upon the lower frequencies present. The idea is that most higher frequencies are harmonically related.

I have an audio processing program that is useful for trying to make really bad or mangled recordings more tolerable, things like scratch removal, hiss removal, and subtractive noise removal. It also has this spectral band replication capability; and if you take a really old recording where everything over 7 Khz or so is rolled off, it does make it sound “better”, but it doesn’t necessarily sound like it is supposed to.

In real music, while it is true that most high frequencies are harmonics of lower frequencies in the source, the relative intensity of the harmonics is not fixed, nor is the envelope of the harmonics necessarily the same (and more often than not it is not) as the fundamentals.

Think of a gong. You first hear the lower frequency fundamental tones prominently. Harmonics are present but as time progresses the fundamentals fall off at a higher rate than the harmonics and they seem to peak a second or two after the gong is struck. This is the sort of thing that spectral band replication is not good with.

I would have rather seen them go with OGG, and if they had they might have been able to transmit the high frequencies and still stay within the target bit rate.

Still, if it resurrects AM and makes it usable for the transmission of music again, then that’s a positive thing.

Another KNHC Article

Dan Thorn brought to my attention this article by Tim Shook regarding KNHC which details the equipment that was in use at the time and some of the interesting problems we had with it. There is also a black and white photograph of Larry Adams and Gene Arnold.

Tim neglects to mention in his on-air position at the station. I’d be lying if I said I wasn’t a bit envious of his golden voice that resonates speakers with deep smooth bass.

Time mentions a range of about 20 miles when KNHC was running 1500 watts. Prior to running 1500 watts, KNHC operated with 320 watts, and at that power on one drive to Spokane, I listed to it all the way across the past, lost it when I went down into the Canyon at Vantage, but once up on the other side received it again and continued to receive it about half-way to Spokane and that was with a portable Sony Earth Orbiter radio sitting in the front seat of the car.

Over the years KNHC has received numerous power increased, at one point to 30,000 watts. Around 2001 they moved the transmitter operations to Cougar Mountain and are now broadcasting with an effective radiated power of 8500 watts from that location. Even though the power is lower, the increased elevation makes up for it resulting in fewer dead spots.

Another data point worth mentioning is that Seattle Public Schools is considering selling the station believing it will bring up to eight million dollars. It would be a tragic loss if it were sold because it has launched the careers of many students. The F.C.C. 1st class license I received while attending Nathan Hale was the ticket to the first job I held for 17 years. The value of KNHC to the students is incalculable.

In all probability, if they did sell, they will not be able to get anything near the projected their estimated 8.5 million dollar price because the operating frequency of 89.5 m Hz is within the educational band of 88-92 MHz and I do not believe there are available frequencies in the commercial portion of the band.

AM Medium Wave Box Loop Antenna for DXing

This is a type of antenna known as a box loop that is very useful for AM broadcast DXing. It is a highly directional highly selective antenna. The directionality in particular has a very sharp strong deep null that is useful for rejecting interfering stations, and the selectivity minimizes receiver overload by strong local stations on other frequencies.

The antenna I built used 1×6 wood planks about 30 inches long, notched half-way through in the center with a notch as wide as the plank was thick to allow them to be fitted together in this “X” pattern. The rigidity of this structure is very important. Use a hard wood like walnut or oak, not a soft wood like pine. Make the notches fit is very tight and then glue them together with wood glue.

The variable capacitor mounted near the “X” is a type used in old tube radios, something on the order of 50pf-365pf. With these dimensions eight turns of #16 copper enameled wire was just right to allow the antenna to cover 530-1600 kHz. If you build it larger then fewer turns may be required, and conversely if you make it smaller, more turns will be required.

The wire used is a trade-off between “Q” (how selective the antenna is) and ability to actually wind the wire tight on the form. For example, #12 wire would result in an even more selective antenna but you’d break the form trying to wind it tight enough to be rigid. In my experience #16 is a good compromise.

The type used in most transistor radios with the plastic spacers provide neither adequate range nor the ability to easily fine tune, The best type is the type out of the old radios that has a built in planetary drive so that it takes three or four revolutions to open or close the capacitor completely.

The turns should be spaced about 1/2 inch apart. Make notches in the end of the planks to accommodate each turn so that the spacing remains fixed.

Connect one end of the loop to the “ground” (chassis) side of the capacitor. Connect the other side of the loop to the terminal on the capacitor that is the non-ground side.

On the first loop from the ground end of the winding, make a tap consisting of a wire leading back to the center of the “X”. Attach the ground shield of your coax to the ground side of the capacitor and the “tap” at the first loop to the center of the coax.

The idea of having the coax feed off a tap of just one turn is that it allows very little loading of the antenna and thus the “Q” remains high providing good selectivity and sensitivity at the resonant frequency.

A large wooden or plastic (something non-conductive) table is a good place to work with this antenna and your receiver. Connect the other end of the coax to your receivers antenna and ground connector. Having a physical ground on the receiver will improve the systems overall stability by making the receiver end of the arrangement insensitive to body capacitance and other stray effects.

Set your receiver to a weak station around the center of the dial and tune the variable capacitor on the antenna. You will encounter a point where the signal is enormously stronger and if you build the antenna to the specifications given here that point will be very sharp. The selectivity, if built as described here, will be sufficient that it will actually attenuate the sidebands of the station you are tuned to. This can be helpful in situations where you have adjacent channel interference and your receiver is not sufficiently selective.

