Comments on Commercial FM TV Standards

as Applied to Amateur Operations

by J. R. Mathison, WB9OQM

GO TO Home Page

Please see a later edition of the ARRL Handbook for basic information and definitions concerning video, pre-emphasis, and FM TV transmission systems.

CCIR Recommendation 276-2 "Frequency Deviation and the Sense of Modulation for Analog Radio-Relay Systems for Television" paragraph 1. recommends "that the value of frequency deviation without pre-emphasis in radio-relay systems for television should be 8.0 Mhz peak-to-peak referring to the nominal peak-to-peak amplitude of the video-frequency signal (see Recommendation 567). Recommendation 567 is the specifications for video and defines the "video- frequency" signal as excluding synchronizing pulses. If the video-frequency signal deviates the FM carrier 8 Mhz peak-to-peak, then the composite video (including sync) will deviate the FM carrier 140/100 IRE x 8 Mhz equals 11.2 Mhz p-p or 5.6 Mhz peak deviation. It is assumed that deviation is directly proportional to amplitude at the input of the frequency modulator.

CCIR Recommendation 276-2 in paragraph 2. recommends "that when pre-emphasis, in accordance with Recommendation 405 is used, the relative deviation of 0 db in figure 2. of that Recommendation should correspond to the value of deviation without pre-emphasis given in paragraph 1.". Figure 2. of Recommendation 405 is the video pre-emphasis frequency response curve of the pre-emphasis network. The "relative 0 db" point on the 525 line curve occurs at 762 Khz. CCIR 276-2 paragraph 2. means that through the pre-emphasis network, a 762 Khz sine wave should modulate the FM carrier 8 Mhz p-p or 4 Mhz peak.

The "relative 0 db" point at 762 Khz should NOT be confused with the pre-emphasis network "crossover" which is at 400 Khz and actually has no significance for FM TV deviation calibration procedures.

The 762 Khz sine wave is not equivalent to pre- emphasized video. See my figure 1.B. for the video waveform at the output of the pre-emphasis network. A test tone (sine wave) of 1.0 Vp-p at 2.333 Mhz and 1.0 Vp-p of composite video applied to the input of the pre-emphasis network will be exactly the same amplitude at the output of the pre-emphasis network. Note that the sync does not contribute to the peak-to- peak amplitude at the output assuming fast (125 nanosecond, as per CCIR 567) rise time from pedestal to white and fall time from white to pedestal. The 1.0 Vp-p at 2.333 Mhz test tone goes through the pre-emphasis network 140/100 IRE (2.92 db) higher than a test tone of 1.0 Vp-p at 762 Khz.

Since by definition the 762 Khz test tone should modulate the FM carrier 8 Mhz p-p (4 Mhz peak) deviation, then the 2.333 Mhz test tone -and pre-emphasized video- will modulate the FM carrier 140/100 IRE x 8 Mhz equals 11.2 Mhz p-p or 5.6 Mhz peak deviation.

Note that this is exactly the same total deviation described above without pre-emphasis except that in this case the sync does not contribute to the peak-to-peak deviation. The 100 IRE (0.714 Vp-p) of video (exclusive of sync) applied at the input of the pre-emphasis network is transformed to 140 IRE at the output. (This 140 IRE is normalized to the level of the 2.333 Mhz test tone at the network output or approximately 0.94 Vp-p). The transformed video gain is 140/100 IRE or 2.92 db which is exactly the same as the attenuation for the 762 Khz test tone through the pre-emphasis network compared to the 2.333 Mhz test tone. The selection of 762 Khz as the "relative 0 db level" on the pre-emphasis curve is an attempt to correlate video (a square wave) with the sine wave response of the network.

Pre-emphasis for FM TV is much more than a noise reduction scheme. The pre-emphasized video has its "average picture level" reduced by a factor of approximately 6.2. The net result is a reduction in the asymmetrical deviation of the video and a corresponding reduction in necessary system bandwidth at RF. The necessary bandwidth can only be calculated for sine waves. If the deviation is held constant at 5.6 Mhz peak, modulation frequencies of 762 Khz, 2.333 Mhz, and 8.0 Mhz will require a flat bandpass of approximately 16.5 Mhz, 25 Mhz, and 45 Mhz respectively at RF. Note that for a given deviation the higher modulating frequencies require greater bandwidth. The necessary RF bandwidth for a complex modulating waveform such as video can not be calculated directly.

