I want to measure differential power at different frequencies. I want to use AD8364 detector .How do i calculate power for frequencies whose corresponding slopes are not given in the datasheet?
The newest revison of the evaluation board (AD8364-EVALZ, see Rev. A datasheet) has a 50 ohm broadband match. So no addional matching over your frequency range is required.
Regardless of whether the slope and intercept is specifiedc at your frequency of interest, you have to calibrate the detector. The slope and intercept specs that are given in the datasheet will vary from device to device so they cannot be relied on to give you any level of accuracy. The attached application note will hopefully be of assistance.
Thanks for your reply
Now my aim is to calculate power at frequencies ranging from 2 to 2.5 GHz with 100kHz spacing.
does this algorithm work well
yes, your approach is sound. I'm not sure that you need to calibrate at 100 KHz intervals. The behavior of AD8364 does change as you go from 2 GHz to 2.5 GHz but the change is fairly slow. Steps of 10 or 50 MHz are probably fine.
Thanks Eamon your pdf and opinion have been very much helpful.
I still have a doubt .
Do i have to consider the impedance matching at the input (as i am driving RF input by 50 ohm source)?
Or Does the AD8364 -evalz evaluation board provides impedance matching at different frequencies?
Or do i have to consider various baluns to characterize AD8364?
In the datasheet for AD8364 , there are different max and min input levels for different frequencies .
Are these experimental limits used for calliberation of slope and intercept?
Can i use input power exceeding that given in the datasheet for the frequencies?
For example ,
for 2.14Ghz frequency, Max input level (Chan A/Chan B) - -2/-4
Min input level (Chan A/ Chan B) - -57/-51
Can i use max - +5 and min -55?
Retrieving data ...