- Ask a related questionWhat is a related question?A related question is a question created from another question. When the related question is created, it will be automatically linked to the original question.
I have found a problem in CCS v.4 with autoscaling in the case of a FFT Magnitude Graph.
A test case to highligh the problem is the following.
A vector contains the impulse response of low-pass filter with 1kHz pass-band, 0dB amplification in pass-band, 2dB ripple in the pass-band and 60dB attenuation in the stop-band. The sampling frequency is 48kHz. We would like to graph the filter frequency response (simply obtained by plotting the FFT of the vector).
When we configure the Single-time graph and the FFT magnitude plot for the vector, we obtain the following output in CCS v.4:
The FFT magnitude output in CCS v.3.3 is, instead, the following:
In CCS v.4 the Single-time graph is correct while the FFT magnitude plot is not correctly scaled. As you can see, only a portion of the frequency response is visible in the pass-band. This happens since, when we set the logscale for the Y-axis of the graph, the autoscaling function imposes a minimum value for the Y-axis equal to 1. In this filter, however, it is very intresting to observe the behaviour in the stop-band, which has e minimum value in the order of -80dB (10^-4). Also this analysis requires a logscale for the Y-axis.
Please note that this problem can be solved also by allowing the user to manually configure the Y-axis datarange for graphs (as requested by other users of this forum).
I have attached a file with the data contained in the vector. The data type is 16bit-signed with Q-value=18. The CCS version I'm using is 4.2.2.00033.
