RF Column #44 - July 1995 - Doug Lung Copyright (c) 1995 H. Douglas Lung ALL RIGHTS RESERVED TOPICS: RF FTP Site announcement FTP access tips using Web browers Changes at the FCC Web Site sci.engr.television.broadcast USENET newsgroup debuts NAB 95 - Comments on reflections & HDTV Coverage Comparision: ITS-830 & Acrodyne TLU-1KE 1KW UHF transmitters ----------------------- This month I'll finish my report on the technical sessions at NAB with a report on Dr. Oded Bendov's interesting presentation on the coverage impact from structures located near the ADTV antenna. I'll also give you my first impressions after installing both the I.T.S. ITS-830 1 KW LPTV transmitter and Acrodyne's new TLU- 1KE 1 KW LPTV transmitter. First, however, I'd like to announce that I now have an FTP site open for the programs and spreadsheets I write for this column. After emailing various copies of my path calculation spreadsheet via America On- line, Compuserve and uuencoded email on the Internet I felt there had to be a better way. Worse, I lost track of who I'd sent copies to and who still needed them. No more! You can find all the path calculation spreadsheets in PKZIP archive format in my FTP directory. Look for the spreadsheet extension inbedeed in the filename to tell which file will work with your spreadsheet program. Archives with "XLS5" in the name are for Excel V5.0. Archives with "WK3" are for Lotus 1-2-3 V3.0 or later. Those with the old "WK1" Lotus extension should work with most any spreadsheet, but without the graphs. The address for the files is "ftp.gate.net". The files are located in the directory "/pub/users/dlung". I may have added subdirectories for different topics and some back issues of the columns in ASCII text format by the time you read this. To access my files from any Internet FTP program simply enter "ftp.gate.net" as the address (ignore the `"`s when entering any of the addresses, file names or directories I mention here), log in as "anonymous" and use your email address as the password. Note that both America On-Line and Compuserve try to help you out by providing a list of popular FTP sites. Don't bother looking for my site there. Check the menus for an option that allows you to enter a specific FTP site or click on and replace the default FTP address (whatever that is) with "ftp.gate.net". My FTP server permits only 20 connections at a time. You may get a message that it can't connect or that the connection was refused if you try to connect at a busy time. Either keep trying until someone signs off or try again during a more quiet time. A Web browser is the quickest way to get to the files. Enter "ftp://ftp.gate.net/pub/users/dlung" as the URL and it will take you right to my top directory listing. When I get a home page designed you may see it as well. If you are using Netscape's browser, holding down the SHIFT key when you click on the file you want will allow you to save it directly to disk. If you click on the file without the shift key it will try to view the file and put up an error box. Click on the "save to disk" option if that happens. Email me at "dlung@gate.net" if you have any problems getting the files or have files you would like to contribute a file. I've found the "ftp://" URL trick handy the last few weeks while the Federal Communications Commission has been messing around with its Web site. The new FCC Web site is now on line. Gone is the simple interface and the link to the FCC Gopher for files. You now get a full color picture of the Commissioners and a limited button menu for navigation. I don't like it. I don't mind the picture, even though it takes a while to load. (Most browsers allow you to turn off graphics.) What I do mind is that it seems to take MUCH longer and MANY more clicks to find the files I want. Furthermore, for some reason Netscape doesn't seem to cache the directories, so I have to wait for them to reload every time I go back a page. Here's the way around it. You can use your browser to access the FCC Gopher or FTP site by typing "gopher://gopher.fcc.gov" or "ftp://ftp.fcc.gov/pub" respectively. The FTP directory is updated before the Web or Gopher directories. I expect the FCC will improve the Web interface to their site. The new interface was up only for a few days before this column was completed and I'm sure in the future they will return to more descriptive file listings and a more straight forward interface. One final bit of Internet news. Last month I mentioned the Call For Votes for a new news group called "sci.engr.television.broadcast". The group is now official and news server administrators should have received notification about it. If you don't find it on your system, contact your system administrator or Internet provider. Early message traffic on the system seems to have a high "signal to noise" ratio, although without the breadth of BPFORUM on Compuserve. I do see more of an international presence in the group than on Compuserve. Here is the charter of the news group, as posted in "sci.engr.television.broadcast": ======== Newsgroups: sci.engr.television.broadcast Subject: About this newsgroup From: robgc@freenet.edmonton.ab.ca () Date: 27 May 1995 00:35:26 GMT It only took about 5 months of planning, but now we finally have a newsgroup of our own. I'd like to thank Mark James for all his advice during the process. CHARTER: sci.engr.television.broadcast This newsgroup will be used for the technical discussions of equipment relating to the operation of a broadcast television facility. This scope will include equipment