RF Interference Measurements

The Automated Radio Measurement System (ARMS) developed by EMA is used to measure the broadband performance of RF systems. In order to accurately predict RF interference (RFI) between multiple RF systems in complex environments, one must use high fidelity measured data. There are many ways that transmitters can interfere with receivers. The root cause of the interference is often not intuitive and the limited performance data provided by vendors makes developing accurate models a challenge. Manual measurements for transmitters and especially receivers can be very time consuming especially when considering radios that can operate over hundreds or thousands of channels.

Analysts must often make educated guesses or use worst-case assumptions in their RFI analysis, resulting in missing real interference problems or over-engineering the solution for interference problems that do not exist. This approach has major implications on time/resource allocations that can result in overly complicated equipment. Unfortunately, measured data, especially broadband measured data, is not readily available for most RF systems. RF system vendors are required to demonstrate that their equipment meets the various military and commercial standards with respect to transmitter emissions and receiver susceptibility. However, the raw measured data is generally not made available to the analyst. Even when such data can be obtained, it very often lacks the fidelity necessary for RF interference analysis.

With the ARMS, EMA can rapidly and very accurately characterize the performance of RF systems. Automation and high dynamic range measurements are key aspects of the ARMS. For transmitters, the ARMS measures the fundamental, harmonics and spurious emissions. For receivers, mixer product responses and spurious responses of the receiver are captured using different susceptibility metrics such as SINAD, BER and C/No.

An example of measured transmitter data collected with the ARMS for a commercial handheld radio tuned to 146 MHz is shown below as the orange trace. A simple model for the transmitter based upon information found on the specification sheet for the radio is overlaid with the black trace to show the important differences between parametric models and measured data. One can clearly see numerous spurious emissions that occur both below and above the fundamental frequency. The parametric model based upon specification sheet information has no knowledge of the spurious emission frequencies. At best, a specification sheet will provide a maximum spurious emission amplitude limit but again, no frequency information is provided. It can also be seen that the harmonic amplitudes based upon specification sheet information are far different from the measured amplitudes of the harmonics. In fact, most of the measured harmonics are 30 dB or more lower in amplitude than the harmonic amplitudes based upon the specification sheet information. Clearly, the missing spurious emissions and the gross inaccuracy of the harmonic amplitudes could lead to very inaccurate RF interference predictions.

The Rx susceptibility is the interference power level required to degrade the Rx performance by a specified amount in the presence of an intended signal and is a function of interferer frequency. Within the Rx tuned bandwidth, susceptibility is closely related to Rx sensitivity. However, out-of-band susceptibility behavior is complex. The detailed frequency-dependent susceptibility profile is not available from a manufacturer’s specification sheet but can be obtained by measurement using two signal sources. The ARMS Rx measurement system automates and accelerates the Rx measurement process to rapidly and accurately create models for RF interference analysis.

Measured receiver susceptibility data for a handheld commercial radio tuned to 420 MHz is shown below. The measured data is the black trace while a parametric model based upon a simple mixer product model is the red trace. For the simple mixer product model, there is no knowledge about the amplitudes of the mixer products except that they shall not exceed a specified threshold.

The mixer product model predicts 17 mixer products (assuming 3rd order mixer products) but measurement shows that only 3 mixer products actually exist. Further, the amplitude of one of the mixer product model responses matches the amplitude on the specification sheet (-70 dBm), but the other two mixer products are over predicted by greater than 20 dB. Similar to the transmitter case, this example demonstrates the pitfalls of using simple models for the broadband behavior of transmitters and receivers.

To learn more about how EMA can help your organization with RF measurements, Contact Us.