I have a question regarding a real resolution of RSSI readouts that shall be placed at 6 bits when SSI Frame structure is defined as below:
16-bit I/ and Q, EAGC = 1, AAGC = 0:48 (figure 2 - fourth diagram from top, page 16 of AD9874 data sheet).
Do we have 6 bits significant in this mode or only 5 bits significant with the LSB that shall be neglected as in another SSI mode where AAGC=1:40 DATA BITS (thus ATTN and RSSI values are alternated with I/Q data)?
I did some measurements where input RF signal level was changed by 0.1 dBm to catch precisely enough moments of switching RSSI values to cover its full 6 bits scale.
To AD9874 a set of external attenuators, filters and VGA stage are connected to set reference point for my application. I have external AGC system that was disabled for this experiment. Below I present picture of maximal values of RSSI versus RF power. The funny thing is that I was NOT able to set RF level to get even RSSI values. RSSI values were captured as a single, maximal peak value of buffer consisting 8 RSS samples.
Can this effect be caused by to coarse RF power changes 0.1dBm?
Shall I do this with greater resolution?
Certainly this measurements can be biased somehow, or be effect of my mistake (I hope so) therefore it would be nice to have a final statement from AD9874 designers.
Greetings and thank you in advance,