RF Column 21 - June 1993 Copyright (c) 1993,1995 H. Douglas Lung ALL RIGHTS RESERVED TOPICS: Trends in RF Technology at the 1993 NAB What's the story with HDTV? Update on HDTV transmission system proposals My comments on HDTV system requirements No klystron transmitters at NAB! Panel antennas for emergency use Waveguide switch maintenance -------------------------------------------------------------- This month I'm going to discuss some of the trends in RF technology revealed at this year's NAB in Las Vegas. By now, you've read the columns in the NAB issues outlining the products on the floor at NAB. I'm not going to concentrate on specific manufacturers here - TV Tech's NAB reporters did a good job of that. Instead, I'm going to pass along information I gathered trolling the floor, hospitality suites and technical sessions. There's a growing interest in HDTV. The most common question (technical) I was asked at NAB was "Is HDTV going to happen?" followed quickly by "When?". The NAB program committee obviously felt it was going to happen before April's convention. One session in HDTV World was titled "U.S. HDTV System: And the Winner is...". As it turned out, that session was restructured to give each of the four digital system finalists a chance to talk about their system improvements and how a grand alliance might result in a common standard. I did not get the feeling a grand alliance would happen in the near future. The difference in RF modulation systems didn't seem to provoke arguments. Much of the discussion and arguments were about the relative value of progressive scan (used by Zenith/AT&T's Digital Spectrum Compatible System and MIT and General Instruments' Channel Compatible DigiCipher) and interlaced scan (used by Sarnoff/Phillips' Advanced HDTV and General Instruments' DigiCipher). For the full story on the merits of the two approaches, I'll refer you to Merrill Weiss' HDTV columns here and the 1993 NAB HDTV World Proceedings. The Proceedings have an excellent summary of the tests the Advanced TV Research Committee (ATRC) performed on the various HDTV systems. Contact NAB Services at 1-800-368-5644 or write them at 1771 N Street N.W., Washington, D.C. 20036-2891 for information on ordering a copy. To summarize this massive document, the trained observers subjectively rated the interlaced scan systems higher than the progressive scan systems. After the tests, coding errors were found in the progressive scan compression which resulted in more noise and artifacts in the picture. This was originally blamed on the cameras used to make the test tapes. With the errors corrected, the progressive scan proponents are eagerly waiting the next round of testing, which they feel will vindicate progressive scan. What does this mean to RF engineers? In reality, not much. It is of major importance to engineers involved with program production, since the progressive scan systems appear to be more difficult to synthesize from line doubled NTSC video. The ATRC found the capital cost to construct an HDTV facility about the same for any of the four digital standards. However, if up-conversion from NTSC is more expensive, it will make it more difficult for small stations to "get by" with an NTSC to HDTV black box at the transmitter site. Which is better? The thrifty side of me likes interlaced scan, since the interlacing itself is a form of compression, reducing the data rate and it should be easy to up-convert to 1050 line interlaced HDTV from 525 line 16:9 NTSC. Many of the new cameras shown at NAB and Sony's Digital Betacam support 16:9 525 line video. The purist side of me prefers progressive scan, since it has the potential to offer a cleaner picture. This subjective opinion comes from experience with high resolution computer displays, where IBM's interlaced monitor has been soundly rejected in favor of non-interlaced VGA displays. Read the Proceedings, read Merrill's column, draw your own conclusions. Like video compression, this summary isn't loss-less! Getting back to RF topics, there have been some improvements in the RF portion of the digital HDTV systems. Interference from NTSC transmissions posed some problems for the HDTV systems tested at ATRC. Sarnoff and General Instruments both improved tuner filtering in their receivers since the tests. Sarnoff modified their Advanced HDTV two carrier transmission system to permit varying the ratio between the priority and high quality carriers at the transmitter. Responding to broadcast engineers' concern about the high peak to average ratios required for DigiCipher transmission, General Instruments added a circuit to permit up to 1.5 dB of peak clipping. While the official field tests await the Committee's selection of a winning system, manufacturer sponsored field tests seem to indicate HDTV will work, perhaps even better than NTSC. So how did I answer the question about HDTV? I still believe it is coming. Field tests so far are positive. I don't expect a standard to be selected until sometime next year. After the selection, I believe some of the better funded TV stations will have HDTV on the air within a year. For the rest of us, I don't think the transition will be as expensive as the estimates show, if the FCC permits use of less than maximum power without future penalties. Forget about using your existing