At NAB2004, broadcast engineers were focusing on getting ready for the shutdown of analog TV. At past NAB conventions, the date was uncertain. Some engineers believed analog TV would never be shut down, and I don't think anyone thought the Dec. 31, 2006, date set by Congress and the FCC was realistic. The concensus at this year's NAB seemed to be analog TV broadcasting will end sometime in 2009 or 2010. This realization had broadcasters concerned about DTV reception and coverage issues, both before and after the transition. I'll look at some of the papers that focused on these issues this month. Most of them were not in the NAB2004 Broadcast Engineering Conference Proceedings and are based on notes I took during the sessions.
One major concern among broadcasters is how well current ATSC receivers perform in today's congested TV bands. Andy Bater reported on ATSC T3-S10's recommended practice for ATSC DTV tuners, which was sent to the full ATSC group in April. The recommended practice had not been posted on the ATSC Web site ( www.atsc.org ) when this column was written, but here are some highlights from Andy's paper.
Receivers should be able to handle signal inputs ranging from -83 dBm up to -8 dBm. The -83 dBm input is based on the minimum signal level needed for DTV reception. The -8 dBm maximum level was determined by looking at the expected signal level from a UHF DTV station operating at the maximum power the FCC allows-1,000 kW-using an antenna with an elevation gain of 30. The elevation gain is inversely proportional to the elevation beamwidth of the antenna and thus affects the amount of signal hitting the ground close to the tower.
Andy Bater noted that transmitter performance can have an impact on DTV reception. Many NTSC transmitters put out excessive spurious signals on adjacent channels, making receiver adjacent-channel rejection moot. Also, although there is a phase noise recommended practice for receivers, some DTV transmitters transmit signals with excessive phase noise, eating up much of the margin provided by the receiver. He emphasized the need for broadcasters to emit a clean signal.
When the original DTV planning factors were developed, it was assumed DTV tuners would utilize double conversion receivers. In practice, most are single conversion. Based on this, some of the UHF channel taboos eliminated for interference into DTV tuners may have to be reinstated.
William Meintel addressed this situation in his paper,.Computing Interference under Moderate and Strong Signal Conditions-Proposed Changes to the FCC TV Analysis Model, which detailed problems with the initial planning factors and inappropriate use of the FCC Bulletin OET-69 and Longley-Rice for analyzing such things as analog transmitter site moves. OET-69 never considered aggregate interference from multiple stations. The desired-to-undesired ratio used in OET-69 assumed weak signal conditions. The antenna elevation patterns that stations actually use, especially LPTV stations, do not match those in the planning model. As Meintel pointed out in a previous NAB paper, the Longley-Rice model overpredicts signal strength under many conditions.
He recommended developing an improved methodology for predicting coverage and interference for use in repacking DTV channels when analog ends. Some of the improvements needed include moderate and strong signal desired-to-undesired (D/U) ratios, taboo channel D/U ratios for DTV and consideration of the aggregate of interfering signals. When looking at interference D/U ratio, the placement of the transmitters should be considered, as the weak signal D/U ratios can be relaxed by 10 to 15 dB for collocated transmitters.
Meintel's paper was not published in the proceedings and he rushed through the recommended D/U ratio slides too fast for me to catch them all. For taboo channel interference offset from the desired channel by 2 to 15 channels, he recommended D/U ratios of -20 dB in strong signal environments and -40 dB in moderate signal environments. In a weak signal environment, the ratios vary depending on the offset.
Of course, the overprediction of signal level by Longley-Rice poses problems in determining what constitutes a strong, moderate or weak signal level. This problem could be addressed to some extent by including land-use clutter factors in the Longley-Rice calculations or development of a new propagation model.
Victor Tawil described the DTV repacking dilemma in his presentation at NAB. It now appears likely analog broadcasting will be shut down in 2009 or 2010. While this seems like a long time, a lot of work is required to move stations with out-of-core DTV channels back into the core, determine what channel stations with two in-core DTV channels want to keep after the transition and accommodate other channel changes stations may need to make to avoid interference and maintain coverage.
This is not a trivial problem. Tawil used the New York and Philadelphia markets as an example. These markets are close enough that channels cannot be shared between them without creating areas of interference that reduce coverage. There are 43 DTV licensees in the market and five UHF channels assigned for land-mobile use. After the transition, 49 channels will be available if low VHF is included. Tawil asked rhetorically, is there enough spectrum? He said "yes," but warned stations may not have the same service area. Can VHF be abandoned? Yes again, but this will also result in some service reduction.
