5G promises significant improvements over previous generations of wireless communication technology, particularly in speed, latency, bandwidth, and quality. Most of the gain comes from using the 5G frequency band 2 (FR2) found in the millimeter wave (mmWave) spectrum. mmWave spectrum is attractive for use in wireless communications because these bands are relatively untapped, which means there is plenty of available bandwidth. Millimeter wave transmissions are smaller than other wireless communications signals, making them ideal for high-speed transmissions in dense urban areas, where many devices operate in close proximity. However, mmWave’s advantages for 5G connections are partially offset by several technical challenges. For starters, mmWave doesn’t spread very far – mmWave transmissions are easily absorbed by the atmosphere and don’t penetrate trees, building walls, and other infrastructure. It is also difficult to measure the performance of mmWave devices using over-the-air (OTA) testing equipment and methodologies. mmWave’s wide bandwidth – an attractive feature of 5G connections – also reduces the signal-to-noise ratio (SNR) because the energy from the signal is spread across the bandwidth. Finally, mmWave uses higher-order modulation schemes to improve spectral efficiency, which in turn requires error vector size (EVM) performance improvements. Signal Analyzer Block Diagram As the signal strength decreases, the noise from the test system used to measure it will also reduce the SNR, affecting the results. Therefore, signal analyzers are designed to handle multiple types of testing applications, including high and low power, narrowband and broadband signal patterns, spectrum patterns, or vector patterns. This diversity introduces many possible components into the signal path, though, including low-noise amplifiers (LNAs), pre-amplifiers, attenuators, pre-select filters, and more. Applying or modifying some of these components can improve measurement accuracy for different test scenarios. This article examines some of the technical challenges posed by mmWave frequencies and the difficulties these challenges pose for accurate and reproducible measurements. The article also suggests strategies to improve the accuracy of the measurement with the signal analyzer through the use of different signal path settings. Path loss Excessive path loss is among the most annoying and commonly cited challenges for 5G mmWave connections. Loss of path between the device under test (DUT) and the measuring equipment reduces the signal-to-noise ratio (SNR), making accurate measurements of metrics, such as EVM, adjacent channel power, and spurious emission difficult. Complicating matters further, the small size of the components and antenna arrays eliminates the possibility of placing probes for the tests performed, necessitating the use of OTA testing – or radioactive testing. OTA test requirements, along with excessive signal path loss for mmWave transmissions, require control and calibration of the radiated environment around the test setup. Signal path loss balancing requires a flexible signal analyzer hardware and software that enables the creation of the optimal solution for a given signal and measurement. For example, a signal analyzer can apply attenuation at higher power levels or a preamplifier at lower power levels to measure a variety of input signals. Signal analyzers provide multiple RF signal paths to reduce noise, improve sensitivity, and reduce signal path loss. Low-level signal metering (default signal path) By default, in the standard signal path of the signal analyzer, the input travels through the RF attenuator, amplifier and preset before reaching the mixer. This signal path is ideal for measuring low-level signals with a bandwidth of less than 45MHz. Measuring millimeter wave broadband signals can be particularly challenging when analyzing broadband vector signals (microwave pre-selector bypass path). Bypassing the signal analyzer preset is a good choice when increasing the RF analysis bandwidth to analyze broadband vector signals because it allows broadband signals to pass unhindered down the RF chain. Bypassing the preset not only enables wideband analysis, but also eliminates amplitude drift and preset passband ripple, further improving the overall accuracy of the measurement. Bypassing the microwave limiter Optimizing modulation analysis (low-noise signal path) The low-noise signal path is well suited for EVM measurements and other measurements testing transmitter modulation quality at higher power levels. Since the amplifier gain, frequency responses and insertion losses double at higher frequencies, lossy switches in the preamplifier path bypass and preamplifiers provide the optimal path for the RF signal. This path reduces path losses, frequency responses, and noise from preamplifiers and switches. Choosing this signal path for broadband EVM measurement leads to higher frequencies, increases measurement sensitivity, and improves signal accuracy. Wideband modulation analysis (full bypass signal path) The full bypass signal path reduces path loss, improves signal accuracy, and increases measurement sensitivity. A full bypass signal