Last month I offered some suggestions on using low-power DTV transmitters to meet FCC deadlines for an interim DTV build-out with less expenditure of time and money. This month I'll offer some tips on choosing a low-power DTV (LP-DTV) transmitter. I'll also look at a technology for improving the efficiency of high- power IOT amplifiers that's finally making it into products after several years of development.
LOW-POWER DTV TRANSMITTERS
At NAB last year, Steve Kuh from K-Tech asked me if I thought broadcasters would be interested in his new low-power (100 W average) DTV transmitter. I said I thought broadcasters in smaller markets that couldn't afford to build out a high-power DTV installation would be very interested in it, if the price was right.
Interest in low-power DTV increased significantly when the FCC released Order 01-330, allowing many broadcasters to begin DTV operation at reduced power without losing protection for authorized higher-power DTV facilities. As a result, at NAB2002, K-Tech found a lot more competition in the LP-DTV transmitter business. This year, most transmitter manufacturers had low-power DTV transmitters in their booths.
One of the first parameters to consider when selecting a transmitter is output power. I've covered the power versus coverage trade off in previous articles and won't repeat it here. My analysis of DTV location options last month was based on a transmitter output power of 1 kW. Most transmitter manufacturers now offer 1 kW output DTV transmitters in the $100,000 to $150,000 price range, with some of the more expensive packages including an ATSC encoder and a small 8-bay slot antenna such as the Andrew AL-8, Scala SL-8 or Dielectric DL-8.
All of the manufacturers of LP-DTV transmitters I visited at NAB offered the ability to start at one power level and add modules later to increase power. For example, Harris' Ranger series DTV transmitters uses the same 460 W LDMOS modules used in its DiamondCD high- power solid-state transmitter. You can start with a 460 W transmitter, add a second module to get to 900 W and use both modules in a high power DiamondCD transmitter. In its Affinity LP-DTV transmitter, Thales used LDMOS amplifiers originally designed for telecom transmitters. Power levels start with a DTV modulator/exciter at 50 W, which can be upgraded to 100 W, 200 W, 500 W or 1 kW by adding amplifier modules. Larcan offers LP-DTV transmitters at 150 W (DTT150) and 500 W (DTT500). K-Tech's XMT-100 is normally configured for 100 W output, but amplifier modules can be added or removed to obtain power levels from 25 W to 1 kW.
As you can see, there is no shortage of LP-DTV transmitter options. To narrow the selection, it is important to consider the transmitter's ultimate use. One thousand watts is not likely to be enough for a full-power DTV transmitter installation. Most of the DTV facilities I've designed have required transmitter powers from 15 kW to just over 40 kW to make the maximum FCC-authorized effective radiated power (ERP). While 15 kW may be a practical output power for a solid-state transmitter, higher-power solid-state amplifiers will be expensive to purchase and, as I'll show later, expensive to operate. If your plan is to use the LP-DTV transmitter as a backup, translator or booster in the future, you can skip to that section now. If you plan to keep the LP-DTV and upgrade it for higher power, read on.
Anyone who bought one of the first solid-state analog transmitters knows how soon solid-state amplifier devices can become obsolete and difficult to obtain; designs change quickly. My recommendation is that if you buy an LP-DTV transmitter with the intention of upgrading it in the future, you obtain a commitment in writing from the transmitter manufacturer that they will be able to supply the upgrade modules for a specified price when you need them.
If you know you will require a high-power tube amplifier to meet your FCC ERP requirements, consider buying an LP-DTV transmitter that can be used as a driver for an IOT amplifier. Conceivably, any of the LP-DTV transmitters in the 100 W to 1 kW power range could be used as a driver for a tube amplifier. However, correction requirements for tubes are usually different than those for solid-state amplifiers. In addition, protection circuitry is required for the IOT amplifier. If the LP-DTV transmitter is to be used as a driver for a tube transmitter, the safest option is to buy it from the same company that will supply the IOT amplifier. Axcera is one company that will sell you an LP-DTV transmitter in the 250 W - 1 kW range that can be later used as a driver for its Diacrode or IOT-based high-power amplifiers. Others offer similar options. As with solid-state amplifier upgrades, have the manufacturer commit to the availability and price of the upgrade before you buy the LP-DTV transmitter.
Another item to evaluate when choosing an upgradeable LP-DTV transmitter is the quality of the DTV exciter and its ability to correct for both linear and non-linear distortions. Harris and Thales use the same DTV exciters in the Ranger and Affinity as they use in their full-power DTV transmitters. Both exciters offer adaptive correction capability and the ability to add 8VSB monitoring software - "CD-Eye" and "Scout" respectively. Some manufacturers use simpler exciters with less correction and monitoring capability with their LP-DTV transmitters. Even if the LP-DTV transmitter will be replaced with another high-power transmitter later, a high-quality full-featured exciter may be useful either as the main exciter in the new transmitter or as a backup.
