[RFI] Trying to avoid Solar RFI

[Jim Brown]

On 12/16/2022 12:45 PM, James Gordon Beattie Jr wrote:

> Thoughts?

Several. Because charge current regulators control the width of a train 
of square waves, it's critical that all DC wiring be twisted pair, with 
no interruptions of the twisting where panels and batteries are 
connected. That DC circuit is 100% modulated by that wave train, so it 
must be viewed as RF, not DC. Just as Elecraft shapes its CW keying to 
optimize readability while minimizing clicks, charging waveforms must be 
optimized for efficiency and to minimize RF noise. N6KR describes what 
he's doing as a "sigmoidal" or raised cosine shape. Wayne first did that 
for the K3; Flex subsequently adopted it after ARRL tests of their 6500 
series rigs showed rather wide clicks.

  You're right to run the system at the highest practical voltage, which 
proportionally reduces the current. RF noise is proportional to the 
square of current. It's also good practice to insert effective common 
mode chokes in that wiring. The winding guidelines for parallel wire 
transmitting chokes are a good starting point. The second link for 
VHF/UHF coax chokes can be used as a starting point for parallel wire 
chokes.

http://k9yc.com/2018Cookbook.pdf
http://k9yc.com/ChokesVHF.pdf

Proper bonding of everything on the AC side is critical, and twisted 
pair should be used for the current-carrying conductors as well.

With respect to the well -- starting current for the motor is probably 
the most demanding spec. My 6kVA Honda inverter generator won't run 
mine, which is on a 100 ft well.

I've seen a few reports that Tesla battery systems are RF quiet. My 
Model 3 is quiet both driving and charging at home -- I have VHF/UHF 
running in mine, and I've seen reports from others that HF is relatively 
quiet as well. Unfortunately it doesn't run in the family -- electronics 
for my Starlink system required extensive choking to cool off.

73, Jim K9YC