Recently , I'm woking on a Rx frequency changing project using the AD9361.My goal is to change the Rx frequency 1000 times in a second.We decide to change the Rx LO Frequency from 1GHz to 3GHz in 0.1s.Everytime we increase 20MHz.So we need 100 times change.Nect 0.1s,we do the same thing.At last ,we change the frequency tatolly 1000times in one second.
In a simple experiment ,the input signal is fixed that is 2.405GHz generated by the external signal source with the 4MHz bandwidth.Rx LO frequency is changed from 2.4GHz to 2.405GHz in a millisecond.Then ,the Rx LO frequency is changed from 2.405GHz to 2.41GHz in a millisecond.Lastly , the Rx LO frequency is changed from 2.41GHz to 2.415GHz in a millisecond.
There is some question about the Rx signal .Here is the spectrum.
Demodulated signal has a symmetrical spectrum.We conclude that the QEC tracking mode leads to the Rx signals unbalanced quadrature.When I set the change time from one millisecond to one second , the symmetrical spectrum is disappeared.Like that
Obvious , such a time interval can't meet our requirements.Our project goal is to change the Rx frequency 1000 times in a second.
We have try the shot QEC mode .It also doesn't work.
We have noticed that any time LO Frequency changes more than 100MHz should updated the three calibrations highlightred yellow.
I have some questions.
1.Is there something else that caused the wrong received signals?Am I wrong?
2.So can I change the Rx frequency quickily and ensure that the Rx signal spectrum is normal?
3.Eveytime , we only change the Rx LO frequency to increase 20 MHz.But , totally our goal is from 1GHz to 3GHz.Is it contradictory to the user guide?When should I update the three calibration?