[RFI] Solar Panel Manufacture Schedules Fix For Solar Panel RFI

Jim Brown jim at audiosystemsgroup.com
Thu May 17 00:17:02 EDT 2018


On 5/16/2018 8:21 PM, KD7JYK DM09 wrote:
> "it emphasizes the importance of twisted pair and shielding for all 
> wiring carrying noisy current"
>
> Which reminds me, how does this work for AC mains feeding various 
> noisy things, such as LED lighting? 

The same -- both are fundamental physical principles. In one case the 
noise is riding on DC, in the other on AC. Twisting minimizes 
differential mode coupling, shielding kills radiation of common mode 
current.  For shielding to work, it must be continuous.  Twisted pair is 
typically 30 dB or more better than zip cord, glorified or otherwise, 
for reducing differential coupling. That's how CAT5/6/7 cables work to 
prevent crosstalk between pairs. When I lived in Chicago, I ran the 
telco lines for my home and biz on CAT5; it drastically reduced 
crosstalk at baseband (audio) as compared to ordinary telephone wiring.

On the HF bands, most noise is radiated as a common mode signal.

> Would twisting romex be better, than sheathing it in ferrite, or, what 
> about just putting it in conduit?

Steel conduit provides very effective shielding if it is continuous over 
the entire path.  Wiring in my shack, and all of the new wiring I added 
to the building that the shack is in, is in EMT (thin-wall).

Twisting Romex would do little if anything to kill RF noise. BUT -- we 
wind #12 Romex through two common mode chokes to kill noise from our 
Yamaha i2000 generators for portable operations.  One choke is 8 turns 
on one #31 2.4-in o.d. toroids, the second is 8 turns on two #31 2.4-in 
o.d. toroids, and they are wired in series. This choke goes right at the 
output of the generator.

> How about twisting, AND conduit, for a 120VAC circuit?

PVC conduit does nothing, and twisting kills only the differential 
component. Only steel conduit provides shielding. And where it leaves 
the shielded partition, there must be a common mode choke.

Below are links to plots of VNWA measurements of the common mode 
impedance of the two chokes.  Orange is Z(mag), Magenta is Rs, Black is 
Xs. Plots are 500 ohms/div, zero for Z and Rs are the bottom of the 
plot, zero for Xs is the center of the graph. Cursor readouts are in the 
top right corner.  The Blue curve is S21, from which these values are 
derived.

http://k9yc.com/AC-Romex-8T1core.png
http://k9yc.com/AC-Romex-8T2core.png

I use a specially built test jig that W6GJB built for me; it places the 
choke (or other impedance I want to measure) in series between the VNA 
output and the VNA input; S21 is the "gain" of that path, and provides 
the voltage divider ratio between the choke and the 50 ohm input Z of 
the VNA. Math functions built into the VNWA software (by DG8SAQ) works 
the voltage divider equation backwards to find the complex choking Z = 
Rs +jXs, and plots all three quantities.

73, Jim K9YC


More information about the RFI mailing list