Welcome to RF Current, a weekly electronic newsletter focusing on Broadcast
technical and F.C.C. related issues. This newsletter is part of The RF Page @ www.transmitter.com, a web
site devoted to TV Broadcast RF engineering. For more information see the What is... guide to the R.F. Page site.
Issues are dated each Monday, although recently I've needed an extra day or
two to complete each issue. Articles may be posted earlier if time permits or
if there is a major, breaking story.
<<< Back to April 10 - Issue 209
April 17, 2000 - Issue 210 Final Edition
- DTV - Sinclair Demonstrates COFDM Advantages at Acrodyne NAB Booth (Apr. 11)
- Broadcasters had a chance to see indoor reception of DTV at NAB. In the Acrodyne booth at the Las Vegas Convention Center, Sinclair demonstrated COFDM's capability to receive both HDTV and SDTV signals on indoor antennas from KVWB-TV channel 29, on Black Mountain.
Sinclair was able to show reception inside the Las Vegas Convention Center of a signal from KVWB-TV, about 12 miles away, on a hand-held "Mediascreen" tablet from Nokia. People had to make an effort cause the Mediascreen video to break up. Mark Schubin, in his Monday Memo, said he had to crawl inside a transmitter cabinet with the Mediascreen and close the door before reception was interrupted. The signal the Mediascreen was receiving was a standard definition signal with some additional webcasting data added. However, using DVB-T Hierarchical modulation, Sinclair was also able to transmit a 13 Mbps 720p HDTV program with one 192 kbits/sec. Dolby stereo audio channel in the same 6 MHz channel. Receiving this signal required a somewhat more complicated antenna. Sinclair used a Radio Shack Dual Bow-tie with reflector sitting on top of the plasma HDTV display.
Here are the technical details behind the demonstration. According to Sinclair, KVWB-TV was transmitting a COFDM with 8k carriers, uniform 64-QAM hierachical modulation and a guard interval of 1/32. The imbedded high priority SDTV stream appeared as a QPSK signal, with 1/2 forward error correction (FEC) and 204 bytes per packet. The bits used for the low priority HDTV stream appeared as a 16-QAM signal, with 2/3 FEC and 204 bytes per packet. The transmitter on Black Mountain had an effective radiated power of 300 kW.
The modulation method chosen for the high priority SDTV data required a carrier to noise ratio of only 10 dB at the receiver. This 5 dB advantage over the 8-VSB modulation method could provide a way for COFDM proponents to show weak signal reception capability similar to or better than the performance of the 8-VSB even after a reduction in average transmitter power output to accomodate the 2 dB greater peak power COFDM requires. However, it is important to note that this 5 dB advantage exists only for the SDTV high priority data stream, not the HDTV stream. According to DVB, the HDTV signal requires a C/N of 19.6 dB, over 4 dB worse than that measured for 8-VSB.
Sinclair's Mark Aitken provided RF Current with short tutorial on Hierarchical Modulation from DVB. See DVB-T Hierarchical Modulation - A Brief Introduction. Information is also available in the DVB News Release Another World's First - Providing 6 MHz DVB-T Dual Service in A Single Channel. Both of these links point to PDF files. An Adobe Acrobat Reader or browser plug in is required to read them. The Nokia Multimedia Terminals web page has information on the Nokia Mediascreen portable DVB-T receiver/terminal and existing set-top DVB-T products.
