[time-nuts] Noise source measurement

[Jim Lux]

On 10/22/11 5:50 AM, Javier Herrero wrote:
> Hello all,
>
> Perhaps a bit OT, but I'm measuring the output noise density of a noise
> source at a puntual frequency. I've fed the noise output to a 8566B
> spectrum analyzer, BW set to 1MHz and video BW set to 1kHz so the
> displayed trace is flat. I obtain a measurement of -45dBm, and I
> understand that the noise density then is -105dBm/Hz.
>
>  From design variables, I was expecting a somewhat lower value, around
> -110dBm/Hz, but between the NoiseCom noise source and the output there
> are several things (attenuator, filter, amplifier, directional coupler,
> variable attenuator, ...), so perhaps there are slight differences
> between estimated insertion gains and losses accumulate up to 5dB.
> Before dismount the system and look directly at the noise source output
> and measure the losses/gains of each element, I would like to know if I
> am doing this mesasurement right or am I commiting some mistake?

Two things spring to mind:

Spectrum analyzers don't have very low noise inputs, in general.  Did 
you account for the additional noise from that source?

The noise bandwidth is not necessarily the same as the resolution 
bandwidth (the shape of the filter isn't rectangular, after all).  You'd 
have to look at the phase noise measuring ap notes from HP, I think they 
give the correction factor to use (it's more than a dB, as I recall)