Adjacent Channel Distortion Measurements Using the NI PXIe-5665 VSA

Increasing performance requirements for production test warrant more accurate measurement methods. Varying hardware architectures in vector signal analyzers (VSAs) often present an opportunity to tailor measurement methods to use individual hardware capabilities. Modifying measurement techniques for a specific device ensures the best possible measurements. This tutorial applies the idea of device-specific measurement techniques to adjacent channel power measurements using the NI 5665 VSA architecture and NI-RFSA software suite. The tutorial covers the basic receiver architecture for the NI 5665 VSA and how that impacts ACPR measurements. Further discussion addresses how to improve the quality of such measurements with the Minimum ACPR feature introduced with the NI-RFSA 2.4 software driver. Last are two examples and a method for using the NI-RFSA 2.4 Minimum ACPR feature and NI Spectral Measurements Toolkit 2.5.1 Adjacent Channel Power calculation function.

1. Introduction
2. Receiver Architecture
3. Measurement Example
4. Using SMT for ACPR Measurments

Introduction

Measuring adjacent channel distortion is an important parameter when characterizing wireless transmission systems. Radio system interference of adjacent channels can be measured with two modulation methods: intermodulation distortion (IMD) and adjacent channel power ratio (ACPR). The IMD technique tests a system’s nonlinear distortion by stimulating with multiple sinusoidal tones in the main channel and measuring the peak power ratio of the main channel to the adjacent channel. ACPR, sometimes referred to as adjacent channel leakage ratio (ACLR), is an improvement on the IMD method. The ACPR technique stimulates with a signal specific to the usage application, for example CDMA, WCDMA, and so on. A ratio of the measured power in band of the main channel to the measured power in band of the adjacent channel is calculated. With either method (ACPR or IMD) the measurement is often performed on the most nonlinear components in a transmission chain such as the power amplifier. Both methods are similar in that the main channel contains a larger signal than the adjacent channel. Making this measurement accurately can be a challenge, requiring a high-dynamic range receiver, filtering, and noise correction. This tutorial explains how to improve such measurements using the NI PXIe-5665, NI-RFSA software driver, and the NI Spectral Measurements Toolkit (SMT).

Receiver Architecture

The hardware architecture plays an important role in how the ACPR measurement is taken. Characteristics such as bandwidth, dynamic range, and image rejection greatly affect the quality of the measurement. Different combinations of these device specifications may be more or less important for different measurement applications. 

Dynamic Range Considerations

For an adjacent channel power measurement, the large main channel signal power is often present in a receiver’s bandwidth. This fact forces setting the reference level higher than desired when measuring the lower power adjacent channel. If the dynamic range of the receiver is sufficient, this may not be a problem. The easiest solution is to filter the main channel signal when measuring the adjacent signal. This gives the RF receiver the ability to apply more gain to the adjacent channel signal and improve the quality of the measurement.

NI PXIe-5665 Receiver Architecture

The NI PXIe-5665 contains an NI PXIe-5603, which is a three-stage super heterodyne downconverter. Figure 1 shows a block diagram. The earlier RF/IF stages have a wider bandwidth of 80 MHz. The final IF stage has a selectable 300 kHz bandpass filter, which is useful for ensuring that adjacent or main channel signal power is not leaking into the channel being measured.

Figure 1. NI PXIe-5665 Receiver Block Diagram

When performing a power measurement on the adjacent channel, the main channel signal energy can be present in the pre-IF filter stages. If this is the case, then the gain of these earlier stages should be set according to the main channel signal power to ensure the signal chain is not overdriven. The gain after the 300 KHz BPF can be increased because the main channel signal is rejected. Figure 2 shows a conceptual view of these bandwidths and signal levels. Minimum ACPR is a quantity that NI-RFSA uses to appropriately increase the post filter gain.

Figure 2. Channel Bandwidths and Receiver Bandwidths

Measurement Example

The following example illustrates the proper method for taking an ACPR measurement using the NI PXIe-5665 and NI-RFSA.

