SPRACP3D November 2019 – August 2026 AWR6843 , AWR6843AOP , IWR1443 , IWR1642 , IWR6443 , IWR6843 , IWR6843AOP
When performing the peak power measurement under EN 305 550 or 47 CFR 15.255, it is possible to see an apparent EIRP spike in the carrier. This can often be seen at the start and stop frequencies of the chirp. The chirp start frequency spike is typically the highest power of the two. In actuality, these EIRP spikes are not real, but rather measurement artifacts. The spikes originate in the test BW of the spectrum analyzer.
It is important to take a moment to discuss the background here: in a perfect world, the BW of spectrum analyzer would be much larger than the bandwidth of the signal of interest and have a sampling rate high enough to sample the full signal on the first pass. However, for many signals (such as multi-GHz FMCW radar chirps) this is not practical or possible. Additionally, it is not uniform for all regulatory applicants. To seek uniformity in testing, ANSI recommended that the BW of the spectrum analyzer be fixed at 1MHz.
This causes significant changes in measurement performance. Imagine if you will, the signal of interest is zipping by the detector so fast, that there is now not enough time for energy to accumulate on the detector. This causes large under measurement of radiated power, known as desensitization. To combat this, ANSI provides calculations that can be performed to compensate for this reduction in detected power. This is often called the desensitization correction factor.
At the beginning and end of a chirp, there can be overshoot and undershoot events as the VCO and synthesizer reset for the next chirp. While the total power emitted does not violate regulations in actuality, the desensitized detector now is able to collect additional energy from the non-FMCW overshoot and undershoot. This results in an apparent spike in the carrier, which the test-house will apply a correction factor to. This will significantly erode margin and in some cases cause regulatory excursion. Since this occurs at the start and stop frequency, it is difficult to prove it is not part of the FMCW signal. Therefore, we recommend increasing the TX start time until the apparent EIRP spike dissipates. This will reduce the BW marginally; however, it will reduce the power of the spike, and it will push the spike down into the spurious region. Greatly improving margin and passing. Typically this can be achieved with 1.5μs to 2μs of TX start time delay.
Figure 6-8 shows an example of this. The scope has been placed in 1MHz RBW, causing desensitization of the detector. The blue trace shows a chirp with EIRP spiking that can cause FCC regulatory failure. The black curve shows the same chirp configuration but with a 1us TX on time. Notice the EIRP spikes remain but are now in the OOB region.
Figure 6-8 Example of EIRP Spike and Resolution