RF Column 25 - November 1993 Copyright (c) 1993,1995 H. Douglas Lung ALL RIGHTS RESERVED TOPICS: SBE 1993 - Miami - Wrapup: Low data rate video compression for news-gathering PBS Plans for compressed digital video on satellite Transmitter and antenna manufacturers' HDTV designs HDTV Advisory Committee plans & progress - no cheap sets. The elusive Westinghouse Silicon Carbide RF transistor --------------------------------------------------------------- I delayed writing this month's column so that I could give you an update on this year's SBE (Society of Broadcast Engineers) convention in Miami, Florida. It was combined with the RTNDA (Radio Television News Directors Association) convention. As a result, there were a lot of SNG and ENG truck manufacturers represented. Unfortunately, the technical sessions didn't leave me much time to view the exhibits - the exhibit hall closed at the same time the sessions ended! I skipped one afternoon session to take a quick run through the hall and noticed a few new products. As with last year's convention, I found the technical sessions to be a bit less of a sales job than at NAB. The smaller size of the sessions made it easy for the audience to question the speakers. Read on to find out what went on in the RF related sessions, what interesting products were shown and some comments on one of the hottest "non-products" talked about by transmitter manufacturers. Whenever I mention video compression I get a lot of phone calls and E-mail, so I'll cover that area first. Colby Systems Corporation (415-941-9090 voice, 415-949-1019 fax) has shown their "VideoModems" (a trademark of Colby Systems) before, however, at SBE/RTNDA they showed a number of enhancements. The "VideoModems" do not use the conventional processing most engineers are used to seeing in digital video equipment. Colby uses an array of Texas Instruments DSP (Digital Signal Processor) chips. I found their little $895 box most interesting. It permits full color video transmission up to 1 frame per 12 seconds over a conventional dial up phone line. While not too hot for breaking ENG, it should do a fine job for such things as traffic reports (put one in a building near each major intersection) or security monitoring of the transmitter site or other remote locations. You could even hook one up to the video output of a TV set to check video from a remote translator or LPTV site. You'll need one modem at each end. This technology might also help alleviate a problem discussed in Saturday's SBE session on broadcast auxiliary services. Police departments are licensing microwave units in the 2 GHz. band for surveillance purposes, on the same frequencies used by broadcasters. One of these "VideoModems" is a lot cheaper and by using phone lines, it should be a lot more secure. For more serious news gathering, Colby introduced a briefcase sized ENG unit! The unit included a cellular phone (3 watt), a data interface, a 12 dB yagi antenna, an LCD color monitor, and a more sophisticated "VideoModem". The "VideoModem" used here is capable of higher speeds than the little $895 box. You can get fairly fast frame rates (one or two seconds per frame) with reasonable quality, provided the image doesn't change much. I wanted to pan the camera they had in their booth, to see how the modem handled that, but it was sealed up in an enclosure and the remote motor control disconnected. Charles Colby, President of the company, recommended grabbing a series of slow speed, full frame, single frames with a still store, then compressing the live image (the "VideoModem" permits remote adjustment of frame rates and size) so that fewer pixels need to be transmitted. This makes the motion look more realistic. During the newscast you show the stills full screen with the live reporter in the corner. This more sophisticated "VideoModem" costs $11,999, and again, you need one back at the studio as well as in the field. In addition to the "VideoModem", Colby recommends a Motorola 3260 Codex for the data transmission, at an additional $1,199. This technology could change international news gathering - Colby can marry his "VideoModem" to one of those new suitcase sized Immarsat satellite telephones for use anywhere on earth. Immarsat can support up to 64KBs data rates, so image quality is better than the cellular system. The Immarsat unit supports two way voice communication simultaneously with the one way video. Video compression plays a major role in Public Television's future, according to Howard Miller, Senior Vice President of the Public Broadcasting Service. During his Thursday morning technical session, Mr. Miller outlined the PBS of the future - one based on a very flexible digital distribution network. AT&T's Telstar 401 satellite will be largely devoted to educational uses. PBS plans to use this satellite not only to provide video feeds for its over the air affiliates, but provide specific programming to education institutions and companies on demand. Video quality levels will vary, depending on the end application. PBS is planning to use General Instrument's Digicipher system. I understand initial installations will use Digicipher 