NTSC transmitter on HDTV - you'll be broadcasting NTSC for another 10 to 15 years, after which you'll be ready to dump that transmitter. Comark's "Dual Use" transmitter offers hope to those stations forced to buy a new transmitter for NTSC now, but expect to do HDTV soon. If this isn't an option, get something on the air quick using a low cost tetrode transmitter operating at 5 or 10 KW and feed it to an inexpensive low power TV antenna, like Andrew's Kerry Cozad proposed in his talk at NAB. Generate the HDTV signal from the NTSC STL using a converter like Snell and Wilcox or Faroudja are likely to offer. As studio equipment is upgraded, buy cameras and VTR's capable of 16:9 aspect ratio 525 line operation. These are available now. Line double them for HDTV operation. The transition will take at least five years. During that time, most of today's TV equipment will need replacing anyway. The FCC could change all of this if it decides that the digital bandwidth offered by the HDTV compression algorithms could be used to transmit two or more NTSC videos in one TV channel. Regardless of what happens, digital transmission will change the way transmitter engineers handle TV. An extra degree of differential phase or an extra percent of linearity error doesn't mean much now. In the world of digital TV, it will shrink the coverage area, with the potential of having many viewers lose their picture. You'll be busy! If you were looking for a new UHF transmitter you might have noticed that no U.S. manufacturers were showing klystron transmitters, MSDC or otherwise. The choices were transmitters using the Inductive Output Tube (I.O.T.) from EEV, tetrodes, primarily from Thomson and solid state devices. Which of these will work for HDTV? Error free HDTV transmission demands amplifiers which can deliver peak powers several times that of the average power. The exact ratio depends on which system we're talking about, but 5:1 is a good starting point. If you've ever watched what happens when you try to drive a solid state amplifier 100% above its rating, you know that they won't work for high power HDTV unless a large number of additional devices are used to generate the high peak powers. Solid state amplifiers capable of producing 30 KW peak power can be built, but with the 5:1 peak power requirement, the average power will be less than 6 KW, not likely to be enough for the maximum coverage area the FCC will allow. Solid state transmitters will be an expensive option for HDTV. General Instruments indicated they were adding a clipper to their DigiCipher HDTV encoder, which they calculate could result in a 1 dB effective improvement in radiated power - the 1.5 dB power gain is offset by an estimated 0.5 dB loss caused by the errors introduced by the clipping. I heard Harris expects to have a solid state transmitter capable of meeting these peak power requirements in a year or two. I didn't hear what technology would be used, but there was a lot of discussion among engineers not associated with any manufacturer about Westinghouse's new silicon carbide transistor, which can tolerate high temperatures and can handle the peak powers needed for HDTV. If this technology does not pan out, it appears solid state transmitters will be out of the price range for most stations wanting enough power for maximum legal coverage. What about tetrodes? Thomson gave an interesting talk about using tetrodes for HDTV transmission. Acrodyne has done some tests with tetrodes for HDTV transmission. One figure that impressed me in the Thomson presentation noted that their 40 KW tetrode was capable of producing over 100 KW for a 20 millisecond pulse. Such a tube, operating at 15 to 20 KW average power, should be capable of handling the peak powers HDTV will demand. The Thomson presentation showed good uncorrected linearity up to about 25 KW, for NTSC operation. There was some concern that the intermodulation distortion may not meet HDTV requirements unless correction is used. I did not see any details on the phase or linearity specifications of the tube when putting out that greater than 100 KW pulse. From the numbers I saw at 40 KW, I surmise it would require correction to provide an acceptable error ratio for HDTV. This technology does show some promise. Tube life numbers from existing 40 KW tubes is encouraging. The I.O.T. is the only proven device for high power HDTV transmission. I know I said I wouldn't mention manufacturers, but I have to admit I was very impressed with the specifications of Comark's new IOX transmitter. It is the only one I saw on the floor that was truly HDTV-ready, needing only a 45 MHz. signal from the HDTV modulator. The linearity correctors used in conventional NTSC exciters do not permit phase and gain adjustments over the full channel bandwidth. They usually have one circuit for low frequency signals, for luminance and another for chroma, centered around 3.58 MHz. Comark designed their linearity correctors to work across the channel, providing flat gain and phase characteristics at all power levels at any frequency in the channel. One caveat - HDTV encoder manufacturers may decide to incorporate linearity and phase correction inside the encoder. Last year Zenith mentioned they were considering a system which could automatically linearize the entire transmitter chain. I'm