> I note that in map65-fft1levels.htm you recommend a value of about 10
> for the value of Floor
> in the Amplitude Margins sub-window when integer arithmetic is used
> for Backwards FFT1.
> This is I think at the output of FFT1 and before the conversion to
> integer for the First Backward FFT.
>
> While playing with parameters this morning I noticed that Floor is
> about 21 in my setup.
> I infer that this is rather high and will limit my dynamic range
> somewhat.
Yes. But not much.
> If I understand this correctly I need to reduce the gain somewhere.
>
> I have at present:
> 1st FFT Bandwidth 100Hz
> 1st Forward FFT version 1
> 1st FFT Amplitude 1000
> 1st Backward FFT version 1
> 1st Backward FFT attenuation 3 (changed from 6)
> 2nd FFT Bandwidth factor 4
>
> If I understand this correctly, then the only place I can reduce Floor
> in Linrad
> is with the 1st FFT Amplitude. Otherwise with the SDR attenuation.
You can reduce 1st FFT amplitude, but that will not improve your dynamic
range although it might help the blanker operation just a little but
only under certain circumstances.
I assume you are using the SDR-IQ. For that particular radio the dynamic
range is limited by the USB transfer that is 16 bit only.
Optimizing a system should begin at the antenna. Does the LNA lift
the noise floor appropriately? None of the sucessive amplifiers
should lift the noise floor much.
If you have access to a signal generator and a couple of attenuators you
can find out at what level the different processing stages saturate.
By switching them on/off you can see what noise floor contribution they
give. Presumably you will find that a signal saturates the 16 bit
USB transfer before it saturates the A/D converter. In that case a
larger RCF output shift should be selected.
73
Leif