that is used to transmit, store, manipulate, or generate broadcast quality signals. Engineering standards, practices and troubleshooting will also be discussed. Recognized groups (such as NAB, SMPTE, etc) will also be encouraged to discuss upcoming conferences and information relating to the broadcast industry. (end of charter) Thanks for all your support. Rob Chambers. ======== I've always made it a point to attend Dr. Bendov's NAB presentations because his papers have focused on how we can't use the same assumptions we do for analog TV for digital TV. Last year he pointed out how high UHF digital channels would suffer in coverage versus low UHF digital channels, even though the same effect goes unnoticed for NTSC. This year, in his paper titled "A New Approach to the Analysis of Adjacent Structure Effects on HDTV Antenna Performance", he showed that side mounted antennas that work perfectly well with NTSC may have big problems with digital signals. Here's a quick summary of the paper in my words. We know that structures adjacent to a radiating antenna will reflect the signals from the antenna. These reflections, because they arrive so close to the time of main signal, do not appear as ghosts. However, the reflections will mix with the main signal. If the signals are in phase they will add. If they are out of phase they will subtract. The stronger the reflected signal is, the greater the effect. Because we are talking about structures close to the antenna, some of these reflections may approach the power of the main beam. Note that the phase of the reflected signal, which will determine whether it adds or subtracts from the main signal, is frequency dependent. Since most of the energy for an NTSC signal is concentrated at the visual, chroma and aural carriers in fairly small bands, chances are for most areas the frequency notches caused by the reflected signals will not affect the picture. For a digital signal, the result is much worse. Energy is distributed across the spectrum and all of it is needed. A notch at any frequency in the band will severely limit the ability to recover the data. Equalizers can handle small variation, but if the notch from the reflected signal takes the received signal into the noise, nothing will help. What does this mean? It means that reflections from adjacent structure, be it a tower leg, transmission line or another antenna on a tower top candelabra will produce significant notches in the coverage pattern. If a side mounted antenna is used for digital transmission, pattern selection to avoid putting energy back into the tower will be extremely important. The best option is to use a top mounted antenna. Stations with top mounted antennas for NTSC might want to consider switching to a side mounted NSTC antenna to allow a top mounted ADTV antenna. Another solution, which Oded didn't mention in his paper, is one that his company, Dielectric Communications, is offering. They have replaced a number of existing VHF antennas with new antennas that include a dummy section below the VHF antenna. When the station is ready to go ADTV, they can replace the dummy section with the ADTV antenna. This avoid the reflection problems side mounted antennas would face. I suspect the final result won't be quite as bad as Oded predicts. That's because there will be more than one reflection and chances are that with the combination of the reflections there will be enough signal at all frequencies for the ADTV receiver's adaptive equalizer to work with. I noticed this while testing a side mounted antenna with a predicted deep null (greater than 25 dB) in one direction caused by the tower. I obtained an STA from the F.C.C. to use a small Scala "Paraflector" antenna fed with about 50 watts tapped off the transmission line to fill in the hole. My tests showed it didn't make any difference. The energy reflected from buildings in the area contributed more than enough signal to fill in the hole, with multipath, of course. On the other hand, because ADTV coverage will be an all or nothing issue, it behooves us, as RF engineers, to recommend options that will provide the best coverage possible. When ADTV goes on the air I'm sure we will see a much greater variation in coverage between stations transmitting from the same location than we do now. The reason is things that don't matter much for NTSC coverage will become critical for ADTV. They include transmitter linearity, antenna linearity and antenna pattern performance. If you've been reading my columns over the last few months you may remember I planned to install two different 1000 watt transmitters for my next two LPTV stations. I put the Acrodyne TLU-1KE into service at W32AY in Boston the end of March and completed installation of the I.T.S. ITS-830 at K40DX in Abilene, Texas the end of April. It's too early to start comparing reliability, but both transmitters are working fine as I write this (knock on wood...). I have one of my new remote controls monitoring W32AY and except for a forced shut-down to test the system it hasn't logged a single hit since it went on the air. Due to difficulty in getting a phone line to the Abilene site I haven't been able to monitor it 24 hours a day. Reports from the local engineer monitoring the transmitter show everything is working fine. You may wonder why I decided to compare two different transmitters in a real world environment. First, I'd had some problems with the performance of some earlier Acrodyne solid state transmitters I'd bought and I wanted an alternative. I'd heard good reports