Joe,

To be honest, though I have a portable system intended to approach the
propagated noise floor over LF-HF I continue to learn about impediments
to achieving this.  My current best method, subject to change, is to
only use a WiFi interface to the Kiwi. For a BBG/Kiwi I use the "YT"
attached via USB, as documented on the kiwiSDR forum.  For the BBG/AI I
use its built-in WiFi. These are colocated with a very symmetric 2 m
reference dipole active antenna of known characteristics and high common
mode rejection.Even so, the power management IC within the BB's as well
as other digital activity on those boards is still perhaps visible in
some situations.  For some measurements I have constructed a 3D printed
AzEl rotor, also interfaced by WiFi and without any connections which is
also used to learn about polarization and direction of signal and noise
sources.  I am leaning toward using only linear power supplies to
convert to the various voltages needed by the Kiwi, preamps and rotator.

Doing it this way, the RPI has no wired connection to the kiwi, either
by power supply or network conductors. This is important to reduce the
possibility for common-mode current injection which can degrade the
noise floor.  I'm tending to use an access point (but not the routing
function built into YT and similar wireless/routers) to communicate with
the stand-alone Kiwi/AxEl rotor system and either the RPI or monitoring
laptop which can be located at significant distance so don't contribute
to the measurement data.

Glenn n6gn


On 7/14/21 10:17 AM, VE3VXO wrote:
> Hi Glenn
>
> Thanks for the very quick and informative response.  Since I have no
> experience with the Pi, before I buy one do you have any other
> recommendations given that I intend to run this from a battery and am
> concerned about shielding emissions from the Pi while keeping it cool,
> the typical plastic cases they push at the Pi outlets are probably not
> what I want.  I see the Pi 4B requires 3.1A, I was initially
> envisioning running the Kiwi on one of those 1MHz buck converters
> which are relatively RF quiet but I guess that idea won't work for the
> Pi-4B as the SMPS is rated for 3A max.
>
> Joe