Hello Kei,
the first issue is how the sin is being generated. The blocks you are using package the sine output into frames, so it isn't a single signal, but actually a frame with 1024 data points. This isn't how the data can be formated for the blocks you have connected the sin waves to. I would start with the standard sine wave block for your first attempt. You can add a zero order hold to convert it to a fixed time step for the spectrum analyzer. I set the time step to 1/(20e9)/10, which is 10 samples per cycle of your fastest frequency. This step could be skipped if you set all three sin wave blocks to a fixed step, but I like to start with a continuous system.
Regards,
Joel
