From 8:00 PM - 11:59 PM CST on Friday, April 28, ni.com will be undergoing system upgrades that may result in temporary service interruption. We appreciate your patience as we improve our online experience.

Optimizing Advanced Hardware Settings on NI Vector Signal Analyzers

Publish Date: Apr 26, 2017 | 0 Ratings | 0.00 out of 5 | Print

Overview

This article expands on topics discussed in Optimizing Basic Power Measurements on NI Vector Signal Analyzers, included in the Related Links section below. In order to get the most out of a spectrum or signal analyzer, it is important to have an understanding of dynamic range and how hardware settings affect the analyzer’s ability to detect signals. It should be noted that there are more driver settings and properties than are discussed in this article.

Table of Contents

  1. Understanding Dynamic Range
  2. Key Advanced Settings
  3. Conclusion
  4. Related Links

1. Understanding Dynamic Range

Before discussing more advanced hardware settings, it is important to first understand the concept of dynamic range. Signal analyzer dynamic range is the maximum power ratio (in dB) between a high-power signal and a low-power signal that are present at the input of the signal analyzer – such that both signals can be discerned and measured at the same time or in the same display. Electronic Design

While the above definition may sound simple, instrument characteristics such as noise floor and linearity each affect the analyzer’s ability to discern a low-power signal in the presence of a high-power signal. Moreover, for most of these characteristics, their impact on the ability to discern a low-power signal varies with the signal power that is present at the RF input of the signal analyzer. Electronic Design

For most spectrum and signal analyzers, a graph is used to illustrate the dynamic range of the device. The graph plots the power of noise and distortion signals relative to the mixer level, for a range of mixer levels. The dynamic range graph provides insight into two key performance metrics of a device: noise floor and linearity. The Average Noise Level describes the amount of noise that the equipment will add, and the Second/Third Order Distortion describes the harmonic and intermodulation distortion performance of the device. When using an RF signal analyzer, a signal too low in power could be lost in the noise floor of the analyzer while a signal too high will cause distortion and compression, and in extreme cases, damage to the analyzer. The Dynamic Range graph for the NI PXIe-5668 VSA is shown in Figure 1.


Figure 1: Dynamic Range plot for PXIe-5668 at 1 GHz, preamplifier disabled (nominal)

 

Ideally, an analyzer would be able to measure any signal without adding any noise or distortion to the measurement, but electrical components are not ideal. By observation, to minimize noise and intermodulation distortion, the intersection point of the Average Noise Level and Third Order Distortion lines shows the optimal mixer level. In Figure 1 above, the optimal mixer level for this analyzer is about -38 dBm.

As the mixer level increases, the noise decreases. Conversely, as the mixer level decreases, the noise increases. This can also be described in terms of the signal-to-noise ratio (SNR). As RF attenuation increases, the SNR decreases because the signal is being pushed down closer to the analyzer’s noise floor. The mixer level is simply the RF attenuation subtracted from the input signal level, as shown in the formula below:


Mixer Level = Input Signal Level – RF Attenuation


So far, only the basic concepts of dynamic range and interpretation of the dynamic range graph have been discussed. However, several other conclusions can be drawn from a device’s dynamic range graph. To learn more about dynamic range as it applies to spectrum and signal analyzers, refer to Part 6 of Super-Heterodyne Signal Analyzers in the Related Links section.

Back to Top

2. Key Advanced Settings

The following sections will describe six hardware settings to fine-tune an analyzer. While many of these settings are common to general spectrum or signal analyzers, this article will focus on how these settings are used in National Instruments Vector Signal Analyzers. For reference, the simplified analyzer block diagram below shows the various key hardware components:



Figure 2: Simplified analyzer block diagram showing key hardware elements discussed in later sections

 

RF Attenuation

RF attenuation is often the first step in the signal path of a spectrum or signal analyzer. From the dynamic range section above, it’s clear that the power level of the signal when it reaches the first mixer is important. Typically, a variable attenuator is used to allow the user to fine tune the attenuation in order to optimize the signal level at the first mixer.