An empirical approach is practical and in my experience an RF carrier deviation of 5.6 Mhz peak by pre-emphasized video will require a flat bandpass of approximately 25 Mhz at RF. Microwave radio equipment designed to carry the video of CCIR 567-1 and FM carrier deviation of CCIR 276-2 is usually designed for a -3 db system bandwidth at RF of 30 Mhz to achieve a pancake flat response of 25 Mhz centered on the carrier. Deviation can not occur within the full -3 db, 30 Mhz bandwidth at RF because intermodulation will occur in the non-linear portions at the shoulders of the response curve.

The receiver IF filters are normally the narrowest bandpass in the system and therefore set the system bandwidth. The bandwidth of 30 Mhz will allow headroom for 5.6 Mhz peak deviation by pre-emphasized video plus a margin of approximately 15 percent for a maximum deviation of approximately 6.5 Mhz peak before serious distortion occurs.

Those of you who are microwave radio technicians for video systems are probably already familiar with the deviation calibration procedure for CCIR Recommendation 276-2. A test tone of 1.0 Vp-p at 2.333 Mhz is applied at the input of the pre-emphasis network. The RF carrier is monitored and the deviation control adjusted for a first bessel carrier null which indicates a modulation index of 2.405. The RF carrier deviation at the first carrier null is 2.333 Mhz x 2.405 equals 5.6 Mhz peak deviation. Deviation is directly proportional to amplitude at the output of the pre-emphasis network which is also the input to the frequency modulator.

The pre-emphasized video at the output of the pre-emphasis network is the same amplitude as the 2.333 Mhz test tone and its deviation is also 5.6 Mhz peak. If a 1.0 Vp-p at 762 Khz test tone is applied to the pre-emphasis network it will be 100/140 IRE times the level of the 2.333 Mhz test tone at the network output. In other words, the deviation for the 762 Khz test tone will be 5.6 Mhz x 100/140 equals 4.0 Mhz peak deviation as per CCIR Recommendation 276-2 paragraph 2. requirements. The microwave radio equipment and deviation calibration procedure described above does not allow adequate linear deviation headroom for the addition of any subcarriers to the video. Deviation beyond the linear limits causes distortion in the demodulated signal and intermodulation between the various components of video and the subcarriers. The demodulated output of the subcarriers will in turn contain products of intermodulation which will seriously degrade their signal-to-noise ratio.

CCIR Recommendation 576-1 paragraph D.3 assumes no subcarriers on the video, not even for program audio. The chrominance subcarrier is the only one mentioned and it is considered an integral part of the "video signal". Amateurs wishing to use "commercial" FM TV standards incorporating subcarriers must look to standards other than the "European" CCIR Recommendations. Luckily these standards exist as the "North American" standards of EIA-250-C.

"Electrical Performance of Television Transmission Systems" EIA-250-C allows for subcarriers and in practical systems up to four are used. Four subcarriers may seem like more than enough but when you consider their use for program audio. intercom, control, and metering just to mention a few, they are quickly used up.

Microwave radio equipment designed to carry the video and subcarriers of EIA-250-C usually has a -3 db system bandwidth of 22 Mhz in order to obtain a pancake flat bandpass of 16 Mhz through the IF center frequency. This allows for a maximum operating deviation of 4.0 Mhz peak plus a tolerance of 10 to 20 percent before nonlinear operation occurs. Calibration is usually accomplished by applying a test tone of 1.0 Vp-p at 762 Khz to the input of the pre-emphasis network and adjusting the deviation control for a first carrier null. As before, the first carrier null indicates a modulation index of 2.405 and 2.405 x 0.762 Mhz equals 1.833 Mhz peak deviation.

