A swept-frequency superheterodyne spectrum analyzer uses a mixer to convert the input signal to the intermediate frequency (IF), where it is amplified, filtered, and detected. Pre-selection filters (sometimes low-pass filters) are primarily used to filter out signals at mirror frequencies. The reference level displayed on the spectrum analyzer screen is related to the gain of the IF amplifier, which only adjusts the vertical position of the signal on the screen and does not affect the level at the input attenuator. The horizontal axis of the screen represents frequency, and the vertical axis represents the measured signal level, typically expressed as a linear voltage (Volt) or logarithmic dB. The amplitude accuracy of a spectrum analyzer is usually categorized into absolute accuracy and relative accuracy. Absolute accuracy refers to the power level accuracy of the signal, measured in dBm; relative accuracy refers to the accuracy of measuring the difference between two signals, where one signal serves as a reference for the other. For example, when measuring harmonic signals, the power ratio of the harmonic to the fundamental frequency is typically measured. Both types of accuracy can be improved by measuring a calibration source with highly accurate amplitude and frequency. The signal processing components at the front end of the spectrum analyzer, such as amplifiers, filters, and mixers, are all sources of amplitude measurement error. In many spectrum analyzer designs, using better components can improve accuracy. Agilent Technologies' high-performance PSA series spectrum analyzers (as shown in Figure 2) employ a complete set of digital intermediate frequency (IF) filters, which can avoid amplitude variations inherent in analog IF filters. However, simply improving some components in the entire signal processing chain is insufficient to eliminate all sources of error. A better understanding of the interactions between the various modules of the spectrum analyzer helps reduce errors and improve amplitude measurement accuracy. Why is amplitude measurement accuracy so important? For example, some communication standards require that the modulated carrier power cannot exceed a certain value, which imposes requirements on absolute accuracy; excessive harmonics or spurious signals can interfere with other communication systems, which also imposes requirements on relative accuracy. Amplifiers in these systems must meet specific linearity requirements to ensure that high harmonics and spurious signals are not generated. For filters in these systems, both passband and stopband characteristics must be measured simultaneously. The interactions between various spectrum components are one source of error. Table 1 lists some sources of amplitude measurement error. Most instrument manufacturers specify both absolute and relative uncertainty in their product specifications. Because relative uncertainty affects both measurements, this article will focus on discussing relative uncertainty. The frequency response flatness of a spectrum analyzer is one of the main sources of amplitude error. This index describes the functional relationship between relative amplitude uncertainty and frequency, and is affected by the frequency response flatness of the input attenuator, mixer, local oscillator amplitude, and input filter. Frequency response uncertainty is generally expressed in both absolute and relative ways. Relative uncertainty describes the maximum possible amplitude uncertainty relative to the center frequency across the entire frequency range, and is generally smaller than the absolute uncertainty for the same frequency band. However, to obtain the frequency response uncertainty of a relative amplitude measurement within a specific band, the relative frequency response index value must be multiplied by two to reflect the peak-to-peak value of the frequency response across the entire band, which usually results in it being higher than the absolute frequency response index. Spectrum analyzers typically use YIG tuned filters as pre-selection filters, and the YIG filter also affects the frequency response characteristics of the spectrum analyzer. This filter must be precisely tuned and aligned to avoid introducing additional frequency response variations. Due to the limited sweep speed of the local oscillator, the YIG filter also requires some delay and compensation to ensure that its center frequency is synchronized with the local oscillator. A low-pass filter is typically added to the front end of a spectrum analyzer. When measuring signals at low frequencies (usually below 2 GHz) that cannot be reached by a YIG preselector, this low-pass filter is used to filter out high-frequency signals. Although this filter also affects the overall frequency response, its impact is much smaller than that of the YIG filter. Because some spectrum analyzers use harmonic mixing technology, the instrument actually has many internal mixing bands, each with a specific frequency response. Therefore, switching between these bands introduces uncertainty. For example, the PSA E4440A spectrum analyzer, which operates up to 26.5 GHz, has five internal mixing bands: 3 Hz to 3 GHz, 2.85 to 6.6 GHz, 6.2 to 13.2 GHz, 12.8 to 19.2 GHz, and 18.7 to 26.5 GHz. When the set frequency span exceeds two mixing bands, the instrument automatically switches internal mixing bands, thus introducing amplitude uncertainty. When measuring the relative values ​​of two signals in different mixing bands, the total uncertainty equals the sum of the frequency responses of the two bands plus the band-switching uncertainty. If the band-switching uncertainty is not specified in the specifications, the total measurement uncertainty for each band can be determined using the absolute frequency response parameters referenced to the calibration source (see Table 1). Another source of uncertainty in spectrum analyzers is the reliability of the range. When measuring two signals located at different vertical positions (ranges), the reliability of the range will affect the results. The linearity of the detector and ADC, and the linearity of the logarithmic/linear amplifier, all affect the reliability of the range. For most logarithmic amplifiers, the linearity deteriorates as the input amplitude decreases. For two signals with similar amplitudes, the range uncertainty is about a few tenths of a dB; for signals with significantly different amplitudes, this uncertainty can reach 2 dB. Typical range reliability specifications are ±0.4 dB/4 dB with a cumulative maximum of ±1.0 dB. The ±0.4 dB/4 dB specification is applicable to signals with similar amplitudes, while the cumulative specification is applicable to signals with significantly different