not sure they considered the wide range of correction required for high power amplifiers. While the Comark design was clearly designed with HDTV in mind and represents the third generation of Klystrode/IOT designs from Comark, I commend Harris for their first IOT transmitter. The result of a joint effort between Harris and TVT in England (which Harris owns), it is very clean. I like the layout, the service-ability and quality of construction. I'd better mention T.T.C. too before I get in trouble. They tell me their air-cooled I.O.T. transmitters are working reliably in Australia. T.T.C. has also participated in HDTV tests - you might have seen their solid state transmitter at NAB last year demonstrating the G.I. DigiCipher system. Which is the best transmitter for HDTV? We really won't know until the FCC decides on a standard, which could be a year away. The progressive scan HDTV proponents don't really want to hear about interlaced systems and vice versa. Zenith HDTV tests indicated a 300 watt transmitter could cover a 30 mile radius. I'm a little suspicious of these tests. While the Zenith speaker showed a slide of a bow-tie antenna used for reception, it was on top a roof on a mast! I still wonder how much power is needed to reach that HDTV set with a UHF loop on the back, smashed up against a wall. I'm hoping to get a chance to try out some typical antennas during the field tests, whenever they happen. Current transmitter output power requirement estimates from the various digital proponents show peak powers very close to that full power UHF stations are using today. What about antennas? Several manufacturers of UHF panel antennas touted their broadband response as perfect for multi-station HDTV antennas. Most of the panel antenna expertise has come from Europe, where such designs are commonly used with UHF transmitters operating at less than 30 KW, low by U.S. standards. Two U.S. companies decided to offer UHF panel antennas at NAB - Harris and Dielectric. The Harris design looks solid - four bay gain per panel with one feed. The Dielectric design is interesting - their brochure shows the antennas elements are circles. I'd like to have heard the story behind that design! Dielectric has made panel antennas before. The UHF panels used by UHF stations at the World Trade Center in New York were built by R.C.A., Dielectric's predecessor. These panels were not wideband, however. I'm running out of space this month, but I'll try to grab a few more inches of space to drop in a few tech tips. Two months ago I described an emergency transmitter system for TV broadcasters. Since that article was published, Micro Communications has put together a system using wideband all channel UHF panels. Four panels offer a wide range of patterns and enough power handling capability for emergency operation. Contact Dennis Heyman for a brochure on this system. Dennis out did himself coming up with the illustration for it. Ask him about it. Micro Communications can be reached at 603-624-4351 or 603-624-4822 (fax). Finally, a warning for UHF TV engineers with waveguide switches in their RF system. Our group has had failures (some worse than others) in three of its four stations using waveguide switches in the past two years. After the third failure and some discussion with the experts at NAB (especially Todd Loney, from Micro Communications) it seems this problem can be prevented. Even though rectangular waveguide appears to be a sealed system, dust from inside the waveguide run seems to collect in the switches, where it interferes with the contact between the finger stock on the moving vane and the walls of the waveguide in the switch. To prevent the pain of having the switch between the diplexer and the antenna fail, it is important that these switches be kept clean. Swab 'em out with a rag and alcohol, being careful not to leave any lint or rag behind. My experience indicates this should be done every two years or so. Todd pointed out that most waveguide installers do not swab out the line before installing it. The aluminum used in waveguides often has a thin coat of oil on it, which he feels will cause problems eventually if it isn't cleaned out. There really isn't anything to fail in the plain waveguide runs, but pay attention to those switches. Next month we'll get back to more conventional topics, with a free offer for TV Technology readers and details on a nifty isolation amplifier for those pesky meter outputs that aren't referenced to ground. Special thanks to all those I met at NAB who offered comments on the column. Your response and feedback are the real reason I take the time to write this article. Please send me your tips and stories. If you've got a problem where our readers might help, let me know. No promises, but if it intrigues me, I'll write it up and offer my suggestions and invite others to comment. To reach me, leave a message for me via CompuServe - my I.D. is 70255,460. MCI Mail, Internet and other services can send messages to CompuServe I.D.'s. If that isn't possible, phone me at 305-884-9664, fax me at 305-884-9661, try my L.A. number at 818-502-5739 and leave a message or write me at 2265 Westwood Blvd., Suite 553, Los Angeles, CA 90064. ((8/95 > UPDATE! - Use dlung@gate.net for e-mail!)) Copyright (c) 1993,1995 H. Douglas Lung ALL RIGHTS RESERVED