1, a non-MPEG system and later upgrade to Digicipher 2, which will include full MPEG2 compression. PBS clearly sees its future in services beyond over the air TV. It will be interesting to watch PBS implement this system. It could become a model for commercial network television ventures - perhaps substituting home shopping for education! The SBE devoted its Friday morning sessions to HDTV. Even though the Grand Alliance hasn't chosen a modulation system, transmitter and antenna manufacturers are busy designing products and positioning themselves as HDTV suppliers. This positioning sometimes involves making some comparisons and tests that, while correct, can be misleading unless you watch carefully. For example, a test comparing two amplifiers' performance showed roughly equal performance, until a close look showed one amplifier was putting out more power and therefore was working closer to saturation. The results might have been the same at equal powers, but why the difference? Another antenna comparison showed amplitude and phase response differences in the vertical pattern between two types of antennas. Very briefly, the purpose of the comparison was to show the superiority of center fed pylon antennas over end feed antennas. The end fed antenna had a gain of over 20, while the center fed antenna had a gain less than 15. All things equal, a lower gain antenna will always have better amplitude and phase response in the vertical pattern. I'd expect at equal gain, the center fed antenna would still be better than the end fed, so why use antennas with such different gains when the point could have been made more scientifically? I suppose because the low gain graphs looked so much better. Again, to the credit of the manufacturers, no attempt was made to hide the data and anyone viewing the data critically would know we weren't seeing "apples and apples" comparisons. In a future column I'll talk some more about interpreting test data and what to look for on a transmitter spec sheet. Enough editorializing. Joe Zuba from Dielectric had some interesting data on an RF system they are building for Public TV Channel 28 in Salisbury, Maryland. The panel antenna had to be designed to work on any of four UHF channels they might be given for HDTV, so that when they receive the channel, they won't need a new antenna. The antenna is fed with rigid line. As you probably know, the length of rigid line sections is different depending on the channel. This is necessary because a certain wavelengths (channels) the reflections from the flanges add up. Dielectric avoids this problem with their DigitLine rigid line - it can be used across the band. Channel 28 its HDTV channel could be 25, 28, 51 or 60. Dielectric's tests showed the worst case VSWR for the DigitLine over these five channels (including 28) was 1.06 to 1, on channel 51. Harris Corporation's Robert Ponka and Comark Communication's Nat Ostroff discussed transmitter design for HDTV. I missed Nat's presentation, but a copy of his paper is in the SBE Proceedings. One thing becomes clear reading Nat Ostroff's paper - TV engineers will have to adopt a new way of thinking about transmitter performance. With digital transmission, transmitter quality will be as important as transmitter power in determining coverage area. Nat describes the importance of Comark's broad band linearity correction and selection of Class A rather than Class AB amplifiers in driver circuits. He used a graph comparing symbol error rates (SER) versus carrier to noise ratio from a transmitter with a class AB driver and a transmitter using a class A driver. Both drivers fed an IOT final amplifier operating class AB. From the tests, he concluded that the Class A driven IOT would work with a carrier to noise ratio 5 dB less than that needed for the same SER from a Class AB driven IOT. I noticed that both graphs showed the correction optimized for NTSC - would they be different if correction was optimized for digital transmission? The IOX modulator has wideband correction capability, so such a test should be possible. Robert Plonka talked about the HDTV transmitter testing at Harris. The session was interesting as Bob showed test results using not only using IOT transmitters, but data from the existing Harris MSDC klystron UHF and Platinum Solid State VHF transmitters. His talk is not in the Proceedings, but if you contact Harris, they may be able to provide it. One thing I found interesting was a test signal Harris developed for checking and optimizing linearity over the power ranges HDTV demands. It looked like a conventional ramp, except instead of starting at black or blanking, it started at sync tip! Robert Plonka showed how various classes of amplifiers passed the signal. Paul Misener, an attorney from Wiley, Rein and Fielding who has participated on the HDTV advisory panel had some interesting observations in his talk. One point he made early on was that we shouldn't count on the allocation table released earlier by the FCC staying the same - the question of what to do with the VHF channels has not be decided. Also, the modulation scheme hasn't been decided on yet. The HDTV Advisory