When designing a new DTV transmission facility or planning DTV conversion of an existing analog facility, in addition to selecting a transmitter with sufficient power and an antenna that has the correct tradeoff between gain and elevation pattern, be sure to consider the transmission line.
Power handling capability and loss have to be considered for any design, but don't forget to look at VSWR (voltage standing wave ratio) and group-delay performance.
When reusing an existing transmission line run for a DTV operation, make sure it can handle the maximum DTV transmitter power. Systems designed for analog TV will often show the rated power for the line as peak visual power with 10 percent (sometimes 20 percent) aural power.
DTV power is specified as average, not peak power, and as a result, the maximum average DTV power the line can handle will be significantly less than the maximum peak visual power. For example, at Channel 39, using the Dielectric DASP program (available at www.dielectric.com/broadcast), we see a 1,000-foot run of 3 1/8-inch 50 ohm "digitLine" rigid transmission line can handle 29.9 kW peak visual power but only 21 kW average DTV power.
Both of these ratings are before considering any derating factor. Dielectric's DASP program adds a 0.9 derating factor for DTV, which reduces the maximum average DTV power to 18.9 kW.
It isn't sufficient to stop at the catalog specifications for transmission line power ratings. Look at the fine print! Dielectric's DASP program notes that maximum power ratings are based on a VSWR of 10:1, no line pressurization, an ambient temperature of 40 degrees Celsius (104 degrees Fahrenheit) and a maximum inner conductor temperature of 120 degrees Celsius (248 F).
While the temperature ratings are valid for most tower sites, if the transmission line runs in an enclosed space in a building with other lines, the ambient temperature could be much higher.
Solar radiation has an impact. Myat ( www.myat.com ) recommends using a derating factor of 0.83 for middle latitudes and 0.66 for tropical latitudes to cover the temperature increase from solar radiation.
I have yet to see an antenna with a VSWR of 1.0:1 across the channel, and an antenna VSWR of 1.08:1 is not unusual. Higher VSWRs will reduce power handling capability. It makes sense to pressurize transmission lines, if only to keep out moisture. It is also an easy way to increase the power handling capability of transmission lines, but if pressurization should fail and input power is not reduced to the rating without pressurization, the line could be damaged.
In such cases, it is essential to monitor transmission line pressure, and if the site is not monitored 24/7, there should be some provision to automatically reduce power to a safe level if pressurization fails.
WHAT ABOUT VSWR?
VSWR is an indication of mismatches in the transmission line system that cause some power to be reflected to the generator and thus create standing waves on a transmission line. A mismatch at the antenna can cause power at certain points on the transmission line to exceed the manufacturer's ratings.
For NTSC systems, our main concern was VSWR at visual, aural and chroma subcarrier frequencies, in that order, where power was concentrated. The DTV signal is broadband and noise-like, so it isn't sufficient to optimize VSWR at one or two frequencies.
Dan Barton, an RF engineer at Alive Telecommunications ( www.alivetele.com ) who has worked on many broadcast antenna systems, told me he finds broadcasters tout the VSWR performance of the antenna, but often ignore the transmission line, which can have a significant impact on system VSWR. The number of elbows in the transmission line and the length of line sections will affect system VSWR. At UHF, tuning elbows can reduce their impact on system performance. Manufacturers can tune elbow complexes for best VSWR performance at a given channel.
If you are planning to use an existing rigid transmission line run on a different channel, be sure to check the length of the line sections. While manufacturers have made improvements in inner conductor connections and the Teflon insulators at the end of rigid transmission line sections, they won't exactly match the characteristic impedance of the transmission line. As sections are combined, these variations in impedance at the rigid line flanges will add and create high VSWR at certain frequencies. See the Myat University technical page at www.myat.com for information on how to calculate these critical frequencies, and for a table showing the appropriate channels for 20-foot line sections, 19 3/4-foot line sections and 19 1/2-foot line sections and a listing of critical frequencies for other line sections.
I like waveguides. No inner conductor to burn up. Plenty of power handling capability and low loss. I've installed many waveguide systems, and the only problems I've had with them, except for one undersized coupler, have been at transitions between rectangular and circular waveguides and, more common, the transition between rectangular waveguides and coax.
If you are planning to use a waveguide for DTV, especially on a lower UHF channels, there could be a problem--group delay.
Six years ago, Henry Fries at Thales (which was then called Thomcast) sent me a table he created showing group delay for different lengths and types of waveguide transmission lines.
There is general agreement that group delay less than 20 nanoseconds will not cause noticeable degradation of an 8-VSB signal. Using that as the criteria, a WR1500 rectangular waveguide should work fine above Channel 35 for line lengths less than 2,000 feet. For 1,000-foot lengths, performance should be fine above Channel 24.
He found that a DTW1500 doubly truncated waveguide is okay above Channel 41 for a line length of 2,000 feet or less. With a line length of 1,000 feet or less, group delay is less than 20 nanoseconds above Channel 29. Using a GLW1500 circular waveguide, lengths of 2,000 feet or less are fine above Channel 49; and 1,000 feet or less for Channel 39 up. See Table 1 for more examples. You will notice that the group delay for a given channel increases as the line shape transitions from rectangular to circular.
If you are now looking out the window at a 2,000-foot run of 15-inch circular waveguide you planned to use for DTV on a Channel 45, and worrying it won't work, relax. First, for GLW1500, the group delay exceeds the recommended maximum by less than 5 nanoseconds. Second, there is a way to correct for it in the transmitter!
At the 2002 NAB Broadcast Engineer Conference, four engineers from Harris Corp. (A.S. Mattsson, D. Danielsons, C. Abascal and J. Seccia) presented a paper titled "New Adaptive Techniques for Correcting VSWR Distortion in DTV Transmitters." The paper focused primarily on correcting for distortion due to VSWR, but group delay is also discussed.
The Harris Apex DTV exciter can dial in group delay precorrection based on the channel, type of transmission line, and length. I'm not aware of any other manufacturer offering this capability, but it would seem easy to add, even if the group delay correction had to be manually calculated. After all, it shouldn't change!
For more information on the impact of group delay on DTV transmission, see my January 2002 RF Technology column, "IEEE Broadcast Symposium Part 2," available at www.tvtechnology.com
I welcome your questions and comments. Send them to me at dlung@transmitter.com.
