Unified FFT not giving expected results

Using Quartus Prime Pro 22.4.0 and Questa Starter FPGA 2022.1

I created a Unified FFT with:

Log2Size = 11

Fixed Point

Reset Active High

Input width = 8

twiddle 14

Full word growth (output = 20 bits)

Target frequency 250MHz

I am creating various input STM files as below.

1. With a constant (DC) value in the STM file the output spectrum seems to be OK, with a spike at sample # 0 (the sample where validOut_dut transitions to '1'), and with the rest of the spectrum = 0.

2. When the input is changed to the convolution of a Blackman window and sine wave (plus a small amount of white noise) the resulting spectrum is not entirely clear to me. I am getting a strong DC component as expected, but the input frequency of the sine is around 400kHz and therefore I expect to see this at around bin number 3 or 4. Instead I am observing strong components at bins 256 and 512. In addition, as you can see, the spectrum is not exactly symmetric around it's centre.

3. If the sine input is used without a window function then the spectrum is even more not clear to me (screenshot of this is not included).

Please help me understand what is happening here.

In addition, I do not understand the function of the INPUT valid signal, since the UG mentions that it should NOT be set inactive during a FFT. So what is the purpose then of this signal.

Thanks

David