about the I.T.S. and I was impressed with its design. I didn't drop Acrodyne because, as in the past, they responded to my problems and continued to improve their 1 KW transmitter designs. They built a transmitter for Channel 32 before I signed the P.O. and gave me a chance to evaluate it before proceeding. In an earlier column I compared the features of the two transmitters and felt that each had its advantages and disadvantages. A combination of the two would have been best! I won't repeat that analysis here, but I will tell you how each installation went and how each transmitter is performing. Acrodyne delivered the Boston transmitter to the site, which had a loading dock and elevator to the transmitter room. Because the transmitter went into what was previously a tank room, it had to be lifted over a barrier. I'd hired Richard Dudley's Boston firm D.T.S. Communications to handle the antenna and transmission line installation as well as the transmitter placement. They carefully lifted the transmitter over the barrier and set it in place. Abilene was a more difficult proposition. The site was quite a way off paved roads and, of course, there was no loading dock. Texas Communications worked with me on the installation and fortunately the tower crew was on site when we arrived. By removing the amplifier modules and exciter drawer we were able to reduce the weight of the transmitter to be point it could easily be moved into position. Unfortunately, getting the two six hundred watt amplifier assemblies, each with its own internal power supply, back into the cabinet wasn't easy. I.T.S. uses top and bottom slide rails similar to those you'd find for horizontal rack mounting. Getting them lined up while suspending the amplifier (even if it wasn't a hundred pounds, it seemed like it) in air wasn't easy. I finally ended up using a piece of wood to support the amplifier at approximately the right height and guiding it in. Considering the difficulty involved in doing this, I think I.T.S. would have been better off putting the amplifier on wheels, like a klystron dolly, and letting you wheel it in and out of the cabinet. After all, alignment inside the cabinet isn't critical. One of the things that keeps bringing me back to Acrodyne is their excellent field service. You can purchase Acrodyne's turn-on service as an option. I strongly recommend it. Brian Truong's experience allowed us to put the station on the air quickly. Part of the turn on package involves testing the antenna, and in this case he uncovered a serious problem that fortunately was easy to fix. Out of the box the Acrodyne met all specs. Brian tried to improve linearity, but found that with the current exciter system there was a trade off between linearity and intermodulation performance. Accepting a luminance linearity error of almost 10 percent allowed him to reduce in band intermodulation to better than 57 dB below visual carrier. This was with an aural power of 100 watts, 10 dB below visual. Acrodyne knew I'd be studying the transmitter's performance and had made an effort to get it looking as good as possible. The transmitter included a new Acrodyne intermodulation corrector and an IF linearity corrector. I was on my own with the I.T.S.. The transmitter was given no special treatment at the factory and turn on service wasn't provided. This gave me a chance to get a good feel for the transmitter, not only how well it performed but how easy it would be to maintain. After installation I found that the visual carrier frequency was over 1,000 Hz. off in frequency. Although K40DX does not require precise frequency offset, I'd specified zero offset (+/- 1000 Hz. error) and it was quoted as such. I verified my readings were correct by setting the Tektronix 2712's reference frequency with WWV at 15 MHz. and then measuring WWV's frequency to confirm accuracy better than 0.1 PPM. Adjusting the exciter frequency was easy. After adjustment, it appeared to be stable. I'll be checking it again in early July. I found that the sync level of the transmitter was inversely proportional to exciter temperature. The sync level meter on the front of the transmitter, although stable during APL changes, never gave an accurate reading. After the unit warmed up I had to readjust the exciter's sync level from 36 IRE units to 40 IRE units. After making this simple adjustment I measured linearity and found the luminance linearity to be much better than the Acrodyne transmitter - within 2 percent. Broadband linearity was excellent as well. All the color bar dots fell within or on the two percent boxes and measurements confirmed differential gain was within plus or minus two percent. I didn't have the ability to do a true differential phase measurement, but looking at a vectorscope it appeared to be within two degrees. As with the Acrodyne transmitter, the sync was clean with no significant spikes or slope. I measured the in band intermodulation (aural carrier minus chroma carrier) equal to or better than 54 dB for any test signal (modulated stair step, red field, color bars, etc.) I fed into it. This was with the transmitter operating at 50 watts aural, 13 dB below visual carrier. As far as maintenance is concerned, I.T.S. wins points for a very straight forward, well labeled exciter system. Acrodyne's is spread out over multiple chassis and is not always well labeled (although the manual has clear drawings). I have one major criticism of the ITS-830. It is extremely difficult to do a legal FCC frequency measurement without using a spectrum analyzer. As