As described, this should cover pretty much the entirety of the old AM band, it may not cover much above 1600 kHz because the old style capacitors had too great a minimum capacitance.

Most of the old broadcast radio capacitors had at least two sections, one that tuned the local oscillator and one that tuned the input to the mixer. On better receivers with one or more RF stages, there would be additional sections for each RF amplifier stage.

The section for the oscillator is smaller than the others. One thing you might try to extend the antenna to cover the 1600 – 1710 kHz section is to take one turn off the antenna which will raise the overall frequency range of the antenna, and then add a switch to cut in another capacitor section to cover the lower range.

By adding more turns and additional capacitance, this type of antenna can also be made to work efficiently down into the long wave portion of the radio spectrum. This is a good option for a long-wave antenna when you don’t have room for larger alternatives.

Back when I was using this antenna, I was in Seattle WA, and at the time there was nothing on either 1200 kHz or 1210 kHz locally. At night, the only station on 1200 kHz was WOAI from San Antonio TX. With just the built in ferrite rod antenna I would get maybe 1-2 on my S-meter, but with this antenna I could peg the S-meter with the same signal.

Being the somewhat insane type that I am I also tried a modification of this basic design where I took another tap at 1/2 turn from the ground end, and wired a one transistor amplifier as an oscillator using the ground, 1/2 turn tape, and 1 turn tap in a Hartley oscillator except the only capacitor was the variable across the entire coil. I put a potentiometer between the 1-turn tap and the base of the transistor to allow me to adjust the “gain” so that I could adjust it to be just on the verge of oscillation so instead of being an oscillator it became a regenerative amplifier.

I included a link to a site describing Hartley oscillators in case the reader is not familiar with them. The bottom turn of the coil with the half turn tap was used as the entire coil for the oscillator which wasn’t allowed to have quite enough gain to oscillate. I used a bipolar transistor with bias, but you might have better performance with an FET as shown on the site I linked to. Whichever method you use, you will want to insure that the transistor operates at very low power so that if it does break into an oscillation you do not damage your receivers front end.

This antenna didn’t need more gain, but the regenerative amplifier allowed me to narrow the bandwidth still more (to several hundred hertz if I got it right on the edge of oscillating) and I could adjust how narrow the bandwidth was by the adjusting the pot.

The reason I added this is that I was trying to receive some foreign stations that operated on frequencies other than the 10 kHz allotments used in the United States, things like 654 kHz, and the additional selectivity was desired. I was never successful though at receiving any of these, too much power line noise and other noise sources to receive very weak signals where I was.

Without the regenerative amplifier, I also tried using this antenna for transmission but because the impedance was so high across the full coil of antenna, the voltages across the capacitor exceeded the value required to arc it over at very low power levels so this is not useful for anything more than a few hundred milliwatts unless you got capacitors with a huge plate spacing.

DXing (long distance reception)

Radio and television DXing has been a long time hobby of mine.

My first interest was DXing AM radio. Initially, the stations I received were not real technically challenging given that at the time clear channel allocations existed, so there were frequencies such as 1200 Khz where WOAI in San Antonio, TX, was the only station at night.

I received a Sony Earth Orbiter receiver as a gift from a close family friend and that receiver had greater sensitivity, selectivity, and an external antenna terminals.

At first I tried a long-wire antenna but found that the receiver severely overloaded resulting in spurious signals all over the band. Living in Seattle, WA, there was a plethora of high power (5-50 KW) AM stations nearby.

Then I learned of a type of antenna called a box-loop. I’ve given instructions on how to build one like I had built. I highly recommend this type of antenna, the performance exceeds that of just about everything else I tried. The selectivity of the antenna eliminated the problem of overload by strong local stations, and the directivity allowed another way to reject unwanted signals.

At the time the environment was just too noisy to receive the really weak signals of interest and my reception was limited to stations in the US, Canada, and Mexico.

Another area of interest for me is television DXing. At one time I had a reasonable setup consisting of a Wineguard deep fringe yagi antenna, Wineguard mast mounted pre-amp, and twinax (shielded twinlead) downlead.

A yagi antenna in which the elements come out 90 degrees from the mast, most modern TV antennas have gone to the log periodic design which has the elements swept back.

For DXing I prefer Yagi antennas to log periodic. I found the directivity of the yagi to be superior to log periodic.

One thing I have subsequently read that others have done, but I did not think of, is to orient the antenna so the elements are vertical rather than horizontal. Some VHF stations have gone to circular polarization because it provides rejection of reflections which causes ghosting if the receiver is equipped with a similar circular polarized antenna. But for those still using horizontal polarization, mounting a Yagi with the elements vertical provides rejection of ground based signals. This was frequently an issue for me as E-skip happens most frequently on channels 2 and 3 and we had a powerful local station on channel 4 that tended to bleed onto channel 3 and to a lesser degree 2.

Most of the E-skip I received came from the mid-west region, only rarely from the south. The local station was to the south, but I was near the top of a ridge, the station was on Queen Anne hill, and it was 100 Kw line of sight so even with the Antenna aimed away the signal was considerable and if I was trying to receive something from the south then it was really bad. Multi-hop E-skip when it occurs often results in a change to the polarization, so a vertical polarized antenna might be helpful both in rejection of local stations and double-hop E skip reception.

I found some interesting resources on the Internet.