- DTV - Improved ATSC Receiver Performance at NAB (Apr. 11)
- In room N207, somewhat hidden away from the main NAB activity at the Las Vegas convention center, Zenith was showing the progress it had made in improving 8-VSB performance. In the demonstration, receivers ranging from the first generation set-top box like the one used in the Sinclair tests in Baltimore to a fourth generation prototype receiver consisting of breadboards in a card frame were subjected to different signal degradations, ranging from multipath to impulse noise. Zenith also had two older COFDM receivers - one a set-top box and the other a TV, operating in an 8 MHz bandwidth. Throughout the range of tests, the third and fourth generation 8-VSB receivers maintained a 4 dB threshold advantage or better over the COFDM boxes. In tests using an ATSC multipath ensemble with phase rotation added to simulate dynamic multipath, the 3rd generation receiver maintained reception with at 7 Hz rate while the fourth generation design was able to hold reception until the variation reached 9 Hz. The 8-VSB and COFDM receivers were subjected to impulse noise generated by a hair drier coupled into the test setup using in interesting coupler taped to the drier's power cord. Zenith tested the threshold performance of the receivers by reducing the desired signal (COFDM or 8-VSB) until the receivers lost the signal. In this test, the second through fourth generation 8-VSB receivers had threshold performance several dB better than that of the COFDM receivers.
I tried to convince the Zenith engineers to attempt indoor reception with one of the new receivers and an Antiference "Silver Sensor" directional indoor antenna. I couldn't talk them into that, but they did demonstrate successful reception using the house antenna system. Indoor signal levels in the convention center were very low and the room the demo was in was not on an outside wall. I didn't really expect an indoor antenna would work, but if it had it would have provided some balance to the Sinclair demonstration. Last year a Thomson/RCA DTC100 managed to identify as digital most of the stations transmitting 8-VSB at NAB 1999, although it wasn't able to lock on any of them, using one of my pocket paperclip indoor antennas. Zenith has uncovered a problem with indoor reception that wasn't addressed in the latest generation receivers. The ATTC's RF capture project found indoor reception often includes pre-echos with delay (advance?) that exceeds of the range of the 8-VSB demodulator's adaptive equalizer. Newer receivers are able to handle pre-echos up to 4 microseconds before the main signal. In an indoor environment, these pre-echos can occur 15 microseconds or more before the main signal. Inside the equalizer range, the demodulator can handle echos within a few dB of the main signal. Outside the equalizer range, the echos appear as noise and must be 15 dB below the main signal when combined. 8-VSB receiver designs with the wider equalization range should make it into silicon within the next year.
Zenith has not posted the papers handed out at the demonstration on their web site and I did not have an opportunity to obtain a copy at NAB. A very brief description of Zenith's work may be found in the Zenith Press release Third Generation Chip Enhances Indoor DTV Reception.
In the Receiving DTV portion of the NAB Engineering Conference, Walter Sidas from CBS reported on indoor reception tests CBS conducted in Philadelphia. In the tests, reception of KYW-DT was compared with the reception of Channel 29's analog signal. At 42 indoor sites, the Antiference log periodic antenna worked best, followed by the Radio Shack dual bowtie and reflector. A loop antenna performed the worst. Indoor DTV reception was possible at 32 of the 42 sites. Ninety-four percent of the sites with better than CCIR grade 1 reception on the analog station were able to receive KYW-DT with an indoor antenna. Sidas said they found clock jitter is a problem on some transmitted DTV signals. The jitter was found to reduce the DTV reception margin by as much as 4 dB.
Chris Strolle from NxtWave authored the paper Feasibility of Reliable 8-VSB Reception. Du ring the presentation at the Receiving DTV NAB session, Strolle said that a system wide solution was needed. This included the transmitter and, most important, its location. Height and high power were important. Better antennas, better tuners with better equalizers and error concealment were needed on the consumer side. He explained that newer tuner designs used both the in-band and quadrature components of the 8-VSB signal. In simulated tests, the new Nxtwave NXT2002 chip was much better than the ATSC Blue Rack receiver in dynamic multipath performance. He said that contrary to popular belief, an 8-VSB tuner can cancel 0-dB multipath and blind EQ does indeed work. Strolle concluded that in two years multipath won't be seen as a problem for indoor reception of 8-VSB signals. (He cautioned that this did not include mobile reception.) Consumer electronics with the new chips should be available in a 12 to 18 month time frame, with PC card designs available in perhaps as few as 6 months.
