Lessons Learned at IEEE
During the weeks following the attack on the World Trade Center I was busy assisting our local engineering staff in restoring the WNJU signal at the Armstrong Tower in Alpine, N.J. You may have noticed I didn’t have a chance to write for last month’s TV Technology.
I don’t have much to say about the tragedy that hasn’t been said before, but I will add my voice to those asking you to support the Society of Broadcast Engineers’ trust fund to assist the families of the broadcast engineers lost in the attack. Details are available at http://www.sbe.org
The efforts expended by manufacturers, the trades, Chuck Sackerman and his staff at the Alpine Tower site and, of course, engineers from the NYC stations were amazing. After the horror of the destruction of the World Trade Center, I consider myself fortunate to have been able to help restore a part of the broadcast facilities lost.
In the case of our station, electricians, technicians and engineers from Thales as well as WNJU’s engineers and various trades took an empty room, enlarged it, and obtained and installed distribution for up to 600A of 480VAC power. We installed a Thales/Comark IOX IOT transmitter capable of 120 kW output, ran almost 300 feet of rigid line, installed a side mount Dielectric TFU-18DSC antenna (in less than optimum weather) and had the station back on the air at 1,333 kW in slightly over two weeks!
Anyone who has done a high-power UHF transmitter installation knows this is nothing short of a miracle.
In my October article about useful RF information on the Web, I complained about how difficult it was to get tabulated antenna pattern data out of the Dielectric Antenna System Planning Software (DASP). If I had looked a bit closer, I would have noticed that in the File menu on the main page there is an Export option that will output the pattern data as a comma separated value (CSV) text file or files that can be imported directly into EDX or V-Soft propagation software.
The software is available at http://www.dielectric.com/broadcast/
. Thanks for Bob Miers for letting me know about this!
I also lamented the removal of Zenith’s excellent technical papers on 8-VSB from the company’s Web site. I was happy to find them back – but due to Web site redesign, you may have to search a bit to find them. Try the direct link http://www.zenith.com/index.asp?url=./sub_hdtv/hdtv_index.html
If that doesn’t work, go the www.zenith.com. At the bottom of the home page, there is a menu bar with "HDTV" as one of the options; select "About HDTV." The technical papers can be accessed from the menu on the right of the "About HDTV" page. Tutorials are available on VSB, MPEG and receiver technology. IEEE Broadcast Symposium
As I’ve said before, the IEEE Broadcast Symposium is the best conference I’ve found for learning about advances in RF broadcast technology.
While attendance at this year’s Symposium was less than last year’s, the quality of the papers remained high. I’ll discuss some of them in the next two columns. This month, I’ll look at papers covering DTV propagation and reception.
One of the most ambitious papers, DTV Coverage and Service Prediction, Measurement and Performance Indices, was authored by Oded Bendov, John F.X. Browne, Charlie Rhodes, Yiyan Wu and Pierre Bouchard. As Charlie has his own column here at TV Technology, I’ll leave it to him to describe that paper!
An interesting paper from Communications Research Centre Canada (CRC) was presented, describing Comparative Tests for Digital TV Transmission Systems. The paper showed the results of tests of the latest generation of VSB and COFDM receivers, both in the laboratory and in the field.
The most notable difference between receivers in the laboratory tests was in the range of multipath that the receivers were able to handle. Performance of both VSB receivers dropped off significantly when leading echoes arrived more than about five us before the dominant signal.
But one of the COFDM receivers was able to handle echoes arriving approximately 70 us before or after the dominant signal. Over this range, it was able to handle echoes within one dB of the dominant signal even with a carrier-to-noise ratio (CNR) of 22 dB – four dB worse than the 18 dB theoretical threshold for this type of modulation.
Most of the pre-echoes encountered in indoor reception, however, are a couple of us or less and this became evident in the results of the indoor reception tests.
In the absence of multipath, 8-VSB has a theoretical advantage of more than 3 dB in CNR. The laboratory tests showed that with the new receivers, 8-VSB receivers at the 19.39 Mbps data rate had a 3.6 dB or better advantage over COFDM receivers with a 19.76 Mbps data rate (64QAM, 8k carriers, _ code rate).