RF attenuation helps bring the input signal into the dynamic range of the analyzer by attenuating the power level down to the optimized mixer level. As reference level increases, RF attenuation will increase. It is important to keep in mind that the signal entering the analyzer is theoretically infinite in terms of frequency and power. For example, if a signal has high-power components outside of the frequency range of the analyzer, a high-power signal is still coming through the input port, which can damage components.
With NI Vector Signal Analyzers, the NI-RFSA driver will adjust the RF attenuation based on the reference level. However, the RF attenuation can be read and adjusted with the RF Attenuation property. Refer to the Devices section of the NI RF Vector Signal Analyzers Help for more information.

Preamplifiers

A preamplifier, or preamp, is usually a low-noise amplifier (LNA) intended to boost low-power input signals. As shown in Figure 2, the preamp in an analyzer comes before the first mixing stage. In some devices, the preamp may be adjustable or only available within certain frequency ranges.

Let’s discuss why a preamp is beneficial to optimizing an analyzer’s dynamic range. The preamp helps boost the signal, improving the SNR, which in turn improves the dynamic range of the analyzer. While a preamp does amplify noise as well, it is specifically designed to minimize the noise contribution to the output signal. Therefore, the amplification of the signal is more significant than the noise that is added, thus helping to separate the signal from the noise.

Figure 3 shows how the preamp can allow an analyzer to measure a low-power signal that would normally be obscured by the noise floor of the analyzer. It’s important to understand the benefit to the dynamic range of the analyzer. The preamplifier increases the amplitude of all incoming signals, and by doing so, the impact of analyzer-added noise is decreased proportionally; the analyzer can now detect small signals which would have been lost due to instrument noise. Although the preamplifier increases the signal level relative to the mixer, the instrumentation typically will adjust the displayed values to normalize down the detected power, resulting in an apparent decrease in instrument noise, as shown in figure 3.



Figure 3: Left - Enabling the preamp boosts the signal into the dynamic range of the analyzer and improves the displayed noise floor. Right - With the preamp disabled, the signal cannot be resolved since it's buried in the noise.

 

To enable or disable the preamp, use the Preamp Enabled property from the NI-RFSA driver API. To understand the preamp capabilities of your NI Vector Signal Analyzer and to see a more detailed block diagram, refer to the Theory of Operation portions under the Devices section in the NI RF Vector Signal Analyzers Help.

As mentioned in the RF attenuation section above, signals are broadband in nature. Because of this, the entire signal present at the input of the preamp will be amplified, including any high-power out-of-band signals. The next section discusses the use of preselection filters to remedy this problem.

Preselection Filters

Preselection filters are bandpass or lowpass filters designed to attenuate out-of-band signals before they arrive at the first mixer. Without preselection filters, you would be forced to set the mixer level based on all signal content, potentially decreasing the effective dynamic range of the analyzer. Preselection filters can be fixed or adjustable in frequency, depending on the design of the analyzer. Fixed filters tend to have a higher Quality factor or Q-factor (narrower bandwidth), but design considerations must be made if multiple filters are needed. On the other hand, adjustable preselection filters may take up less circuit board space than multiple fixed frequency filters, but the Q-factor will vary over frequency.

The preselector filter is automatically tuned by the NI-RFSA driver. For analyzers with a configurable preselector, there are a few properties that can be used.

The Preselector Present property can be used to programmatically determine if the analyzer has a preselector. This can be useful in test systems in which different analyzers may be present to selectively control the preselection filter on modules that support it.
To enable or disable the preselection filter(s) on the analyzer, the Preselector Enabled property can be used. By default, analyzers with a preselector filter will enable the preselector when it’s in the signal path. For example, when the PXIe-5668 is tuned to 3.6 GHz or higher, the preselector will be enabled by default. To disable it, the Preselector Enabled property can be set to Disabled.

The PXIe-5667 analyzer utilizes a bank of preselector filters. The RF Preselector Filter property can be used to specify a particular preselector filter.

To learn more about the preselection filters available in a specific NI analyzer, refer to the device-specific Theory of Operation section in the help documentation included with the NI-RFSA driver as well as the Super-Heterodyne Signal Analyzers article in the Related Links.

Mixer Level

As discussed in the Understanding Dynamic Range section above, the mixer level is an important parameter for achieving an optimal dynamic range. The mixer level is the power level of the signal at the input to the analyzer’s mixer.

