Dynamic range unraveled

A spectrum analyzer’s distortion and noise floor directly affect its dynamic range. Understanding the measurements you need to make for a given application can help you evaluate whether an analyzer’s dynamic range is adequate and may prevent you from buying a more expensive analyzer than you need.

Figure 1.  An RF transmitter’s function blocks have different performance specifications.

Table 1 gives an example of transmitter specifications at the block outputs in Figure 1. Typically, the signal at the output of the digital-to-analog converter (DAC) has the tightest spectral performance requirements.

Figure 2.  DANL, SHI, TOI, and noise limit a spectrum analyzer’s dynamic range.

Figure 3.  Active components such as mixers add harmonic distortion to a received signal.

For measurements at RF and microwave frequencies, the third-order and fifth-order intermodulation products of active components such as mixers and amplifiers dominate the instrument’s ACPR measurements. For an ACPR measurement to reflect the true performance of the transmitter under test, the analyzer’s distortion must be significantly lower than that of the transmitter. If the spectrum analyzer’s distortion level is 18 dB below that of the device under test, the measurement uncertainty will be less than 1 dB (Ref. 1).

The TOI slope in Figure 2 shows that for every 1 dB that you reduce the mixer’s input signal level, you reduce the TOI distortion products by 2 dB. You can achieve even lower levels at the mixer’s input by increasing the analyzer’s input-signal attenuation. Increasing the attenuation, however, directly impacts the noise floor and reduces the dynamic range. Optimal attenuator settings, therefore, involve a tradeoff between TOI and noise-floor performance.

You can use two-tone analysis as a starting point to better understand a spectrum analyzer’s performance, because you can represent wide-bandwidth signals as a summation of tones (Figure 4). Third-order intermodulation products drop in amplitude to the instrument’s noise floor from the center of the carrier to the edge of the adjacent channel.

Figure 4.  Third-order and fifth-order distortion decreases as the frequency offset from the carrier increases.

DANL—a measure of a spectrum analyzer’s sensitivity—also affects dynamic range and is typically specified at 0-dB input attenuation and normalized to a 1-Hz resolution bandwidth (dBm/Hz). Spectrum-analyzer noise forces an increase in DANL as the mixer level drops in response to greater attenuation of the signal in front of the mixer.

Using Figure 2, you can find the optimal mixer input level and, with it, the level of input attenuation. The intersection of the TOI and noise-floor lines provides the optimal mixer input level for maximizing dynamic range at about –38 dBm. Figure 4 highlights this intersection and shows the areas where phase noise, TOI, and broadband noise dominate dynamic range.

To further improve dynamic range, some spectrum analyzers apply averaging and trace math to subtract the analyzer’s noise from the measurement. Noise correction can provide up to a 10-dB improvement in noise-floor performance. Noise correction also lets you apply additional attenuation to drive the internal TOI of the analyzer down without increasing the noise floor (for TOI-limited measurements).

The phase noise of an analyzer’s local oscillator can also degrade noise-floor performance at offset frequencies close to the carrier. At offsets greater than 1 MHz from a carrier, the phase noise won’t impact dynamic range.

Figure 5.  At offset frequencies from a channel, different parameters dominate dynamic range.

Resolution bandwidth (RBW) also affects a spectrum analyzer’s noise floor. When you increase resolution bandwidth, you increase the noise floor by (10*log(RBW2/RBW1)), but you shorten sweep time. So, you can trade off dynamic range for shorter measurement time.

SHI, TOI, DANL, and resolution bandwidth combine to form a spectrum analyzer’s dynamic range. By optimizing the combination of these specifications, you’ll find the optimum point of the instrument’s dynamic range.