MEANINGFUL FM TRANSMITTER MODULATION LINEARITY MEASUREMENTS

Abstract

In Frequency Division Multiplex (FDM) systems, nonlinearities in the modulation and demodulation processes of the transmission system produce intermodulation (IM) products which are effectively added to the desired modulation. The effect of these added products is the degradation of data accuracy in the form of noise on the subcarrier data. Currently there are no standard test procedures or specifications that permit the prediction of the level of these IM products during system design. The characterization of transmitter modulation linearity by the measurement of IM, or cross-products, produced by simultaneous modulation by two tones, is considered ideal. This is because the test conditions can closely simulate the highest modulation level subcarriers used and the necessary demodulation equipment can he readily calibrated using common frequency modulated sources. The modulation tones used are both in the upper portion of the transmission system’s baseband and at a modulation level near the level of intended use. Measurement of only the difference frequency IM component, (f !f ), is considered 2 1 adequate for the determination of 2nd order nonlinearities. The 3rd order IM components are measured only at (2f !f ) and (2f !f ) and are normally found to be of equal 1 2 2 1 amplitudes with FM transmitters. All higher order products, as well as direct harmonics, are ignored. From the three IM level measurements, and the two desired tones, the 2nd and 3rd order modulation intercept points (IP and IP ) are determined in essentially the 2 3 same manner as the intercept point, or IP, that is common in specifying the linearity of broadband RF amplifiers. When the amplitude of the various signal and IM components are plotted on log-log scales, the desired signals have a slope of one while the 2nd and 3rd order products have slopes of two and three respectively. On log-log plots the intercept point is the modulation level at which extensions of the low level values of the IM components meet the extension of the desired modulation level. Once the IP values are determined, they may be readily used for system IM calculations. Measured IM levels in a sixteen channel FDM system compared very favorably with predicted levels using the IP values obtained from two-tone tests. The nonlinearities of the demodulator employed in the test system may be evaluated by the use of the “beat” frequency of two independently modulated FM signal sources as the required input to the demodulator. The IM products in the demodulator output in this case are due only to the demodulator’s transfer characteristics. IM product levels of the test system greater than 60 dB below the simultaneous modulation level of ±300 kHz each by 400 and 450 kHz tones are obtained at Sandia Laboratories. The use of two-tone IM tests for the evaluation and specification of FM transmitter modulation linearity is strongly recommended.