The pre-emphasis network is exactly the same one used by CCIR. The 2.333 Mhz test tone and pre-emphasized video will both be 140/100 times higher than the 762 Khz test tone at the network output. As before, deviation is directly proportional to amplitude at the pre-emphasis network output and 1.833 Mhz x 140/100 equals approximately 2.6 Mhz peak deviation by the pre-emphasized video. The amplitude of each standard subcarrier is usually 10 percent (-20 db) of the composite video amplitude or 0.1 Vp-p. Four subcarriers of 6.2, 6.8, 7.5, and 8.2 Mhz each will add together in-phase approximately 600,000 times each second. Their in-phase addition will equal the sum of the amplitudes of the individual subcarriers. 1.0 Vp-p composite video plus four standard subcarriers will add up to 1.4 Vp-p at the input of the pre-emphasis network. At the output, the pre- emphasized video plus subcarriers will be approximately 1.43 times the video alone due to the placement of the subcarriers on the pre-emphasis frequency response curve. From above, the deviation by the video is 2.6 Mhz peak and 2.6 Mhz x 1.43 equals approximately 3.7 Mhz peak deviation by the video plus four subcarriers. This is just under the 4.0 Mhz peak operating linear deviation design limit for microwave radio equipment with a -3 db RF system bandwidth of 22 Mhz.

Some editions of the ARRL Handbook described a deviation calibration procedure using a test tone of 0.457 Vp-p at 762 Khz applied at the input of the pre-emphasis network and adjusting deviation for a first carrier null. Deviation for the test tone is 1.833 Mhz peak. Increasing the level to 1.0 Vp-p causes the deviation to increase by a factor of 1.0 / 0.457 and 1.833 Mhz x 1.0 / 0.457 equals 4.0 Mhz peak deviation for the 762 Khz test tone. This result is exactly the same as required for the "European" CCIR Recommendation 276-2.

Amateur FM TV operators must be careful to use an appropriate calibration procedure consistent with the system bandwidth of the radio equipment used. I am quite familiar with a television broadcasting company which bought a new microwave radio system to use for many hundreds of miles of studio-transmitter links. This was Lenkurt radio equipment incorporating all the research and development of Bell Laboratories and many years experience of the Bell Telephone system -premium microwave radio equipment with a 22 Mhz system bandwidth.

The engineer for the broadcasting company proceeded to calibrate the system for FM TV by using the test tone of 1.0 Vp-p at 2.333 Mhz and a first carrier null. The error was then compounded by adding four subcarriers which resulted in a total deviation of 5.6 Mhz x 1.43 equals approximately 8 Mhz peak deviation. There was plenty of distortion and intermodulation as this system was designed for only 4.0 Mhz peak operating deviation. The video was noisy, distorted, and had poor resolution. Video, sync, and chroma noise was heard in all the subcarriers. This premium microwave radio equipment was labeled "inferior" by the engineer and the company replaced it at considerable cost.

Amateurs using surplus IF strips with a -3 db bandwidth of 30 Mhz can accommodate subcarriers simply by lowering the total deviation. Pre-emphasized video deviation can be lowered to 4.0 Mhz peak (2.86 Mhz peak for the 762 Khz test tone) and the addition of four standard subcarriers will result in 4.0 Mhz x 1.43 equals approximately 5.7 Mhz peak deviation, well within the linear deviation limits of the 30 Mhz system bandwidth.

Calibration can be accomplished by applying a test tone of 0.64 Vp-p at 762 Khz to the input of the pre-emphasis network and adjusting the deviation control for a first carrier null. Lowering the video deviation from 5.6 Mhz to 4.0 Mhz peak should not affect the signal-to-noise ratio too much assuming substantial limiting action.

My residence in Sheridan, Wyoming is many hundreds of miles from any technical library. In my search for expert advisors, I sent inquiries to approximately 20 manufacturers of microwave radio equipment. Responses numbered only about a half-dozen from systems engineers and none claimed to be knowledgeable about FM TV. I did have the good luck to have the chairman of the Electronics Industries Association subcommitte for Television Transmission Systems as one of my advisors. This gentelman in turn was able to interview for me the elderly gentelman who had that charmanship at the time in the 1970's when the CCIR 405 pre-emphasis network was adopted into EIA-250-C.

I had hoped that someone would come up with specific references to Bell Laboratories research which may have contributed to the "North American" standards. I was told that once a recommendation became part of an industry standard that references to the recommendations sources were dropped causing its loss from the literary chain of documentation.

This article reflects ideas and opinions formed during many hours of telephone conversations with my advisors and my own 17 years experience as a technician operating FM TV microwave radio studio-transmitter links spanning hundreds of miles. Some of the ideas and experience presented here are in contrast to the popular belief about FM TV. Hopefully, any controversy resulting from this article will lead to more articles written by knowledgeable people with a background of practical experience.

Please send E-Mail to: mathison (aatt) sdf-eu.org

.

Last revision 2012-12-15