Tie Up Loose Ends Before Making New Resolutions

Before I make any New Year's resolutions, I want to tie up the loose ends that accumulated in 2005. I've covered most of the details of measuring DTV TPO (transmitter power output). Those loose ends involve the proper use of a spectrum analyzer as the power output measurement instrument. I've always said that a spectrum analyzer can mislead the unwary user more ways than a cat has lives:

1) By now you know that your choice of resolution bandwidth strongly affects the indicated power within the channel. The analyzer's RBw (resolution bandwidth) is small compared to the DTV signal bandwidth--5.38 MHz. As the DTV signal power is evenly distributed across its 5.38 MHz of bandwidth, if the RBw were 0.538 MHz, it would read 10 dB low.

Your range of choice of RBw was determined by the manufacturer. The smaller the RBw, the longer it takes to analyze the power distributed across the channel, but the more detail in the resulting spectrum plot.

This gets really important near the edges of your channel. Try this: Set RBw to 1 MHz and observe your DTV signal. You might panic when you see a lot of signal energy outside of your channel, and try as you might to reduce this out-of-channel energy, your analyzer is giving you a false impression. Now reduce RBw to, say, 0.1 MHz and you will find that your DTV signal fits nicely inside your channel. Note that a DTV signal is quite unlike NTSC in this respect.

But there is a little more to it. The frequency selectivity of your spectrum analyzer is determined by a real-life filter, and its frequency selectivity is not rectangular but like that of a tuned circuit, sort of Gaussian.

There is a correction factor for the RBw to be added to (10 dB in the example above).

Not all spectrum analyzers have the same selectivity characteristic, even if they provide the same choices of RBw. Having said that, I cannot tell you what the correction factor for your instrument is, but the manufacturer can. It is in the manual. It is a small number and is often overlooked, but when measuring your TPO, the devil is in the details, and you know how the FCC is about details. I suggest as a New Year's Resolution, you make up a table of correction factors based on the manufacturer's specification for your instrument over its range of RBw, taking into account its RBw correction factor and keep it where it would be hard to ignore.

2) But before you adopt this suggestion, there is a second correction factor you need to know about. This correction factor is due to the way your spectrum analyzer responds to a noise-like signal such as DTV compared to the way it responds to something in the analog world. Again, this correction factor can be found in the instrument manual, but it relates to the way your spectrum analyzer demodulated the noise-like signal to baseband. It is a small number, but should not be ignored in making TPO measurements.

I want to thank Mr. John Tremblay of Larcan for his informative booklet "Digital Translator Handbook," which so clearly explains these fine points. Finally, I must repeat my admonition that the real limit to your TPO is the sideband splatter being generated in your DTV transmitter. You most keep your TPO below the level at which your sideband splatter is at or less than the limit set by the DTV RF mask. I understand that Gary Sgrignoli and many others are working on an IEEE standard for making these compliance measurements. I don't know when the IEEE will approve and publish this, or if it will be the last word on compliance. Remember that the most important component of this sideband splatter is third-order intermodulation products. These increase 3 dB per 1 dB increase in TPO, so only small reductions in TPO change the sideband splatter significantly.

24/7 DTV EAS

Since my Oct. 21 column was published, Congress has started to take action to deal with the subject of 24/7 DTV EAS in all media. In fact, there is an appropriation in the works to fund the development of such a mandated EAS. I am very pleased that this step has been taken. I also learned that within the Enforcement Bureau of the FCC, this matter is receiving the attention it deserves. I believe this topic will be on the agenda of the Feb. 7 ATSC committee meeting. I was honored to be invited to attend this meeting, but alas, by then, I will be back in Oregon.

Last summer, my wife and I were at the Oregon seacoast when the tsunami alarm system was tested. A previous test failed; this one was declared successful. We were there and we never heard those sirens, and no one I met heard them either. A tsunami is caused by an earthquake beneath the ocean. If there were a way to detect the undersea earthquake, then the warning of a possible tsunami could be given much earlier, hours earlier. That would permit evacuation of coastal communities. The December issue of IEEE Spectrum featured an article on the ability to detect impending earthquakes electronically. Note I said impending, perhaps a few days even before the temblor that may launch a tsunami. The article reports that electromagnetic waves are radiated in the extremely low-frequency portion of the radio spectrum before earthquakes. These travel over great distances almost instantly to where they can be monitored and analyzed by computers.

For the first time, we may be able to detect a tsunami before the sea wave is launched. This development makes it all the more timely that we organize a 24/7 EAS for all media, the most universal of which is DTV terrestrial broadcasting. Broadcasters hold the key to our being able to evacuate threatened areas. If I couldn't hear those sirens last summer, I wouldn't be able to hear them when there is a tsunami threat. EAS should not have to depend on air raid siren technology of the 1940s but on the universality of digital broadcasting.