High Powered History

Size: px

Start display at page:

Download "High Powered History"

Randolf Long
5 years ago
Views:

The Broadcasters Desktop Resource www.thebdr.

As participants pass away and memories fade, facts behind events often get lost. So, when telling a story, it is always best to get it from the people who were there.

1 The Broadcasters Desktop Resource edited by Barry Mishkind the Eclectic Engineer High Powered History Building the Sears Tower Site Part 3 By Warren Shulz [July 2014] History can be a slippery thing. As participants pass away and memories fade, facts behind events often get lost. So, when telling a story, it is always best to get it from the people who were there. In this series, Warren Shulz is sharing his first person recollections with us of how the Sears Tower the tallest building in the US for over 40 years came to be the primary broadcast site in Chicago. The story continues. In this third part, we will take a closer look at the FM Master Antenna installed in 1982, and the coverage pattern that we were able to get from the individual antennas. THE FM MASTER ANTENNA At the same time the single-bay Cavity Back Radiator (CBR) antennas were being developed for the Sears Tower, Harris was developing a 12-bay CBR array for a 12-station diplexer for the Senior Road project, which was in Houston, Texas. In fact, the CBR originally was a TV design a panel antenna Harris was promoting for the TV broadcasters. The FM antenna was a spin-off from a Channel 6 TV antenna. In addition to the circular cavity a grid screen at the rear provides isolation from the supporting tower structure for additional isolation from the structure. The Senior Road Project was ahead of the Sears Project, and modeling was done as well as fullsize testing. (There is a sad historical note related to the Senior Road Project: the 12-bay antenna section at the Senior Road Project caught on a guy wire and the ensuing accident toppled the entire tower. So the initial design-build was destroyed and was again rebuilt. The Senior Road tower was rebuilt and the Senior Road Project was completed.) CBR ANTENNA FEATURES The Harris CBR antenna was designed with a number of features: A low windload by virtual of the galvanized steel grid structure. A broad VSWR bandwidth reduced any issues from ice buildup, Each antenna was assembled and tested at the Palmyra test range near Quincy, IL, then taken apart for shipping. De-couple chokes were used on the radome support arms. These were in three sizes low, mid, and high parts of the FM band.

2 Un-jacketed semi-rigid 7/8-inch line was attached on as many places as possible on the tower to feed the CBR elements. Each semi-rigid line was custom fitted onsite The radome bracket I-beams connected to 4-inch square arms, extended to arm length and supported with a Teflon round spacer. Insulator pads behind the round basket were applied to contact points of vertical support bars. Wire jumpers went between the tower legs and back screen of the CBR panel. These were at four locations at each tower leg for a total of 24 per panel. It fit into a 7½-foot tower module. (Remember, it was like a foot looking for just the right shoe size.) PATTERN PERFORMANCE A lot of detail became apparent from full size pattern testing. Each of the five original antennas were assembled on a duplicate tower section and tested for VSWR and pattern optimization. This was done at the Harris antenna test range in Palmyra, MO, across the river from Quincy, the same range Harris used for TV antenna testing. Dick Fry, a retired Harris engineer, reports the site had a reputation for having the best far field measurement performance. When re-assembled at Sears the un-corrected VSWR was re-verified as measured at the test range. Each three-sided CBR array had two feeds per panel for a total of six feed lines, made up of unjacketed semi-rigid 7/8-inch 50 Ohm feed lines. All lines were pressurized to the feed bowl of the antenna element. The six lines terminated to a power divider that stepped 8.33 Ohms to the 50 Ohms power feed input. A fine matcher follows the 3-inch rigid power divider for final tune up. These inter element feed lines were custom fit on-site from a bender Harris brought to the site. (However, I do not know what was done for the added three antennas after the original five were installed on-site.) It is of interest to note that the Harris CBR took 18 months from initial research to delivery and installation. A VERY NICE PATTERN The Harris cavity back radiator was the right choice for the Sears Tower: it has a true circular polarized pattern as shown in the following axial ratio plot: Axial ratio is a measure of the radiated field at all angles of rotation. It can only be measured with a rotating dipole under near free space conditions. The result only could be achieved with a panel antenna and in particular this CBR design by Harris. (The plot shows the pattern for all polarizations and all angles.) HOW THE CBR WORKS Gerald Collins published a chapter about the CBR in the Antenna Handbook, Volume III, (Edited by Y. T. Lo, and S. W. Lee, University of Illinois-Urbana, Chapman & Hall, 1993). At this time Collins was lead engineer at Harris Broadcast Division. The wide-band flat dipoles are fed with equal currents in quadrature phase to excite the cavity in planes parallel to the dipoles. Beam width and radiated pattern are determined by the size of the cavity. On a side-mount, three-sided tower, power can be balanced to produce a directional pattern. 2

On the other hand optimizations can be applied to obtain a very uniform circular CP pattern. In the case of the Sears site, full-size range testing produced proof of performance to +/- 1.

Using it for a single 200 khz FM signal is not a big challenge and in later years Clear Channel diplexed 93.9 and 95.5 on one CBR and 103.5 and 107.5 on another.

