Greetings E2E.
We are performing RF testing on a new 2.4GHz wireless device that uses the CC2652P7 that has an RF circuit designed based on the CC1352PEM-XD7793-XD24-PA24 reference design. During this testing, we observe an unexpected level of variability in the high-power PA Tx output power in the range of 3-5 dBm for a given frequency and output power configuration.
Our measurement setup is for conductive RF power measurements using a spectrum analyzer connected at the PCB antenna input via an SWF type coax connector with isolation switch (muRata MM8130-2600). The radio is configured for continuous transmit at +20dBm output at either 2.402, 2.426, or 2.480 GHz and power is measured until peak stabilization and the peak level is recorded. We have eliminated our measurement setup or procedure as a cause of the variability, and we can repeatedly measure the same device's output - even days apart - with very low variance in measurements.
After trying many approaches (auditing component tolerances, examining board build quality, swapping SWF connectors, etc.), our latest effort to isolate the cause of the output power variability had us swapping the MCUs on two previously tested boards - one with good Tx output power and the other with poor Tx output power. We were somewhat shocked to find that the Tx output performance of the two boards followed the MCUs, so that the board that previously had poor output power now had good output power nearly identical to the other boards previous measurements. Likewise, the board that initially had good performance now had the poor performance as measured on the initial bad board. Our conclusion is that the Tx output power variance is tied to the MCU part, but we do not know if the MCUs themselves are the issue or if there is some other issue that only manifests on certain parts.
As I mentioned, we have reviewed the tolerances of all the parts on our RF circuit. However, we have not been able to find any Tx output power tolerance for CC2652P7 itself in the datasheet, TRM or other technical documentation.
Question: Can you think of any possible RF front-end design issues that could result in poor Tx output power performance on some CC2652P7 parts, but not on all parts?
Question: What is the tolerance of the high-power output power of the CC2562P7?