Why is there unstable measurements in active pulse LP power sweep?

June 12, 2024



I have been making some CW and pulsed loadpull measurements on a FET. I am having some problem getting clean pulsed measurements, the plot below has the power sweep for a FET CW and pulsed, any ideals what would cause the pulse measurements to look that noisy. The period is 1mS and the pulse width is 10uS. 1% duty. If I sat down on a thru line the measurement is clean.

Symptoms / Solutions:

There are four possible errors to investigate:

1: The noisy measurements can come from the Averaging, as we know the IF bandwidth for the pulse is very large and unstable; to compensate we need to increase the averaging setting.

2: An issue with the Weights and Radius setting of tuning in the project. That means for each power level, FDCS will control the active injection signal generator to adjust the impedance to the target impedance, however there is a tolerance. But sometimes depends on the active amplifier’s capability, FDCS can’t tune the impedance to the exact point. This will result in a shift in impedance. This usually happens when tuning near the edge of Smith Chart or with high gamma. If we reduce the gamma slightly to allow the active amplifier to tune, this may result in more stable measurements.

3: Add some delay in the Measurement Editor, because sometimes the PNA-X’s reading is faster than the setting of DC supply. That means when the DC not finishing the setting, PNA-X starts reading already.

4: Try to adjust the RF measurement delay and measurement period to get the proper and stable measurement region.

Keywords: Active LP, Pulse, Power sweep, Delay, Stable, IF bandwidth.

Last date modified:  06 Mar 2017