DTV Coverage Problems
I've written about distributed transmission systems (DTS, sometimes referred to as a single frequency network, SFN) before, but thanks to better software modeling and experience with real-world systems in Las Vegas and New York, I now have a better understanding of how these systems work. This month I'll focus on a DTS application I haven't covered before—relatively low-power synchronized transmitters—and discuss how they can improve reception in the middle of a station's primary coverage area. WHERE WILL MORE TRANSMITTERS HELP?
While it may seem strange to add a transmitter in the middle of an area where coverage already exists, there are situations where it can help improve coverage.
Indoor reception of DTV, especially at VHF frequencies, can be difficult. A prime example of this is in communities 30 miles or more from the transmitter. Outdoor reception should be fine in these areas. Indoor reception may work if the antenna is near a window, high enough, and facing the right direction, but it won't be easy. A second transmitter with an ERP of only 1 or 2 kW should provide plenty of signal over a radius of 1 to 2 miles for indoor reception. More power will cover a larger area.
Many stations planning to initiate mobile DTV service may be using a horizontally polarized antenna. Adding a vertically polarized signal from additional transmitters provides a way to improve reception on mobile antenna, likely vertical whip antennas, and on handheld devices where the receiver's antenna polarization will vary.
FCC low-power rules, Section 74.750(f), allow horizontal, vertical or circular polarization. This would apply to low-power on-channel repeaters licensed under Part 74. For a DTS licensed under Part 73, a waiver may be needed for vertical only polarization. Vertically polarized transmit antennas offer another advantage—reduced interference to viewers' outdoor antennas in weak signal areas.
Finally, it isn't unusual to find some small shadowed areas in the middle of otherwise strong signal areas. Look at the Google maps coverage map I created for WNBC
(zoom out if the overlay colors disappear).
You will notice there are many pockets of predicted weak signal due to obstruction by buildings and hills. Even in Los Angeles (see http://dtvcoverage.xmtr.com/knbc-dt.html) there are areas obstructed by terrain (parts of Hollywood and Culver City) and by buildings (Century City). Low-power transmitters could fill in coverage in many of these areas.FILLING IN THE HOLES
I'd like to see broadcasters be given the flexibility to use low-power (maximum 30 to 50 watts transmitter output) on-channel repeaters inside their coverage area with streamlined FCC processing, perhaps registration-only with notification to adjacent-channel stations, to fill in these coverage holes when and where needed.
|KBLR-Stratosphere Signal Difference and Delay map |
Firing up a fill-in transmitter inside the coverage area of the primary station takes careful planning, particularly if high power is required. The first step in the system design is to determine the desired signal strength in the area to be covered. If the goal is indoor reception, I'd recommend aiming for at least 88 dBµV/m.
On the other hand, if indoor or mobile reception is not critical and viewers are used to receiving weaker signals on outdoor antennas, a level of 68 dBµV/m or even less may be sufficient. The next step is finding the optimum transmitter location. If a strong signal is needed over a small area, putting an antenna on a tall building or hill in the middle of the area may work. This approach works well in congested urban areas to boost signal in the "canyons" between buildings.
When determining the best transmitter location to boost signal in an urban area, consider the "wrong side of the street, wrong side of the building" problem. Locating the transmitter at the edge of the area with the antenna aimed back towards the primary transmitter site may help in this situation.
If the weak signal is due to terrain shielding, the ideal location is likely to be near the top of the terrain creating the shadow. This allows the use of a narrower azimuth pattern, reduced transmitter power and perhaps an on-channel digital repeater.
The antenna's azimuth pattern should concentrate the signal over the area of interest. In congested areas, the elevation pattern is particularly important. If the signal close to the transmitter is too strong, it could interfere with adjacent channel stations. If it is located on a high-rise building, the antenna will need to be located near the edge of the building or on a tower on the roof and use enough beam tilt to provide a good signal to the streets below.
Once the location and transmit antenna pattern are known, the effective radiated power needed to provide the desired signal strength can be determined. If the transmitter power can be kept around 30 watts and the site is in a strong signal area, an echo-canceling digital on-channel repeater may be able to be used, eliminating the need to install a microwave or fiber to the site.