The paper Feasibility of Reliable 8-VSB Reception by C. H. Strolle, S. N. Hulyalkar, and T. J. Endres, all with Nxtwave Communications, is available on the web in Adobe Acrobat (PDF) format.
- DTV - SET Reports on DTV Tests in Brazil, Proponents Respond at NAB Breakfast (Apr. 11)
- SET invited broadcast engineers to a Tuesday morning breakfast at NAB. SET Group Coordinator Fernando Bittencourt from TV Globo had SET engineers involved with the testing explain the results of some of the key tests. ATSC chairman Robert Graves commented on the performance of the ATSC 8-VSB modulation system while Peter Macavock represented DVB. Mr. Yamata spoke for the Japanese ISDB system.
The lab tests of threshold performance showed ATSC 8-VSB (ATSC) the winner, with a measured C/N of 14.6 dB, slightly better than the expected 15 dB performance. The DVB 2k COFDM (DVB 2k) system measured 19.0 dB, although DVB indicated the performance limit could be as good as 17.5 dB. The DVB 8k COFDM (DVB 8k) tests showed a C/N of 16.9 dB. Peter Macavok indicated that with improved receivers it may be possible to improve COFDM threshold performance by perhaps 1.5 dB. The ISDB-T (ISDB) system had a measured C/N of 18.6 dB, with Mr. Yamata saying that 16.6 dB may be possible with some receiver improvements.
The ATSC system also performed well in laboratory tests of rejection of impulse noise. In these tests, the ISDB system was second best and ATSC was second best, followed by DVB 8k. The DVB 2k system tested the worse. DVB commented its performance would improve if it replaced the pulses affected by the noise with zeros. Mr. Yamata said the ISDB system could be improved even more by using longer interleaving and more forward error correction. Other proponents noted that the same techniques could be applied to their systems.
Multipath tests revealed the known weaknesses with the ATSC system. In the lab tests, DVB 8k had the best results, followed by ISBD and ATSC. The DVB 2k system had the worst C/N performance with static multipath. Robert Graves from ATSC said that systems were senstive to the phase of the multipath and it was important to rotate the phase of the multipath during testing to present all possible phases to the receiver. He felt that the DVB 8k system may not have been tested with multipath at its worst phase angle. This was refuted by the SET engineers, who said that different phases were tested. When asked about future improvements, Graves responded that over 2 or 3 years it may be possible to have ATSC receivers work with 1 dB ghosts.
Examining performance with doppler echos, ATSC was worst. DVB 2k did much better, DVB 8k was somewhat worse and ISBD was similar in performance to DVB 2k. Robert Graves was asked if the ATSC system might someday support mobile reception. Graves replied that a new mode, 2-VSB, that may support mobile reception at lower data rates was under discussion. He did not see mobile reception anytime soon with 8-VSB.
In explaining the results of the tests of outdoor coverage, the SET engineer explained that the most important areas were near the transmitter. In the testing, the DVB 2k system was the worst. DVB blamed this on the poor performance of an early model receiver. ATSC was the next worst. ATSC said the results should be questioned because not all systems were tested at all sites. At problem sites ATSC has seen variations of 55 dB even over short distances. SET responsed that their data is valuable and the for these new tests, 34 sites were retested at the same time. The DVB 8k and ISDB systems tested the best.
One interesting test involved indoor reception. These compared indoor reception under ideal circumstances, indoor reception with a blender generating impulse noise and indoor reception with people moving around in the room with the antenna. Under ideal conditions, ATSC worked at 70 percent of the sites versus 90 percent for COFDM. When the blender was turned on, ATSC worked at 45 percent of the sites and COFDM at 80 to 90 percent of the sites, with the best results for DVB 8k. With people moving around in the room with the antenna, ATSC worked at only 9 percent of the sites compared with 70 percent when COFDM was used.