NI-RFSA Software Primer

The NI-RFSA software driver controls the NI PXIe-5665. NI LabVIEW software is not required; however, all examples in this tutorial use LabVIEW. NI-RFSA is an attribute-driven device. The user can write several attributes together and then commit that state to hardware before making a signal acquisition.  Each attribute contains a default value and is readable and often writeable. Often a user may write a desired value that is not exactly achievable; NI-RFSA does its best to calculate an appropriate coerced value provided the information from all preceding attribute writes. NI-RFSA contains many configurable attributes. The following example needs only a few to demonstrate the concept of improving the receiver gain for ACPR and IMD measurements. The following is a list of attributes used:

NIRFSA_ATTR_REFERENCE_LEVEL
NIRFSA_ATTR_IF_OUTPUT_POWER_LEVEL
NIRFSA_ATTR_MINIMUM_ACPR
NIRFSA_ATTR_SPECTRUM_CENTER_FREQUENCY
NIRFSA_ATTR_SPECTRUM_SPAN
NIRFSA_ATTR_DEVICE_INSTANTANEOUS_BANDWIDTH
NIRFSA_ATTR_DOWNCONVERTER_GAIN (Read Only after commit)

Example Setup

In this conceptual example, we measure the ACPR of a main channel at 1 GHz, channel spacing of 5 MHz, and channel bandwidth of 4 MHz. We estimate that the main channel peak power will be below +10 dBm. The goal is to measure the main channel power in band and the adjacent channel power in band and subtract the two to come up with an ACPR metric in dB. The modulated signal source simulates a 4 MHz wide communication channel, with an appropriate data model for the given application.

Figure 3. Example Hardware Setup

Measuring the Main Channel

Measuring the main channel power in band is the first quantity to obtain. The REFERENCE_LEVEL is set to 10 dBm. MINIMUM_ACPR is set to zero because adding extra gain after the IF Filter only results in overdriving the digitizer. CENTER_FREQUENCY is set to the main channel frequency of 1 GHz. SPAN is set to the channel bandwidth of 4 MHz and the DEVICE_INSTANTANEOUS_BANDWIDTH is set to 300 kHz to force the use of the 300 kHz IF Filter. The use of the narrow filter path is not required for measuring the main channel. In truth, using the narrow filter when not required increases the time of this measurement. The IF_OUTPUT_POWER_LEVEL is set to –6 dBm – this results in a downconverter gain less than -16 dB. The Read Power Spectrum function commits all attributes to hardware and performs a data acquisition to calculate the power spectrum. After the spectrum data is obtained, the power in band is calculated from the power spectrum data. The SMT provides functions and examples to easily calculate power in band. A later example shows more details of SMT capabilities, and some limitations.

Figure 4. Main Channel Power in Band NI-RFSA Measurement Setup

Measuring the Adjacent Channel

Measuring the adjacent channel power in band is similar to the main channel, except we estimate the MINIMUM_ACPR based on some expected distortion performance. This is conceptually shown in Figure 2. For explanation purposes, we assume the ACPR should be greater than 20 dB. We use 20 dB as our value for MINIMUM_ACPR. The CENTER_FREQUENCY is tuned to 1.005 GHz to measure the power in band of the high side adjacent channel. The 300 kHz filter is selected from setting the DEVICE_INSTANTANEOUS_BANDWIDTH property. NI-RFSA errors if MINIMUM_ACPR is greater than zero and the narrow filter path is not selected. With these settings, users would expect a final downconverter gain of +4 dB to get the adjacent channel signal to the desired -6 dBm IF output power level. Because there are a limited number of calibrated IF states available, NI-RFSA may not be able to find an attenuator/gain state with exactly 20 dB of extra gain after the IF Filter. NI-RFSA searches all states and selects one that has a prefilter gain appropriate for the main channel power and gets as close to the desired MINIMUM_ACPR without exceeding it. If there is no such state, the MINIMUM_ACPR is coerced to zero and no benefit is gained; however, the measurement is still performed. Typically for higher reference and mixer levels there is more post filter gain available because more IF Attenuation is applied in the configurations. The user can read the coerced value to determine what extra gain is added. In our example, we are able to obtain only a coerced MINIMUM_ACPR of +14 dB, which results in a downconverter gain of -2 dB. Figure 6 shows a conceptual illustration of these signal levels. This additional filtering and gain results in increased noise performance and a more accurate measurement.

