Adaptive filters
-
lilltroll
- XCore Expert
- Posts: 956
- Joined: Fri Dec 11, 2009 3:53 am
- Location: Sweden, Eskilstuna
Changed to arraybased channels to make the code more simple to read. Also connected I2Sout->Decimate->FIR->Interpolate->I2Sin, and it is working. Time to add the filterupdate as well - and prepare for looong filtercoeffs.
Probably not the most confused programmer anymore on the XCORE forum.
-
lilltroll
- XCore Expert
- Posts: 956
- Joined: Fri Dec 11, 2009 3:53 am
- Location: Sweden, Eskilstuna
Yesterday I was able to supress my first sine-wave with realtime subsampled ANC on XMOS.
Due to a poor estimate of the secondary channel, only the magnitude of the frequency response adapted correctly - the phase became delayed compared to the correct one - since the error signal and the reference signal must be perfectly in sync for the LMS to work.
But it worked for some sine-wave frequencies - since sine-waves is the same for 360*n degrees
Due to a poor estimate of the secondary channel, only the magnitude of the frequency response adapted correctly - the phase became delayed compared to the correct one - since the error signal and the reference signal must be perfectly in sync for the LMS to work.
But it worked for some sine-wave frequencies - since sine-waves is the same for 360*n degrees
Probably not the most confused programmer anymore on the XCORE forum.
-
lilltroll
- XCore Expert
- Posts: 956
- Joined: Fri Dec 11, 2009 3:53 am
- Location: Sweden, Eskilstuna
ANC is now working as supposed to - but I killed my CODEC :(
Probably not the most confused programmer anymore on the XCORE forum.
-
lilltroll
- XCore Expert
- Posts: 956
- Joined: Fri Dec 11, 2009 3:53 am
- Location: Sweden, Eskilstuna
My first multichannel ANC is now up and running on one L1 - that is soon maxed out.
Probably not the most confused programmer anymore on the XCORE forum.