"Phil Goddard" <
ph...@goddardconsulting.ca> wrote in message <jl79n7$7uh$
1...@newscl01ah.mathworks.com>...
I too am interested in extracting the Simulink Spectrum Scope data.
I am modelling a signal path where the signal source (data.mat file) is loaded in the workspace using the "From Workspace" source block. Using the Spectrum Scope block to view the spectrum of this data, I see the expected spectrum, with the noise floor some 130 dB down. Connecting the "From Workspace" block directly to a "To Workspace" block in the Simulink model, I am able to export the data (through Simulink). Post processing the output data in Matlab, the resulting spectrum is as expected (with the noise floor some 130 dB down).
Expanding my Simulink model by inserting a filter (plus some other blocks) in the signal path, I use the Spectrum Scope to examine the resulting spectrum along the signal path and it is as expected. However, using the "To Workspace" block to probe the nodes between my blocks in the signal path and post processing in Matlab, the resulting spectrum is different from that observed by the Spectrum Scope with the noise floor of the Matlab produced spectrum severely degraded (by some 70 dB). I suspect that I am not controlling the time step correctly when sampling the data using the "To Workspace" block. I am using "Sample time (-1 for inherited)", but there must be some subtlety about the timestep control that I am missing.
Being able to extract the Scope and the Spectrum Scope data would help debug my problem.
The version of Matlab is R2007b.
John.