EFFICIENT HIGH-POWER DTV TRANSMITTERS
Earlier, I mentioned that the UHF DTV stations I'd been working with needed transmitter powers from 15 kW to slightly more than 40 kW to make their FCC-authorized maximum ERP. So far, tube amplifiers have been the least expensive way to obtain power levels in this range. While solid-state amplifier costs have decreased, tubes are likely to remain the device of choice for high power.
I wasn't expecting any breakthrough high-power transmitter technology at NAB2002. This year's surprise was the interest in multistage depressed collector (MSDC) IOTs for high-power UHF DTV. MSDC technology is not new - MSDC klystron TV transmitters appeared in the late 1980s. Robert Symons, at Litton Electron Devices Division (now part of Northrup Grumman) presented a paper on MSDC IOT technology at the NAB Broadcast Engineering Conference in 1997. Marconi Applied Technologies (which reccently changed its name to E2V) has been working on MSDC IOT technology for years as well. Litton/Northrup Grumman calls its MSDC IOT a "constant efficiency amplifier" or CEA. Marconi uses the term "Energy Saving Collector" or ESC to describe its MSDC technology.
In his paper at NAB97, Bob Symons pointed out that the efficiency of an IOT varies as the square root of the power output. An 8VSB signal has a peak-to-average ratio of approximately 4:1. Most of the time the amplifier is using only one-fourth of its peak power capability, which limits its efficiency to under 30 percent.
The CEA, on the other hand, is able to achieve efficiencies of more than 50 percent through the use of the multiple collectors. I'll describe the technology in detail in a future article. If you can't wait, visit the Litton Electron Devices Web site Constant Efficiency Amplifier page at www.littonedd.com/broadcast_products/cea_articles.html for papers describing the technology and its implementation.
Acrodyne and Thales are two of the companies I noticed at NAB displaying CEA or ESC amplifier transmitters. While I didn't have a chance to visit Astre at NAB, its Web site at www.astresystems.com/indexmain.html has details on the CEA, including a table showing the savings in utility costs for the CEA versus a standard IOT at various DTV power levels. At an average output power level of 20.6 kW it shows the CEA will save $20,978 per year if electricity costs $0.10 per kilowatt-hour.
Thales will be offering CEA and ESC tubes in its new Paragon DCX DTV transmitter. Using the Litton CEA tube, Thales engineers reported obtaining efficiency twice that of solid-state amplifiers and 40-50 percent greater than that of conventional IOT amplifiers. The Paragon DCX does not require a crowbar.
Marconi's ESC-IOT has three collectors, compared with five in Litton's CEA IOT. While this reduces the efficiency, it also reduces the complexity of the high voltage power supply. The ESC-IOT can be used in the same cabinet as its existing 30 kW IOT after installation of a 12.5 kV, 2.5 amp depressed collector power supply. Marconi said the efficiency improvement is 15 percent better than that of a conventional IOT operating at 38 percent efficiency.
After looking at the efficiency improvement, it seems like switching to a CEA or ESC amplifier for DTV is a "no-brainer." However, as anyone that has had experience with MSDC klystrons will tell you, extra collectors bring extra complexity.
The increase in HV supply complexity is minimal with the three-collector ESC. The five-collector CEA will require a new HV beam supply. My experience with the MSDC klystron in an analog transmitter was that is was very difficult to correct luminance non-linearity to less than four or five percent because linearity was influenced by the amount of current flowing in each collector, which was determined by the APL. Any change in operating parameters required a change in correction. I've been told that during tests at Thales, the adaptive pre-corrector in its DTV modulator had no problem correcting for the CEA tube. This makes sense because the average power of an 8VSB signal is more stable than the average power of an NTSC signal.
Cooling the multistage collector is also an issue. The MSDC klystron required pure or de-ionized water, which tended to react with metal in the tube. While pure-water cooling systems are commonly used with high-power HF and MF transmitters, I've heard of tube problems at a few MSDC klystron sites where the transmitter engineer was not careful with cooling system maintenance. The Litton CEA offers a unique solution to the problem. The CEA IOT can be oil cooled; a simple plate heat exchanger is used to transfer the heat to a conventional glycol/water cooling system. The use of oil avoids many of the problems associated with pure-water systems.
Is the CEA or ESC right for your station? Perhaps. Many stations have chosen to use solid-state amplifiers at high-power levels, even though they are substantially more costly to purchase and less efficient. These stations feel that solid state amplifiers reduce the need for maintenance, increase reliability and avoid the need to have a transmitter engineer skilled with and comfortable working with high- voltage, high-power tubes. These factors offset the initial cost and efficiency issues. Similar arguments may make the conventional IOT look more attractive than the CEA or ESC, even though the differences are not as extreme between the two tube technologies as they are between tubes and solid state. Transmitter manufacturers have come a long way in making tube amplifiers easier to maintain and service, but it will be another year or two until we know how the new MSDC IOT based amplifiers will work long-term. As more data on the performance of the new tubes becomes available, I'll report it here.