Figure 5. Adjacent Channel Power in Band NI-RFSA Measurement Setup

Figure 6. IF Out Signal Levels With Adjacent Channel Signal Boosted

Example Considerations

Once the power in band of the main channel and the power in band of the adjacent channel have been measured the difference can be taken to calculate the measured ACPR value. Although this example was simplified for explanation purposes, these two measurements can be incorporated into a single routine to automate this process. Additionally, we discussed measuring only the high side adjacent channel; this method can also easily measure the low side and even subsequent alternative channels. For large channel spacing or for channels where the larger main channel signal is out of the receiver bandwidth, users can adjust the REFERENCE_LEVEL and set MINIMUM_ACPR to zero to improve such measurements. Users can also increase the IF_OUTPUT_POWER_LEVEL attribute to obtain more dynamic range; however, increasing this quantity increases the chance of saturating the digitizer ADC. The default value for IF_OUTPUT_POWER_LEVEL is -6dBm, which has designed headroom to decrease the chance of ADC or OSP saturation. Adding averaging may also improve the quality of the measurement – NI-RFSA has attributes to configure automatic averaging.

NIRFSA_ATTR_SPECTRUM_NUMBER_OF_AVERAGES
NIRFSA_ATTR_SPECTRUM_AVERAGING_MODE

Using SMT for ACPR Measurments

SMT is a package of functions that extend the existing functionality of LabVIEW to perform frequency analysis tasks. SMT works with NI-RFSA to provide functions that analyze data or operate the NI-RFSA instrument directly to make certain measurements. The Minimum ACPR feature was introduced in NI-RFSA 2.4 and, as of the writing of this tutorial, SMT version 2.5.1 is current. SMT 2.5.1 and later provides a function specifically for ACP measurements, SMT Adjacent Channel Power.vi. Figure 7 shows the external connection to this function.

Figure 7. SMT Adjacent Channel Power VI

This VI requires that spectrum data be presented in a contiguous span. NI-RFSA 2.4 does not have an automatic way to apply an attribute such as Minimum ACPR to a specific frequency range or others in a single contiguous span acquisition. Hence, the user must perform multiple acquisitions with the appropriate Minimum ACPR applied to the specific channel, as described in the previous section example. The existing SMT ACP function is incompatible with the usage requirements of the Minimum ACPR feature in NI-RFSA. 

Figure 8. SMT Power In Band

To use the NI-RFSA Minimum ACPR feature, the SMT Power in Band can be used rather than SMT Adjacent Channel Power. With this function, the power in band for each channel can be extracted directly. A tool can be created using the iterative approach described in the example above. The tutorial, WCDMA ACPR Example.vi, provides an example of this.

Example Code

The attached example VI includes two key subVIs: Calculate ACP Channel Parameters.vi and ACP with Divided Spectrum.vi. These two VIs are designed to work together in the sense that Calculate ACP Channel Parameters.vi produces an array of channel parameters that specify the center frequency, Minimum ACPR, and reference level for each channel in an ordered array. Users can take this array and loop over these parameters and use NI-RFSA to gather the appropriate spectrum data and push it into an array. The VI ACP with Divided Spectrum.vi operates on this array of spectrum data and extracts the power in band of each channel. The user can then take the difference of each channel to calculate the power ratio relative to the main channel. With this information, the high side, low side, adjacent, and alternate power ratios can all easily be calculated. Figure 9 shows a screenshot of this example. We have omitted the details of the implementation for brevity. The VI, however, should be self-explanatory with the basic knowledge of how the NI-RFSA software driver works as explained earlier. The screenshot shown is a measurement using an NI PXIe-5673 as the signal source using the NI WCDMA/HSPA+ Generation Toolkit and examples with no device under test (DUT) between it and the NI PXIe-5665.

Figure 9. ACPR Example Front Panel

Conclusion

In this tutorial, we covered the receiver architecture of the NI 5665 VSA and its impact on ACPR measurements. We further discussed the basics of using the NI-RFSA software driver and the Minimum ACPR feature to improve ACPR measurements. In addition, we provided an example that illustrated how to integrate the Minimum ACPR feature with the SMT. Users can achieve further improvements to ACPR measurements with data processing techniques such as noise floor correction.

Downloads

wcdma_acpr_example_lv90.zip

6 Ratings | 2.00 out of 5

 Print |  PDF

Reader Comments | Submit a comment »