path can reduce the loss at mmWave frequencies by up to 10 dB compared to the default signal path. The full bypass signal path is a combination of a low-noise signal path and a microwave limiter bypass path, which avoids the multiple switches in the low-range switching circuit as well as the microwave pre-selector. While the advantages of using a full convolution path are obvious, this path has some drawbacks, including potential in-band imaging and low SNR for testing low-power signals. However, eliminating images in the desired range by adding a bandpass filter can improve EVM results by 1 to 2 dB. The external amplifier can also improve the signal-to-noise ratio (SNR) when testing low power signals. Other Considerations Another critical component that affects the accuracy of 5G mmWave measurements is the level of the input mixer. Setting the input mixer level for the signal analyzer provides a trade-off between distortion performance and noise sensitivity. As discussed above, the signal-to-noise ratio (SNR) in 5G mmWave signals decreases due to broadband noise and excessive path loss, resulting in poor EVM and adjacent power ratio measurements that are not representative of the actual performance of the DUT. The Signal Analyzer Input Mixer is another tool that can help overcome the challenges of 5G mmWave frequency measurements. The optimum mixer level setting depends on the measuring devices, input signal characteristics, and specification test requirements. It is also possible to apply an external LNA on the front end of the signal analyzer to improve the mixer input level. Some new signal analyzers, such as the Keysight N9042B UXA X-Series Signal Analyzer, include a signal path low amplifier, along with the amplifier. This allows users to realize the benefits of using LNA to optimize the mixer input level without the need for external components. To get the best EVM measurement results, the mean frequency (IF) noise of the signal analyzer should be low enough that it does not reduce the signal-to-noise ratio (SNR). The input signal to the digitizer should be high enough, but not so high that you overload the digitizer. The optimal balance is a fine jig that requires a combination of RF attenuator, preamplifier, and IF gain value based on the peak level of the signal. New signal analyzers allow users to optimize these hardware settings at the touch of a button, improving the signal-to-noise ratio (SNR) while avoiding digitizer overload. However, manually tweaking settings such as IF gain and RF attenuators is often necessary for optimal settings, resulting in the best measurement results. Components in the signal path Another critical factor to consider for accurate 5G mmWave measurements is the effect of components in the path between the signal analyzer and the DUT. Components in the signal path can degrade the overall measurement accuracy of the signal analyzer. All elements of the test network must be taken into account. Measurement accuracy becomes more important as bandwidth increases and frequencies rise into the mmWave spectrum. With smaller margins for error, engineers need to find ways to eliminate frequency response errors, which occur at different frequencies and influence phase and amplitude responses. Signal analyzers provide an internally calibrated procedure to correct their frequency responses. Cables, connectors, switches and fixtures in the signal path between the signal analyzer and the DUT can degrade measurement accuracy due to frequency response errors. Using different amplitude correction configurations and complex corrections can help remove frequency responses, providing a more accurate picture of DUT performance. Signal analyzers make it possible to configure both amplitude and complex corrections to correct frequency responses (although a high-performance signal generator or bus network analyzer is required to calibrate the test network). Using a signal generator with a power meter and sensor to measure the amplitude and then entering the correction values into the signal analyzer is an effective way to make amplitude corrections. New receiver calibrators specifically designed for signal analyzer receiver measurement systems, such as Keysight’s U9361 RCal receiver calibrators, provide a transmission standard that enables both absolute amplitude and complex magnitude and phase corrections. The frequency extender is connected to the receiver calibrator. Making accurate measurements of 5G at mmWave frequencies The promise of 5G – particularly the mmWave FR2 range of 5G – is clear. It provides step functionality increase in speed, bandwidth, and performance and will eventually enable entirely new use cases and business models. But working with mmWave frequencies presents obstacles, particularly with regard to path loss, which makes accurate and reproducible measurements difficult. Understanding and using the various RF signal path options on a signal analyzer can help you overcome these challenges when performing 5G mmWave measurements. About the author Dylan McGrath is a veteran tech journalist and former editor-in-chief of the EE Times. He is now Senior Director of Industry Solutions at Keysight Technologies. .