Committee recommended additional auxiliary services (studio transmitter links, relays, etc.) frequencies be allocated for stations broadcasting HDTV. Two of the bands the Committee recommended are currently used by the U.S. government and may be released for other use as part of the overall re-evaluation of government spectrum needs. Paul Misener said he expected testing to start towards the end of next year, with an FCC system selection in late 1995, after a comment period. This would start the clock for HDTV construction. I asked Paul if the Committee had discussed low cost, small screen portable TV's in the $300 range for HDTV. Such sets would not have high resolution CRT's and would serve as second sets for the kitchen, bedroom, etc. Paul said no, such a low end TV was not what HDTV was for. The extra cost of the HDTV circuitry in a large screen TV costing over $2,500 would not be that significant and he felt manufacturers would starting including HDTV capability in all large screen TV's as soon as practical. Putting my editorial hat back on, I was very disappointed by this. First, most households are not likely to stick large screen TV's in the kitchen or in the bedroom. (Well, I suppose some may end up there...) Small sets will continue to be used these spots, also in hotel/motel rooms and in many non-affluent households as the primary TV. The big sets that stay in one place will most likely be hooked up to cable. The small sets, except for the hotels and apartments, probably use a loop or rabbit ear antenna for over the air TV. If digital reception will only be available on big screen TV's, which most likely will be hooked up to cable, why bother with over the air HDTV? I feel one of the advantages HDTV can offer is higher quality, cleaner reception on small TV's as well. Even if the tube can't display the full resolution, the result will be better than any over the air NTSC is now. While data shows 60 percent of TV households have cable, it doesn't show how many small non-cable connected TV's there are in the same household. Don't write off this audience! Merrill Weiss had an excellent presentation on budgeting for HDTV, with a number of valid suggestions for reducing the cost. Since he has his own column, he can tell you about it there! Saturday at the SBE convention was FCC day. I moderated a session on common-mode TV transmitters, however, as I'm out of column space for this month, I'll have to postpone my report on this part of the convention until next month. This was the last SBE convention. In the future, SBE will manage the engineering sessions at the NAB (National Association of Broadcasters) convention. I'm sorry to see the SBE conventions go. The small size that made them so attractive also made them financially impractical. There's so much to do at NAB it is tough to devote a few days to concentrating on engineering. SBE and its officers deserve the support of all broadcast engineers and manufacturers. Oh yes! I said I was going to tell you about the hottest "non-product" at the SBE. Westinghouse had a suite at the Fountainebleau Hilton where they described their hot new transistor for high power TV transmitters. The transistor is reportedly capable of power levels twice as high as is now practical for conventional transistors - I heard up to 500 watts. Westinghouse developed the technology for the military and is now looking for non-military uses for it. I've found VHF TV engineers are reluctant to give up the redundancy and simple operation of their solid state NTSC transmitters when they go to high power UHF HDTV. The Westinghouse silicon carbide transistor offers them some hope they won't have to. Digging a little deeper, I found that so far, the transistor hasn't operated at power levels over 20 watts - one report said 1 watt CW (continuous wave). A lot of work needs to be done to increase power levels. Why is this product so hot? A bit of a pun, really. The transistor operates at over 300 degrees Fahrenheit! It must be preheated before it can be used, like warming up a filament. I couldn't get by the Westinghouse suite, but I'll try to get some more information on this interesting device for a future column. For now, don't throw away your tubes! That's it for this month. Next month, the FCC at the SBE convention, evaluating common mode transmitter NTSC performance, some new products from Andrew Corporation shown at SBE and some items from readers. By the way, I have completed an early design for my cheap WWV calibrator/receiver. It should cost under $25 to build. I haven't had a chance to test it yet, but if you would like a very preliminary schematic to build it yourself, let me know. Full details here when finished! As always, I welcome your comments and contributions. Drop me a letter at 2265 Westwood Blvd. Suite 553, Los Angeles, CA 90064, fax me at 305-884-9661, call me at 305-884-9664 or, the best way, send electronic mail at CompuServe ID 70255,460. You can reach my CompuServe mailbox through MCI Mail and Internet, as well most other computer networks. ((8/95 > UPDATE! - Use dlung@gate.net for e-mail!)) Copyright (c) 1993,1995 H. Douglas Lung ALL RIGHTS RESERVED