Last but not least, interference from the fill-in transmitter to the primary signal and to adjacent channel stations has to be calculated. For interference to other stations, the FCC requires using the root-sum-squared method to combine the signal level from all transmitters in a DTS licensed under Part 73.
I didn't see any provision, however, for combining signal strength from an on-channel digital translator licensed under Part 74 and a primary station licensed under Part 73. Self-interference is more difficult to calculate because it varies depending on the receiver equalizer range and type of receive antenna.
I've found it handy to map interference outside the equalizer range by looking at the difference in signal strength between the stations needed to allow reception. If a DTS is used, the timing can be adjusted to minimize interference but if the transmitter is located inside the main coverage area it won't eliminate it everywhere.
The antenna or power levels may require adjustment to reduce interference. If a digital on-channel repeater is used, the only handles for controlling interference are location, antenna and power level.
When the two signals are within a few microseconds of each other, interference is predicted when their amplitude differ by only 1 or 2 dB. Fortunately, it isn't easy to match levels that closely. In testing reception in an area where such interference was predicted from the WNJU transmitter at 4 Times Square and the high-power transmitter at West Orange, none was observed. Indeed, turning off the 4 Times Square transmitter caused reception to fail at low antenna heights. Based on the low probability that signals will be close enough in amplitude to cause interference and the ease of fixing it if they are—move the antenna a few inches; interference in this region can usually be ignored.
Interference from fill-in transmitters is more likely in more distant areas where the primary transmitter is shadowed and the fill-in signal is outside the receiver's equalizer range. Before getting too concerned about this interference, remember the primary signal was already too weak for indoor reception in the area and a directional antenna should provide the gain and rejection needed to eliminate any calculated interference.REAL WORLD EXAMPLES
How does it work? The DTS demonstration at the 2009 NAB Show I mentioned in June's RF Technology column showed multiple transmitters can improve coverage in urban areas without causing interference.
The main fill-in transmitter was located on top of the Stratosphere Las Vegas Hotel and operated at an ERP of 1,000 watts. It used a 200 watt Rohde and Schwarz transmitter and a simple antenna array composed of two vertically polarized Scala log-periodic antennas with a maximum at 189 degrees.
Offset mechanical beam tilt was used to maximize the signal over the Las Vegas Convention Center area while providing a strong signal down the Strip. The primary transmitter operated at 230 kW ERP from Black Mountain using horizontal polarization only.
From previous articles on DTS, you'll realize this system violated one of my rules for DTS—minimize signal overlap in the area between two transmitters. This system worked because the fill-in transmitter ERP was low and it was vertically polarized.
Timing difference increased in locations further south towards the main transmitter, but its much higher power allowed receivers to treat the weaker Stratosphere signal as noise. Interference was predicted in a weak signal shadowed location west of Black Mountain, but the signals were so weak anyone would require an outdoor antenna to receive KBLR. Due to the use of vertical polarization on the Stratosphere and the angle between the sites, interference was unlikely.
I created the "KBLR-Stratosphere Signal Difference and Delay" map using SPLAT GIMP and some PERL scripts. The difference between the two signals is shown in colors ranging from green (16 dB) to cyan or yellow when the difference is only 2 dB. A gray-scale overlay shows the timing difference between the two transmitters, with the bright areas and narrow bands in the middle showing delay under 10 microseconds and the broader bands representing 20-microsecond steps out to 70 microseconds. This figure is based on a 30-microsecond delay in the signal from the Stratosphere. The PERL scripts can be downloaded from xmtr.com/splat/perl/
How did it work? No interference was reported. LG supplied a handheld mobile DTV receiver that was used to check the signal throughout the Hilton and Convention Center. In multiple locations in the South Hall meeting area, reception was poor or non-existent until the Stratosphere transmitter was turned on. Without out the Stratosphere transmitter, conventional ATSC reception was impossible in the meeting rooms; the signal couldn't even be detected. With the Stratosphere transmitter, KBLR could be received on a USB tuner with a whip antenna, although it wasn't perfect.
The mobile DTV signal was rock solid on a Pixtree USB receiver. Along the strip, I heard the mobile DTV signal was being received reliably deep in hotels, even at ground level. E-mail me at firstname.lastname@example.org