[hpsdr] NOISE FLOOR

[Scott Traurig]

Yes, but some of that jump in noise floor is due to the receiver aggregate
noise figure.

The Apache hardware has an approximate aggregate noise figure of around 14
dB or so, as best as I can measure it in the 8000DLE. We know that 6 dB of
that is from the preamp. Losses in front of the preamp (feedline,
connectors, preselectors, etc.) account for probably another 1dB of that
(it's more than 1dB of loss, but the Friis equation used in noise figure
cascade analysis mitigates a lot of that). The rest has to do with the
conversion in the ADC. That's the nice thing about these receivers, they
are so simple. Loss+Gain+Conversion Loss. After that you are in the digital
domain which is essentially noiseless (not entirely, but I hate math ;-).

You can measure noise figure yourself by attaching a dummy load, looking at
the noise, referencing it to dBm/Hz, then comparing that to the thermal
noise floor, which is -174dBm/Hz. The difference is receiver noise figure.

I'd happily take a 3dB reduction in noise figure. I live in a very quiet,
rural environment and regularly see much better than ITU-R P.372-13
predicts. In fact I typically see 10-15dB better.

73,

Scott/w-u-2-o

On Fri, Jan 18, 2019 at 9:08 AM Jim Sanford <wb4gcs at wb4gcs.org> wrote:

> This is my experience.
>
> I just ran a test:  My noise floor on 10 meters went up 20db when I
> connected an antenna.
>
> I do NOT think that I am receiver noise-figure limited.
>
> Remember that too much gain ahead of the front end degrades third-order
> intercept, etc.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20190118/789ef3ff/attachment.html>