The COFDM receivers’ CNR performance improved to 3.8 to 5.6 dB when the COFDM data rate was dropped to 4.39 Mbps (QPSK, 8k carriers, _ code rate).
CRC used a directional active antenna with a low-noise preamplifier for its indoor field tests. 8-VSB was compared with the 19.76 Mbps COFDM.
If COFDM reception wasn’t possible at 19.76 Mbps, the data rate was dropped to 11.71 or 4.39 Mbps. (The results at lower data rates were not detailed in the CRC presentation.)
Anyone who has played around with indoor TV antennas knows that some parts of the room have better reception than others. When conducting indoor field tests, it is useful to know how sensitive reception is to the location of the antenna in the room. CRC did what they called "T-Measurements" to determine location sensitivity.
The antenna was moved along the length of one wall – the area where reception was possible was noted. The antenna was then moved between walls, either near another wall or near the center of the room. These methods allowed reception to be plotted on an X-Y graph. CRC also used the traditional way to measure ease of reception – measuring the angle over which the antenna could be rotated without losing reception.
The CRC indoor tests showed that 8-VSB was able to maintain a CNR advantage of 3.5 dB or better, over 19.76 Mbps COFDM.
What was interesting was that 8-VSB could be received deeper inside a room than the 19.76 Mbps COFDM signal. Although the antenna could be rotated over a wide angle while holding COFDM reception at many sites, reducing the antenna rotation angle to less than 135 degrees gave 8-VSB a slight advantage. Reception of the 4.39 Mbps COFDM signal, of course, was more reliable than 8-VSB.
The CRC has planned additional tests with COFDM data rates of 17.56, 11.71 and 4.39 Mbps. It will investigate the use of antenna space diversity and on-frequency repeaters to improve DTV reception.
Tests are planned on enhanced 8-VSB receivers with better training sequences, better equalizers and multimode (2- and 8-VSB) reception. It will be interesting to see how the enhanced 8-VSB modes compare with the lower data rate COFDM modes.
Check the directory of http://www.crc.ca/html/crc/research/broadcast/
for reports on the Phase 1 CRC tests (filename: cdtv_report_8.2.pdf). When the Phase 2 tests are completed, this may be good place to look for that report. ATSC T3/S9 Results
John Tollefson, from PBS, moderated the Panel and Reports of the Results of the ATSC T3/S9 8-VSB Enhancement Group Studies and Lab Tests. Panelists included Mark Richer from ATSC, Robert Seidel from CBS, and Chris Strolle from NxtWave Communications.
Mark Richer reported the T3/S9 evaluation might last until March or April 2002. It should be finalized in the fourth quarter of 2002.
I’ve described some of the proposals for 8-VSB enhancements in previous columns. Robert Seidel was concerned that some of the proposals would lower the data rate of the ATSC signal. He commented that HDTV should not be compromised to provide better reception for portable sets, which, he said, represent a very small part of the audience.
In the Q&A session, Victor Tawil added that the Oren and Broadcom proposals don’t reduce the data rate. It was noted earlier, however, that these proposals don’t allow reception with a lower CNR or threshold of visibility.
Chris Strolle explained the combined NxtWave/Zenith proposal that enhances forward-error correction on a per packet basis. For example, a 2.5 Mbps robust signal could be combined with a 15 Mbps signal. The trade-off in data rate for the robust signal is approximately 2 to 1. The lower data rate has a 6 to 9 dB CNR advantage over the full data rate signal.
This can also be used to improve adaptive equalizer performance at the higher data rate. A 2.5 Mbps data rate should be adequate for a small screen, portable set, while 15 Mbps should provide enough bandwidth for HDTV when using the most recent encoder designs.
Ralph Justus, from the Consumer Electronics Association, recognized that there was a lot of pressure on the industry to solve the problems with DTV. He warned broadcasters that the reception issues the T3/S9 group was addressing should not distract it from solving other, major problems with the DTV rollout – including cable carriage of DTV signals, copy protection and PSIP implementation.
As always, I welcome your comments and suggestions. Email me at firstname.lastname@example.org