Other issues have to be considered in repacking the channels, including the accuracy of the FCC database, coordinating DTV channels with Mexico and Canada, and the adequacy of the current interference model.
On May 13, MSTV sent the FCC a five-step process for repacking TV channels after analog broadcasting ends. For details on the plan, check my article in the May 18 issue of my weekly RF Report e-mail newsletter.
Wayne Bretl from Zenith and Victor Tawil presented an update on the performance of fifth-generation 8-VSB receiver chips in the paper, "Fifth Generation VSB Receiver Field Test Report." Some of these newer chips are able to take advantage of multiple echoes and provide reception, even when the uncorrected signal-to-noise ratio is less than 15.5 dB. Equalizer response has been extended and is now symmetrical, which should greatly improve reception from distributed transmission networks or on-channel boosters, where multiple transmitters may be received at one location, leading to "pre-echoes" that older 8-VSB receivers had trouble handling. Fifth-generation receivers now initialize the equalizer based on channel impulse noise response, reducing acquisition time. The newer chips also have improved capability for handling reflections with higher Doppler rates, making them more suitable for portable and mobile use.
Tawil also presented the field-testing results of the fifth-generation receivers in Washington D.C. The D.C. sites were the same as those used in the testing of the Linx receiver technology reported at NAB2003. The test procedure was similar to the Linx testing-a dipole antenna mounted on a tripod five feet above street level. The tests found that the fifth-generation receiver provided successful reception for 66 percent of the tests, intermittent reception for 21 percent of the tests and no reception in 14 percent of the tests. An error in transcribing the fractional percentages from the presentation or rounding to the nearest full percentage point resulted in the total of 101 percent.
See my July 9, 2003 RF Column for details on the Linx tests and procedures. In Washington D.C., the Linx prototype failed in 17.7 percent of the tests, provided intermittent reception in 22.3 percent of the tests and provided successful reception in 60 percent of the tests, compared with failure rates of 62.6 and 63.6 percent for each of two reference receivers. The link for the Linx test report in my June 2003 column is no longer working. It is now available here. Note that these tests were based on an older Linx prototype receiver and newer versions should be expected to provide better reception.
Tawil warned that the sites used for the testing were selected as worst case sites and cannot be used to determine service availability. Based on pictures of the sites in the Linx report you can see he isn't exaggerating!
Bretl also reported that based on simulations, adding enhanced VSB (E-VSB) packets to a DTV signal does not offer much improvement in the Brazil A echo ensemble with fourth- and fifth-generation VSB receivers. Adding E-VSB packets to the Brazil C ensemble provided improvement ranging from 20 to 25 percent on fourth-generation receivers, but minimal improvement on fifth-generation receivers. He concluded that recent receiver designs do not need much help from E-VSB-the main benefit is the 6 dB gain in SNR performance on the E-VSB signal itself. Unfortunately, I didn't see any over-the-air demonstrations of E-VSB at NAB this year.
Adding E-VSB packets does improve equalizer convergence, however. In the Brazil C echo ensemble, fifth-generation receiver equalizer convergence time with 5 percent enhanced packets improved from 230 to 83 milliseconds. For the Brazil E ensemble under the same conditions, convergence time improved from 500 to 220 milliseconds. Adding 7.5 to 15 percent enhanced packets improved Doppler performance in early fourth-generation hardware but was not needed with the new demodulators.
That's it for this month. Next month I'll look at some of the new transmission technology at NAB2004 and also give you an update on the performance of low-band VHF DTV. Many viewers and broadcasters responded to my last article with useful comments. One correction: Several CBS engineers alerted me that the CDBS information on WBBM-DT is out of date. The station is currently operating at 3.7 kW with a license application pending. Also, Jack Davis from KTXL-TV in Sacramento pointed out that KCRA replaced its circularly polarized antenna on Channel 3 with a conventional three-bay batwing antenna last summer. He noted the signal at his house increased by 8 dB when they went back to horizontal polarization.
Your comments and questions on any RF topic are always welcome. Drop me an e-mail at firstname.lastname@example.org. Your question may become the basis for my next RF Technology column!
Future US's leading brands bring the most important, up-to-date information right to your inbox