amplitudes. When a spectrum analyzer needs to measure signals of different levels, its flexibility can be achieved by adjusting the reference level; however, adjusting the reference level also introduces uncertainty. The reference flux and input attenuator are related to the intermediate frequency (IF) gain, and their range can be adjusted from the Display Average Noise Floor (DANL) to the maximum input level it can withstand. Adjusting the reference level is essentially adjusting the gain of the IF amplifier, which, like all amplifiers, has a gain that varies with amplitude and frequency. Therefore, any adjustment to the reference level during testing will introduce uncertainty. The reference level is usually calibrated using an internal standard reference source (an external source can also be used). Similar to the built-in standard sources of many power meters, the PSA series spectrum analyzers have a built-in standard source with a frequency of 50 MHz and a power of -25 dBm, with an amplitude accuracy of ±0.24 dB (while the built-in standard source of the ESA-E series general-purpose spectrum analyzers has the same amplitude and frequency as the PSA, but an accuracy of ±0.34 dB). Therefore, the spectrum analyzer achieves the highest measurement accuracy when the reference level is set to -25 dBm and the attenuator is set to 10 dB, as the reference level parameters are calibrated under these conditions. The reference level uncertainty is typically given as ±0.3 dB within -20 dBm; this value increases as the reference level deviates from -20 dBm. It's important to note that different instruments may use different terms for "reference level uncertainty" in their specifications. For example, Agilent Technologies' 8560 series portable spectrum analyzers use the term "intermediate frequency gain uncertainty," while the PSA series uses "reference level accuracy." Since the attenuation value of RF microwave attenuators varies with frequency (and sometimes even temperature), the accuracy of step attenuators is also a function of frequency. Furthermore, if the attenuator settings during reference level calibration differ from the actual measurement settings, uncertainty will also be introduced. The accuracy of most attenuators deteriorates with increasing frequency; the typical uncertainty for attenuator switching is ±1 dB. Because the frequency response of analog filters is not ideal, the output amplitude characteristics of filters with different bandwidths can vary significantly. Therefore, the conversion resolution bandwidth filter also introduces uncertainty during measurement, especially when using analog filters. Digital filters perform well in this regard, but their implementation cost is higher. Therefore, in the mid-range ESA series spectrum analyzers, the digital intermediate frequency (IF) filter is only implemented up to 300Hz; higher bandwidth filters are simulated. The high-end PSA series employs a fully digital design for its IF processing section, including FFT analysis and a digital sweep receiver. This design not only improves amplitude measurement accuracy but also increases scanning speed. Changing the scale of each division on the screen display also affects measurement accuracy. For example, changing the scale from 10dB per division to 1dB per division alters the characteristics of the spectrum analyzer's logarithmic/linear amplifier, introducing uncertainty. Maintaining a constant scale during measurement avoids this error. Typical linear-logarithmic conversion uncertainty is ±0.25 dB at the reference level, but if the spectrum analyzer is displaying a saved trajectory, this uncertainty has no effect on the measurement. The total relative amplitude measurement uncertainty is affected by the sum of all the above factors. Some errors arise from changes in settings. If the attenuator, resolution bandwidth, and reference level settings remain unchanged, all related uncertainties can be eliminated, and the total uncertainty can be minimized. For example, PSA series spectrum analyzers, due to their use of fully digital resolution bandwidth filters, do not introduce additional errors when switching resolution bandwidths, resulting in significantly higher accuracy than spectrum analyzers using analog filters. To improve the accuracy of relative amplitude measurements, the simplest method is to avoid changing settings during measurement: do not change the resolution filter settings, although for fully digital filters like those used in PSA, the resolution bandwidth filter can be changed; ensure the same attenuator settings are used for reference level calibration and actual measurement; and do not change the scale of each division during testing. The signal transmission network connecting the spectrum analyzer and the device under test (DUT) affects the characteristics of the DUT signal, therefore these network characteristics must also be compensated for. Typically, the built-in amplitude correction function of the spectrum analyzer, along with a test signal source and power meter, can be used to measure the frequency response characteristics of the network. The measurement results can be tabulated and stored within the spectrum analyzer, and the data in the table can be used for correction during measurement. For accessories such as antennas and cables that are necessary in certain tests, the above method can also be used for compensation. Furthermore, the instrument can store multiple sets of data to accommodate different settings. Below is a typical example of uncertainty calculation, where the measured signal has a frequency of 1 GHz and an amplitude of -20 dBm. To compare the testing accuracy of different instruments, the high-end PSA series 4440A and the mid-range ESA-E series E4402A spectrum analyzer were selected. All settings were identical: attenuator 10 dB, frequency span 20 kHz, reference level -10 dBm, scan time set to automatic, resolution bandwidth 10 kHz, and video bandwidth 1 kHz. The ambient temperature was room temperature (+20 to +30°C). The nominal absolute amplitude uncertainty of the E4440A PSA (digital intermediate frequency filter) was ±0.24 dB, while that of the ESA (analog intermediate frequency filter) was ±0.54 dB. Adding these two figures to the absolute frequency response of each spectrum analyzer yields the worst-case uncertainty. For higher frequency signals, especially harmonic testing, the uncertainty is greater because the instrument needs to switch internal mixing bands. Using a digital intermediate frequency filter can effectively improve the measurement accuracy of the spectrum analyzer. During the measurement process, proper instrument settings can also ensure that the test results meet the optimal accuracy given by the instrument.