Part Number: TMS320F28335
Tool/software:
I used CCS7.4 to debug the data acquisition program in DMA mode. The collected data was stored in array A[256], and the odd points (left channel) in A were copied to array B[128]. The tools-graph-single time in CCS is used to observe the time domain waveforms of odd-point data in A and array B respectively, which are exactly the same. But when looking at their respective spectrum using tools-graph-fft magnitude, A's spectrum is wrong and B's is right. The Acquisition Buffer of A is 256, the index increment is 2, and the fft order of both is 10. However, if you change the Acquisition Buffer of A to 128, you will get the correct result. Is it a bug in CCS that updates to 8.1 and 12.2 get the same results?
