Technical Note BC-01

Pre-emphasis, Low-Pass Filtering & Why It's So Important


Pre-emphasis is related to maximum frequency by the modulation index, which for FM broadcast in this country, is arbitrarily set at 5 by FCC regulations.
Pre-emphasis is required because noise interference may produce indirect FM by shifting the phase of the carrier; high audio frequency noise produces more equivalent FM than does lower frequency noise. Because signals in the higher-end of the audio spectrum usually have less energy than lower frequency signals, the uncompensated transmitter has rather poor signal-to-noise at the higher frequencies. Pre-emphasis was introduced to correct this condition by exaggerating the high frequency response during transmission and attenuating it by the same amount during reception. Another way to look at the reason for preemphasis and an upper limit on audio response is by working out the formula for modulation index, or deviation ratio:

            Max Deviation
          ------------------ = modulation index
          Highest audio freq

Since the index is arbitrarily limited to 5, we can see how the upper limit of 15khz is derived:

          ------------ = 5
            15,000 hz

The need for an input filter is always present, although much more important than ever when broadcasting a stereo signal. Program material rich in harmonic overtones can easily exceed 15khz and heterodyne with the 19khz pilot tone. Worse than that, it can heterodyne with the lower sideband of the L-R subcarrier modulation, which moves closer to overlapping as signal frequency increases. For this reason, a SHARP cutoff audio filter is required after 15khz. How steep is the rolloff? Well, imagine a "brickwall" filter that begins to roll off and is down 3 db at 15.5khz and then is down more than 40 decibels at 19khz. An 8-pole filter, supplemented by a resonant band reject circuit is suited for this. If such a filter were omitted, frequencies could risk being high enough to overlap. Since the subcarrier spectra is both the sum and the difference of the audio frequency against the 38Khz subcarrier, the higher the modulating frequency, the lower the bottom or left sideband on the chart below would extend, and the higher the upper sideband would extend. Eventually, as the frequency goes high enough, the actual program tone and the lower sideband would overlap with the pilot tone and eachother. The upper sideband may interfere with SCA (Subsidiary Communications Authority) services at 67Khz and higher.
Spectral Energy
Removal of audio program frequencies above 15Khz. is also important in mono transmissions because of carrier sideband generation. The carrier frequency plus the sum and the difference of the audio frequency that is modulating it represents the first-order sidebands. From that point outward, exist additional orders, each spaced further out by the same interval of the modulating frequency, but at diminishing amplitudes. The higher the modulation frequency, the wider the spacing of intervals. And when modulation index of 5 is reached, the order of sidebands rises to 8 on either side of the carrier. If there were no upper limit on audio frequencies allowed to modulate the transmitter, these sidebands would extend into adjacent channels, causing interference.

Please make every effort to be a responsible broadcaster. The more knowledge you have, the better your capabilities to reach this standard and very important goal.

Authored by your friendly "Peg-legged" Bovine One