RF Highlights at NAB2003

At this year's NAB, the focus was not on debates about DTV standards or the transition, but discussions on how to make DTV transmission and reception work and improve DTV broadcast equipment.

At this year's NAB, the focus was not on debates about DTV standards or the transition, but discussions on how to make DTV transmission and reception work and improve DTV broadcast equipment.


At last year's NAB, there were a number of MSDC (multi-stage depressed collector) IOT devices in transmitter manufacturers' booths. This year, Thales reported the installation of its first Paragon transmitter at KLRU in Austin, Texas. The Paragon uses the L-3 CEA (Constant Efficiency Amplifier) MSDC IOT. Robert Symons at L-3 Communications must be very happy to see this transmitter in operation. Symons spent over 20 years developing this technology and has presented several papers on MSDC technology, starting with "Depressed Collector Klystrons for High-Efficiency UHF Television" in 1982. I reported on his work on the CEA MSDC IOT in this column as early as 1997.

Gordon Gummelt and Fred Stefanik described the Thales Paragon transmitter in the paper "New Generation Ultra-efficient DTV Transmitter Using MSDC IOT" at Sunday's engineering conference. The Paragon DTV transmitter installation at KLRU uses two of L-3's CEA IOTs to generate 43 kW. Measured efficiency of the transmitter was 58 percent-one amplifier required 38.9 kW of collector power to generate 22.6 kW RF power. This is slightly less than the 60 percent efficiency possible with a tube operating at maximum power; the L-3 CEA IOT is rated at 30 kW average RF output. The paper compared the 60 percent efficiency of the five collector L-3 MSDC IOT with three collector MSDC IOT designs-53 percent-and a two collector MSDC IOT-49 percent. A conventional single collector IOT has an efficiency of only 37 percent for DTV. The efficiency of solid-state amplifiers is even less, about 18 percent for DTV.

The Thales engineers said there is little difference in cost between amplifiers using the five collector IOT amplifier and amplifiers with less collectors. The main difference is the HV power supply, which requires multiple outputs. In the Paragon transmitter, Thales replaced the high voltage crowbar used in all earlier Thales transmitters and most other IOT-based transmitters with its "Soft Arc" technology, which limits the rise time of the current through the tube and keeps the peak current through the tube within the limits specified by the tube manufacturer. The only path the current takes is through the tube and, consequently, the total energy dissipated when an arc occurs with this system is far less than that dissipated when a thyratron is triggered, resulting in less stress on AC mains components and, with less EMF being generated, less stress on other amplifier components.

Upgrading an existing IOT-based DTV transmitter to an MSDC IOT may be possible. Since the MSDC IOT uses an oil-cooled collector and an oil-to-water heat exchanger, the existing cooling system can be used for this transmitter, as can the existing exciter and driver amplifier. A new HV power supply is required, of course. In response to a question about retrofitting the Soft Arc protection system to existing IOT DTV transmitters, the Thales presenter said there was potential for it. Applying this technology to analog transmitters will be much more difficult, if it is possible at all, due to the rapid changes in IOT collector current with changes in average picture level.

As far as I know, Thales was the first company to deliver an MSDC IOT DTV transmitter, however, other manufacturers are adopting MSDC IOT technology. On the exhibit floor, Axcera said they have taken an order for an MSDC IOT transmitter and can use the L-3 tube described above, the E2V MSDC IOT or the Thales MSDC IOT. For stations that may want to start out with a conventional single collector IOT, perhaps for analog operation, and later upgrade it to an MSDC IOT for DTV, Axcera offers a power supply that can handle the current requirements for the single collector but has the taps necessary for the MSDC tube. Itelco was showing a prototype DTV transmitter using the L-3 tube and the switching power supply it introduced a few years ago. The switching power supply allows Itelco to limit current through the tube in the event of an arc without using a crowbar. Larcan is looking at building an MSDC IOT transmitter.


When high power is not required, solid-state amplifiers offer high reliability and redundancy. Martin Horspool from Harris Corp. described how solid-state transmitters can be built so that they can be converted from analog to digital operation.

One of the tips he offered in his paper include using common-amplification of analog visual and aural carriers instead of separate amplifiers; the impact on efficiency is minimal. A 10 kW solid-state common amplification transmitter has an efficiency of 27 percent where an externally diplexed transmitter has a combined efficiency of 27.5 percent. With common amplification, a mask filter is required to reduce out-of-band intermodulation products. A mask filter is also required for DTV.