The driver, NI-RFSA, tries to optimize the dynamic range of the analyzer for minimum noise and distortion by evaluating the requested reference level and setting the attenuation and preamp accordingly. However, in some cases, it may be beneficial to further minimize noise or distortion in order to achieve a better measurement than the automatic configuration provides. Consider the scenario where a device produces spurs that need to be measured and are close to the noise floor of the analyzer. By default, the analyzer will adjust RF attenuation to optimize dynamic range. But in this case, it is more important to further minimize noise at the expense of more distortion to be able to resolve the spurs.

The mixer level can be adjusted in a number of different ways. First, as mentioned in the previous sections, the mixer level is directly related to the RF attenuation and preamp. If the RF attenuation and preamp are constant, then adjusting the input signal power will adjust the mixer level. The Mixer Level property can also be used to fine tune the analyzer’s mixer level by adjusting the attenuation applied to the signal before it reaches the first mixer. Refer to the NI RF Vector Signal Analyzers Help for more information regarding the Mixer Level property and other supported properties for a specific device.

IF Filter Bandwidth

The IF filter stage of an analyzer typically precedes the IF gain stage and applies a bandpass filter before it reaches the ADC. A smaller IF bandwidth can provide better dynamic range by attenuating out-of-band signals. However, a smaller IF bandwidth increases the measurement time since more acquisitions must be made to cover the same frequency span.

Choosing the right IF filter depends on the measurement being made. For example, EVM measurements require an instantaneous bandwidth wider than the bandwidth of the signal to ensure proper demodulation. However, for Adjacent Channel Power measurements and Two-Tone Intermodulation Distortion measurements, a narrower IF bandwidth may be more favorable despite the increase in measurement time. A narrow IF bandwidth allows the analyzer to maximize IF gain by only allowing a portion of the RF signal to be present at the IF stage at any given time. Electronic Design Figure 4 shows how a wider IF bandwidth allows more integrated power through, thus limiting the amount of IF gain that can be applied.



Figure 4: Left - A wide IF bandwidth passes more power and requires low IF gain. Right - Narrow IF bandwidth passes less integrated power and allows for higher IF gain. Electronic Design

 

Optimizing the IF filter and IF gain for measurements requires trial and error as well as a deep knowledge of the signal and measurement being performed. To learn more about optimization techniques, see Super-Heterodyne Signal Analyzers and Optimizing IP3 and ACPR Measurements in the Related Links section.

IF Gain

At the end of the down-conversion process, the signal must be scaled appropriately for the ADC.  This is important to ensure the full range of the ADC is used. Even more importantly, the IF gain can further help to raise a low power signal above the noise floor of the IF digitizer. However, too much gain will cause the ADC to clip the signal. Therefore, the key to maximizing the SNR, and in turn the dynamic range, is to maximize IF gain without causing distortion.

See figure 5 for an example. Assume an analyzer input signal containing two sine wave tones: a high-power tone at 1.004 GHz, and a low-power tone at 1 GHz. In this case, the reference level can only be decreased to just above the total power of the signal. Notice in the left trace how the tone at 1.004 GHz is easily seen, but the smaller tone is lost in the noise floor because the IF gain is too low. Compare this to the trace on the right, and note how an increase in IF gain effectively lowers the displayed noise floor so the signal at 1 GHz is resolved.  



Figure 5: Left - IF gain too low to resolve small signal at 1 GHz. Right - IF gain high enough to resolve the small signal at 1 GHz. (reference level – 0 dBm)

 

To adjust the IF gain, the IF Output Power Level (dBm) or the IF Output Power Level Offset (dB) properties can be used. The difference between the two is in their implementations: the former sets the IF output power level directly, while the latter sets an offset from the default IF output power level. The driver will throw an error if both properties are set at the same time.

Back to Top

3. Conclusion

There are many hardware settings available for fine-tuning measurements with a signal analyzer. In this article, we discussed how these settings can be used to increase dynamic range and improve measurements. It’s worth noting this has been a brief introduction to a subset of hardware settings for NI Vector Signal Analyzers. Further investigation is encouraged to learn more about the features of your analyzer and how to maximize its performance.

Back to Top

4. Related Links

Back to Top

Bookmark & Share


Ratings

Rate this document

Answered Your Question?
Yes No

Submit