Re: Weird behaviour of the analog signal source

[Marcus D. Leech]

On 03/02/2020 04:45 PM, Lukas Haase wrote:
> Hi Marcus,
>
> How can (or better: *should*) a fully digital signal source have phase noise?
>
> Also, for 1Hz at 5MSps I always get either 5005789 or 5005790 samples (instead 
> of 5000000) ... this is fairly deterministic.
>
> Experimenting a bit, I see the weirdest behaviour of the analog signal source.
> In my opinion, the source should be fully deterministic. But it is not.
>
> Example 1: I pipe the output of the signal source into a file:
>
> https://snipboard.io/xY1JvE.jpg
>
> and read it with MATLAB:
>
> data = read_float_binary('baszmeg.dat');
>
> Then I compare it to an ideal (=expected) version:
>
> t = 0:1/fs:(length(data)-1)/fs;
> plot(t, [data - sin(2*pi*t)' ])
>
> https://snipboard.io/ecTaFL.jpg
>
> Now, I would not care too much about a constant phase shift or similar, but it 
> can be seen that the frequency slowly drifts (this is also seen if I just plot 
> them on top of each other).
>
> Example 2: I extend the block diagram with blocks that should never alter the 
> behaviour as they are only reading samples:
>
> https://snipboard.io/W6kyF0.jpg
>
> Note that the "Controller" is a simple Embedded Python block that only reads 
> the input samples into a temporary variable:
>
> def work(self, input_items, output_items):
>      squared_wave = input_items[0]
>      squared_wave[np.where(input_items[0] > 0)] = 1
>      output_items[0][:] = input_items[0]
>      return len(output_items[0])
>
> However, now the saved data is distorted:
>
> https://snipboard.io/amyn3X.jpg
>
>
> Any suggestions highly appreciated.
>
> Thanks,
> Lukas
Someone with more insight into the guts of the signal source needs to 
comment here.

It looks like it moves back-and-forth between integer and floating-point 
representations, and doing so, finite-precision problems can easily
  creep in.  But that's just an educated guess on my part.