How to optimize dynamic range for distortion measurements?

One issue associated with signal measurement is the ability to distinguish between a large fundamental signal and smaller distortion products. The dynamic range of a spectrum analyzer specifies the maximum range within which the spectrum analyzer can distinguish between signal and distortion, signal and noise, or signal and phase noise.

AnswerOne
issue associated with signal measurements is the ability to distinguish between a large fundamental signal and smaller distortion products. The dynamic range of a spectrum analyzer specifies the maximum range over which the spectrum analyzer can distinguish between signal and distortion, signal and noise, or signal and phase noise.

When measuring signal and distortion, the mixer level determines the dynamic range of the spectrum analyzer. The mixer level that optimizes the dynamic range is determined by the spectrum analyzer's 2nd harmonic distortion, 3rd order intermodulation distortion, and displayed average noise level (DANL). These specifications can be used to plot the internally generated distortion and noise versus mixer level.

Figure 1 shows the -75 dBc 2nd harmonic distortion point at a -40 dBm mixer level, the -85 dBc 3rd order distortion point at a -30 dBm mixer level, and the -110 dBm noise floor at a 10 kHz RBW. The slope of the 2nd harmonic distortion line is 1 because the SHD increases by 2 dB for every 1 dB increase in the mixer fundamental level. But since distortion is determined by the difference between the fundamental and the distortion components, it changes by only 1 dB. Similarly, the slope of the plotted third-order distortion is 2. For every 1 dB change in mixer level, the third-order component changes by 3 dB, or 2 dB relative. The maximum second- and third-order dynamic range is obtained by setting the mixer level where the second- or third-order distortion equals the noise floor, and the corresponding mixer levels are marked in the figure.

Figure 1. Dynamic range—distortion and noise.

Dynamic range must be increased by narrowing the resolution bandwidth. As shown in Figure 2, the dynamic range increases when the RBW setting is reduced from 10 kHz to 1 kHz. Note that the increase is 5 dB for second-order distortion and 6 dB for third-order distortion.

Figure 2. Dynamic range improvement with reduced resolution bandwidth

A final point is that the dynamic range of intermodulation distortion is affected by the spectrum analyzer phase noise, since the frequency spacing of the different spectral components (the measured spectrum and the distortion products) is equal to the spacing of the measured spectral lines. Figure 3 shows the noise curve obtained with a 1 kHz resolution bandwidth for the measured spectral lines spaced 10 kHz apart. If the phase noise at 10 kHz is only -80 dBc, then for this measurement, 80 dB becomes the ultimate limit of dynamic range, rather than the maximum 88 dB dynamic range shown in Figure 3.

Figure 3. Phase noise limited third-order intermodulation product test.