3 On the other hand optimizations can be applied to obtain a very uniform circular CP pattern. In the case of the Sears site, full-size range testing produced proof of performance to +/- 1.5 db circularity in both H and V polarizations. The bandwidth will accommodate several analog TV channels for diplex applications. Using it for a single 200 khz FM signal is not a big challenge and in later years Clear Channel diplexed 93.9 and 95.5 on one CBR and and on another. For FM broadcast on a three-sided tower a 120 degree beam width is desired and is produced with the cavity basket about 6 feet in diameter and 2.5 feet deep. The cavity grid does the following: Isolates the dipole radiator from the tower. Shapes the pattern vs. a dipole alone. Equalizies beam widths for H and V polarizations. provides isolation from the supporting tower structure for additional isolation from the structure. Another not-so-apparent benefit of using a single bay antenna is that no vertical nulls occur. This is of value in covering the downtown Chicago area and for automotive reception in the area near Sears Tower. Half power occurs at an angle 60 degrees below horizontal. For all the above mentioned reasons this antenna type has performed quite well, and has been in service now for over 30 years. THE INITIAL STACKING PLAN This graphic shows how the original stacking plan was implemented. The following photo shows a CBR module as it was being fitted into the section between the two radome support levels (the arms with flanges). As you look at it, you will notice that the first five stations did not take the top five slots. (The apparent slot at the bottom was not usable, due to the taper of the tower section from its eightfoot face to the 12-foot cylinder which would have required extensive modification of the radome shape to fit a CBR on the taper.) Rather, three of the stations opted for more physical 3

separation to reduce cross-coupling at least until the other slots were filled.. CROSS-COUPLING In 1982 the affects of cross-coupling into FM power amplifies was just beginning to be understood.

4 separation to reduce cross-coupling at least until the other slots were filled.. CROSS-COUPLING In 1982 the affects of cross-coupling into FM power amplifies was just beginning to be understood. Before the antennas were purchased, Harris took the position that, with filtering, any stations that were 800 khz apart would have 56 db of isolation. Geoff Mendenhall even published a NAB paper using products from Broadcast Electronics to analyze 800 khz-spaced cross-coupling leakage and the resulting intermodulation products generated in the power amplifiers. (Harris also went on record that any Harris transmitter would meet the -80 db spurious emission for 800 khz spacing.) Once installed, the CBR antenna cross-coupling was measured to an isolation of 23 db to an adjacent CBR antenna and 44 db from the top slot to the bottom slot antenna. A MATTER OF MATHEMATICS A long discussion occurred. It was clear the 3 rd order product was the most likely to occur. Here is the math: The adjacent station s carrier Fc mixes with 2 nd harmonic and appears in the Power Amplifier (PA) output of F1. That is the product of interest 2(F1) +/- F2. To illustrate: 94.7 mixes with 93.9, producing an IM product at The PA is a current source at carrier and is some wild impedance at other than carrier frequency looking back into the power amplifier. The return signal is simply a standing wave of floating energy, back feeding it to the vacuum tube PA. With the adjacent antennas under power the return level on the sample port would not change if the Narda pass-through was an opened circuit (no load) or terminated into a 50 Ohm, 25 Watt load. Return leakage level was unchanged. REAL WORLD CHALLENGE In the real world, the cross-couple leakage is a non-terminated standing in the feed line of the bandpass filter/pa output. As the site built up, the return leakage could reach the 100-Watt area and RF burns were a risk from an un-filterer feed line. Likewise, when connecting test equipment one needs to aware to use high-power RF pads for active antenna cross-coupling measurements. Solid state transmitter combiners using some form of combiner with reject loads represent a better termination to the return signal and that results in better turn-around-loss from an adjacent channel signal. Since test equipment works in a 50 Ohm world, we discussed whether the specification should be -50 db or -56 db, as terminated in 50 Ohms. ANALYZING THE PROBLEM Some time after that discussion I concluded the cross-couple product from an antenna behaves as a current source. I arrived at this conclusion by using a Narda coupler at the fine matcher on a CBR antenna input as a receive antenna. 4

5 However, because a simple bandpass filter is used it can create other instabilities that are transmitter dependent. These instabilities can show up as spurious emissions not related to the site s 3 rd order products which can include VSWR shutdown for out-of-band spurious products from the PA. These instabilities will only show up when operating into the narrow bandpass filter. The same amplifier will be fully stable into a wide-band 50 Ohm load. (In other combiner designs, such as a constant impedance, 3 db coupler hybrid, the PA would be terminated in a wideband mode. REDUCING INSTABILITY Often feedline length variation between bandpass filter and the PA output feed can bring stability. It is a cut and try situation. The power amplifier needs to be examined for spurious emissions with various drive levels, gain settings, tuning, and loading to assure it is fully stable when terminated into the bandpass filter as a load. It is evident this is not an easy process. Not only do you need to observe the for the 3rd, 4th, and 5 th order products you also need to monitor any near carrier +/- 5 MHz for odd emissions while varying PA operating parameters. THE ONLY CHOICE AT THE TIME If you are wondering why was this site built using bandpass filters, the original group (in the 1982 re-build) felt having fully-isolated hardware would be more reliable than building a combiner system of the time. To locate a bank of diplexer filters on the Sears site would also require more space and trigger additional operating costs in rent, maintenance, and hardware cost. By the time you did that it would have made more sense to just have built a large combiner with a single wide-band antenna. However, the original group build-out in 1974 (and the rebuild in 1982) did not go in that direction. In our final installment on building the Sears Tower broadcast facilities, we will see how the instabilities were finally solved, and how the build-out reached the point where it is today Warren Shulz now is enjoying the retired life, after being Chief Engineer at WFYR-RKO for 15 years then at WLS AM-FM for 22 years. Now, he is out enjoying RVing and riding his ebike. When he is in, you can contact him at: wshulz@cs.com Missed the preceding parts 1 and 2? Just click here. Want to know when Part 4 is posted? Please take 30 seconds and click here for our one-time-a-week BDR Newsletter. Return to The BDR Menu 5

High Powered History

The Broadcasters Desktop Resource www.thebdr.net edited by Barry Mishkind the Eclectic Engineer High Powered History Building the Sears (Willis) Tower Site By Warren Shulz [November 2014] Whenever multiple