FFT source_ready in simulation

Hello,

I have a question to the FFT reaction on source_ready.

After source_eop I set source_ready to '0' for 3 clock cycles and then set it to '1' again.

(picture http://www.directupload.net/file/d/3896/qklbqmig_png.htm)

1 clock after source ready gone to '0' the source valid of the fft shows something like a "glitch"(?) and sets

source valid to '0' two clocks after that. With the next clock edge, source_valid is high again.

The number of clock cycles where source_ready is low seems to have no influence on this behavior. It also happens with 50 cycles

source_ready low.

For simulating I have used the auto-generated testbench from the MegaWizard Plugin Managerand compiled the .vho to the work library of

Modelsim.

Is this behavior normal or is my simulation somehow wrong? What can I do to solve the problem?

thanks in advance and best regards,

Andreas