you may recall from a column I wrote a long time ago, the FCC is concerned with two frequencies as far as LPTV transmitters are concerned. The first is the visual carrier. The second is the aural intercarrier. (Notice that it is the intercarrier, not the actual aural frequency they are concerned with.) While there is a front panel jack on the exciter to monitor the output of the transmitter, there is no switch to turn off the aural carrier. That is necessary in most cases to get a stable reading. While I didn't try it (I had a spectrum analyzer) it is possible to kill the aural carrier by pulling a connection inside the exciter. No provision is made for measuring the aural intercarrier frequency externally. The only way I can see is to go inside the exciter and get the signal off one of the boards. I.T.S. does provide a sample of the IF and local oscillators on the front panel and you can calculate the output frequency from these frequencies. Over all, I like the ITS-830's layout. It's simple, everything is accessible and the exciter cover pops off with quarter turn fasteners, so getting inside isn't as bad as it sounds in the previous paragraph. I'd like to see the current metering switch and connections brought out to the front panel of the final amplifiers to make it easier to do quick system checks, but other than that and the difficult slide system, the design is good. I.T.S. offers a reasonably priced spares kit for the ITS-830. I.T.S. mounts their transmitter in a 24" wide rack, while the Acrodyne unit uses a standard 19" rack width. I.T.S. allows air to flow freely through the cabinet, while Acrodyne filters all air into the cabinet. I'm not sure which is better. I.T.S. says that in a dusty environment the dust will blow through and out of the heat sinks instead of clogging a filter. On the other hand, if you have a dusty environment, filtering the air and keeping the filters clean might reduce maintenance later. Over the years, as I've installed several Acrodyne transmitters, I've pointed out areas that needed improvement and Acrodyne often improved them. Therefore, the TLU- 1KE already included many of my recommendations. Acrodyne dropped the hard to understand amplifier transistor current LED's and replaced them with a multiple position front panel switch that permits viewing the current through each of the amplifier's transistors. This makes it easy to spot a failed transistor. In addition, the aural carrier kill switch is now on the front panel, along with two jacks for measuring the RF output frequency and the intercarrier frequency. Acrodyne has continued to refine their front panel metering and status indicators. My only criticism is that Acrodyne's economy exciter requires outboard correction to work properly, so you end up with video and IF looped through multiple chassis. One benefit of this approach is it is easy to bypass a failed assembly. I didn't try lifting one of Acrodyne's new 800 watt amplifier modules, but I have replaced some of their predecessor 300 watt units by myself without help. Acrodyne mounts their amplifiers at the top of the chassis, almost at eye level, instead of at floor level like the I.T.S. The power supplies are in the bottom of the rack, not in the amplifier. I like the idea of one power supply per amplifier chassis. That's standard on the I.T.S. but is an option on the Acrodyne unit. Both units have a remote control interface. The Acrodyne has more straight forward on-off control, while the I.T.S. unit mirrors the front panel switches, which require two switch changes to turn off the transmitter -- switch auto to manual then switch from operate to standby. I like the D connector remote interface on the ITS-830. You can make up the cable at a comfortable bench and just plug it in. The Acrodyne uses a terminal block on the top of the transmitter. Ideally a manufacturer would give both options - a D connector for a major remote control interface and a few screw terminals for simple control and monitoring functions. The ITS-830 is a clean design and the output has significantly less mechanically induced phase noise than the Acrodyne TLU-1KE does. I've had many transistor failure problems with old Acrodyne 1 KW solid state transmitters. However, the two Acrodyne units I installed in 1993 haven't shown any problems. I.T.S. claims a better record for their ITS-830 units, with no field transistor failures reported. On the other hand, they said part of the reason for this was their transmitter worked at lower collector voltages than the Acrodyne. I didn't see this. Both transmitters showed similar collector voltage, around 30 volts. I'm out of space again! Next month I'll give you the tips I've been promising on using the enhanced Blue Earth BE-485 Micro in remote control designs and I'll also try to finish up the path calculation spreadsheet series I've been delaying for a couple months now. I've received some good tips and info from readers and I'll be passing them along to you as well. Your comments are always welcome. Send them to me at dlung@gate.net or, if you are on Compuserve, use my PPN, 70255,460. I travel a lot and when I'm in the office I'm usually busy. You can try me at my office number, 305-884-9664 after 6 PM eastern time or fax me at 305-884-9661. I maintain a mail drop in Los Angeles for correspondence but be aware that it can take several weeks for me to receive the mail, especially if I'm traveling. The address is 2265 Westwood Blvd., Suite 553, Los Angeles, CA 90064. Copyright (c) 1995 H. Douglas Lung ALL RIGHTS RESERVED