Of interest to anyone planning to install an analog transmitter (tube or transistor) that may be upgraded to digital on the same channel, Harris found that a mask filter designed for DTV could also be used for analog on the same channel-this makes converting from analog to digital easier. Power supply requirements are more stringent for analog transmitters than digital since they have to handle current fluctuations caused by average picture level varying at video rates. Load regulation is less critical for DTV transmitters. The optimum amplifier drive levels and biasing will be different for analog and digital operation. Harris uses an on-board micro-controller to optimize its broadband amplifier parameters for different modes of operation.


Earlier this year I devoted two columns to distributed transmission systems, also called single frequency networks. At NAB, David Hershberger, principal engineer at Axcera, delivered an excellent paper he authored with Michael Pugh titled "Implementation of the ATSC Distributed Transmission System." The paper presented a detailed explanation of the ATSC candidate standard for distributed transmission. I do not have space to do justice to it here, but the paper is in the NAB2003 Broadcast Engineering Conference Proceedings. Once the standard is adopted by ATSC, if there is interest I'll devote a column to a more detailed explanation of distributed transmission transmitter technology.

Hershberger pointed out in the paper that when the RF watermarking feature of the standard is used, adaptive precorrection systems have to be modified so it won't be removed. The RF watermarking is used to identify individual transmitters in a single frequency network and is inserted at low level into the 8-VSB signal. On an eye chart, the watermark appears as a small eye opening in between the large eye openings. On a constellation diagram, the watermark can be seen as a widening (or splitting) of each VSB level. Photo 1 shows a constellation diagram from the Axcera demonstration at NAB showing the splitting of the VSB levels with the watermark data.

(click thumbnail)Photo 1: Constellation diagram from Axcera's demonstration at NAB.
Note that the level of the watermark signal was much greater than what would normally be used. It was increased to make it easy to see how the extra data affects the signal.

Photo 2 shows David Hershberger standing next to his Axcera distributed transmission system demonstration using the DXA2B distributed transmission adapter, two modified DT2B 8-VSB exciters, a Rohde and Schwarz EFA-53 analyzer and two DTV receivers-one an off-the-shelf set-up box and the other a prototype LINX VSB receiver with increased equalization capability.

(click thumbnail)Photo 2: Axcera's David Hershberger with the company's distributed transmission system demo at NAB.The distributed transmission adapter from Axcera will cost about $50,000, since the changes required in the exciters can be handled by software in the Axcera DT2B without additional cost. This equipment is available now, although until the FCC adopts rules for it, any operation would have to be under an experimental license.

Using the demonstration setup, Hershberger showed the LINX receiver was able to handle a 0 dB echo (from the other modulator) with timing differences up to 12 microseconds. The legacy set-top box could handle a 2 dB desired-to-undesired (D/U) ratio. He said fifth generation DTV receivers were also able to work with a 0 dB ratio.

In the paper, David Hershberger noted, "it is in the interest of all parties to develop an interoperable standard." He said that it should be possible to use a distributed transmission adapter from one manufacturer and slave transmitters from two different companies and still have everything work correctly. A Harris representative told me they're actively working on adding distributed transmission network technology to its 8-VSB modulators. A Larcan representative stated that they are committed to the low-power TV and translator market but are waiting for FCC action on the standard.

In many cases where on-channel repeaters are needed to fill in shadows in coverage, the fine level of control offered by the ATSC candidate standard for distributed DTV transmission may not be needed. K-Tech offers regenerative DTV translators that can retransmit the ATSC bitstream without re-encoding it. Since the data stream coming out of the unit is the same as the bitstream going in, the modulator controls in the ATSC standard are not required as long as the unit is used as an off-air booster. GPS is used to lock the frequency and symbol rate so errors do not accumulate with multiple hops. There are trade-offs, of course. While delay through the K-Tech on-channel regenerative DTV translator can be reduced to as little as 1 microsecond, performance suffers with this short of a delay. If the delay is an issue, Steve Kuh of K-Tech suggested the simplest approach for broadcasters wanting to use low-cost DTV translators to fill in coverage holes is to use different channels, rather than the same channel, in the system. In the right circumstances, a few channels can be re-used at different points in a long string of translators.

One thing was clear from NAB2003. While some parts of DTV transmission are maturing, new technologies are coming on the market to make transmission less expensive (MSDC IOTs, for example) and more flexible (distributed transmission, DTV translators). While I didn't discuss it this month, manufacturers are also making DTV modulators easy to set up and monitor.

Next month I'll wrap up my coverage of RF at NAB2003 with a look at DTV reception, RF propagation and DTV test and monitoring. As always, comments are welcome. Drop me a note at dlung@transmitter.com