The proponents were asked give their final comments. Robert Graves, speaking for the ATSC 8-VSB system, said that higher power and a higher antennas than those used in the tests were needed. He also said the Brazil needed to consider the channel planning process. In this area, VSB was better in all the lab tests. He also reminded everyone that the test location in San Paulo was not representative of the conditions in the U.S. DVB representative Peter Macavok said that single frequency networks (SFN) and gap fillers are the solution to reception problems in rough reception areas. He commented that there were 673,000 set-top boxes using the DVB-T standard in the UK. He also said we should expect to see improvements in DVB-T receiver performance. Mr. Yamata said the ISDB broadcasting and manufacturing was moving forward in Japan.
The SET Chairman explained why they had settled on a COFDM system for Brazil. They expect cable carriage of 98 percent in a few years and because, he said, their broadcasters are strong content providers, must-carry was not required. Cable systems would carry their signals. Therefore, the DTV system they choose must compliment cable. Mobile and portable reception are essential and this is the area where VSB fails.
At the end of the breakfast, there was an interesting exchange between Robert Graves (ATSC) and Peter Macavok (DVB). The comment was made that data rate was important, including the familiar refrain "bits are bucks". Robert Graves stated that "The important thing is transmitting bits", to which Peter Macavok responded, "No, the important thing is receiving bits!"
More information on the Brazil tests, including a copy of the earlier test results, is available on the SET web site at www.set.com.br.
- FCC Begins Proceeding to Resolve DTV Receiver Cable Compatibility Issues (Apr. 13)
- In a News Release issued today, the FCC said it "began a proceeding to resolve outstanding issues regarding the compatibility between cable television systems and digital television receivers, set top boxes and other consumer electronics equipment." The FCC's Notice of Proposed Rulemaking (NPRM) addressed two unresolved cable compatibility issues: "(1) how to label DTV receivers with different features, including the proper designation for receivers providing two-way interactive capability; and (2) licensing terms for copy protection technology."
The FCC asked for comment on "how to describe digital television receivers with different capabilities to operate with digital cable television systems." Specifically, the FCC was concerned whether there was a clear distinction between sets equipped with the IEEE 1394 interface and those that didn't include it. The News Release said the FCC "asked for comment on whether the transition to digital requires any changes in FCC requirements for cable operators to offer supplemental equipment to subscribers to enable them to use special features of their digital TV receivers."
The NPRM also addresses copy protection issues. The full text of the Notice of Proposed Rulemaking - Compatibility Between Cable Systems and Consumer Electronics Equipment, PP Docket 00-67 is available on the FCC web site as an ASCII text file fcc00137.txt or a Word dcoument fcc00137.doc. Comments are due May 24, 2000 and reply comments are due by June 8, 2000.
- DTV - Zenith Introduces ATSC PSIP Enhanced VSB Translator at NAB
- The use of PSIP in the ATSC data stream to identify TV channels and map them to specific frequencies has caused some concern among TV stations that use translators to reach viewers in remote areas. If the translator simply converts the incoming DTV stream to another channel, as soon as the consumer's TV receiver decodes the PSIP date it will try to switch the tuner to the originating station's channel, not the translator channel.
At this year's NAB in Las Vegas, Zenith introduced a product to help solve this problem. It was showing an "ATSC PSIP Enhanced VSB Translator" that includes baseband processing. Baseband processing allows the translator to perform error correction on the incoming signal, remove adjacent channel interference and, most important, manipulate the PSIP information to match the output channel of the DTV translator.
More information is available in the Zenith Press Release Zenith Introduces New Product to Advance the Roll-out of DTV Translators.
- NAB Product Annoucements
- Zenith Introduces ATSC Video Server - press release - Server can provide a baseband MPEG stream for system testing or broadcasting of over 100 minutes of stored content. SMPTE 310M and LVDS SPI outputs availalbe.
Additional NAB Items are listed in the April 10 RF Current.
- OTHER Items of Interest
>>>>Next April 24 - Issue 211
Return to The RF Page @
Last modified May 7, 2000 by Doug Lung email@example.com
Copyright ©2000 H. Douglas Lung