We are using AD9361 as a RF to test our Wifi Baseband. I have encountered some problems about the AGC using.
I use AD936X Evaluation Software to generate the config script.
AGC brief configration is shown below:
. FAST ATTACK AGC mode
. Split Table mode
. Energy Lost Threshold : 6
. Stronger Signal Threshold : 7
. Gain Lock Exit Count: 8
. Power Meas in State 5 : 3 ( 128 samples )
. Dec Power Measurement Duration : 2 ( 64 samples )
My digital interface Sampling Rate is 40MHz.
When testing, it is difficult for AD9361 to find the signal burst to lock the gain and the ending to unlock.
below is the CTRL_OUT( 0x035==12) logic sampled by LogicAnalyzer
DI and DQ is the RX digital input from AD9361, it is an 54M non-ht OFDM frame. Signal power is -21dbm.
In my understanding, there would be a stronger_signal pulse at the burst of the frame,and the AGC state jump to unlock state and adjust the gain. And at the end of the frame, there would be an Energy_lost pulse, then the AGC state jump out and relock again.
But the reality is it stays in state 5( lock state) at the burst of the frame. In the middle of the frame, an unexpected unlocking is appeared( mybe it is caused by a LMT overload ).To top it all, there was a stronger_signal at the end of the frame.
I'v tried tons of Stronger Signal Threshold and Energy Lost Threshold to optimize the signal detection, but failed.
How could this happen.....
Thanks a lot~
I use FDD mode. Does this strange thing in connection with this？