We use the weak directional coupler to attenuate the power signal when we use AD8362 to measure the larger power. In 500 Mhz-900 Mhz frequency band, the coupling of weak directional coupler is positively correlated with frequency, and the coupling degree of the highest frequency in this frequency band is 4 dB higher than that of the lowest frequency. When we use AD8362 again, we find that the output voltage of AD8362 can be adjusted with the frequency curve by adjusting the capacitance and other components input to the AD8362 chip. That is to say, the output voltage of AD8362 after adjustment is negatively correlated with frequency. The effect of frequency on AD8362 measurement is solved by eliminating the influence of the two devices on the signal. Because it is still in the stage of theoretical research, I would like to ask if this method is feasible.
yes, you can design input match to introduce an opposite tilt to the response of coupler to compensate the undesired vout shift. better approach is to calibrate the device across the desire frequency range, and build up a set of calibration coefficients in the software and use interpolation to compensate the frequency variations.