I wouldn't really describe this as a "common mode" (although that is used) - it's really better conceptualized as a third wire which is connected to the feedpoint, as Roger said. As noted, the choke adds an impedance in the wire.
Roger's other comment is well taken - what you measure on the bench may not be what's in practice.
You can play around with this in NEC pretty easily. I just add a wire to represent the coax shield (outside) and throw a lumped load in where needed.
Unfortunately, NEC doesn't have a way to have a frequency varying load, but I've done that in Python to generate NEC input files. 4NEC2 or EZNEC might have a way to do it. FairRite publishes .xls files with the properties of the materials, which you can use to to calculate R and X.
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
routines to read ferrite data sheets
and calculate series impedance given frequency, etc
Created on Tue Aug 25 15:40:15 2020
@author: jimlux
"""
import matplotlib.pyplot as plt
import os
import numpy as np
""" typical file from FairRite:
77 material,,
Frequency(Hz),µ',µ''
1.00E+04,1989,14
1.00E+05,2001,37
"""
def readdata(filename):
""" encoding needed, because there's a "mu" character in the file, and UTF-8 dies"""
mu = np.genfromtxt(filename,delimiter=',',skip_header=2, encoding='Latin-1')
return (mu)
def toroidLzero(N,OD,ID,Length):
""" returns Lzero in nanoHenry, given dimensions in mm
"""
L0=0.0461 * N**2 * np.log10(OD/ID) * Length * 10
return(L0)
class ferriteloading:
""" class to support generalized management of ferrite properties"""
def __init__(self):
self.datatab={}
def addmix(self, filename,mixid=None):
""" adds a new mix to the library. Mix is specified as a csv file, from FairRite.
if a mix id isn't specified, the first two characters of the filename are used.
mix ids are *strings*
"""
mutab = readdata(filename)
muf = np.array(mutab[:,0])
mu = np.array(mutab[:,1] + 1j * mutab[:,2],dtype=complex)
if not mixid:
mixid = os.path.split(filename[:2])[1]
if not isinstance(mixid,str):
mixid = "%d"%mixid
self.datatab.update({mixid:[muf,mu]}) #if we make this a 2d array, then frequencies are complex
def mu(self,mix,freq):
if not isinstance(mix,str):
mix= "%d"%mix
x=self.datatab[mix]
mu = np.interp(freq,x[0],x[1])
return mu
if __name__ == "__main__":
mixlist = [31,73,75,76,77,78]
OD = 4.0
ID = 3.0
Length = 1000.
Lzero = toroidLzero(1, OD, ID, Length)
for mix in mixlist:
mixfile = "%2d-Material-Fair-Rite.csv"%mix
print(mixfile)
mu = readdata(mixfile)
f = mu[:,0]
muprime = mu[:,1]
mudoubleprime = mu[:,2]
omega = 2 * np.pi * f
R = omega*mudoubleprime * Lzero * 1E-9
X = omega * muprime * Lzero * 1E-9
plt.figure()
plt.semilogx(f,R)
plt.semilogx(f,X)
plt.title(mixfile)
plt.xlabel("Frequency")
plt.ylabel("Ohms")
plt.legend(["Real","Imag"])
plt.grid()
plt.savefig("%2d-mix.png"%mix)
""" now test the class based implementation """
ferrites = ferriteloading()
for mix in mixlist:
mixfile = "%2d-Material-Fair-Rite.csv"%mix
print(mixfile)
ferrites.addmix(mixfile,mix)
for mix in mixlist:
print(mix,ferrites.mu(mix,1e6))
toggle quoted message
Show quoted text
-----Original Message-----
From: <
[email protected]>
Sent: Apr 16, 2025 10:46 AM
To: <
[email protected]>
Subject: Re: [nanovna-users] How do I measure this Maxwell choke?
Users of a Maxwell or W2DU choke need to be aware that the effectiveness in reducing common mode current is not directly related to the magnitude of the impedance |Z| or S21 loss that they measure using a VNA.
When a balanced dipole is fed with an unbalanced transmission line, like coaxial cable, there will be RF radiation from the coaxial cable. The reason why requires some explanation. The power fed to the antenna travels along the inner conductor and the inner surface of the coax shield. The outer surface of the shield is separated from the inner surface due to the skin effect which takes place at RF frequencies. In effect the outer surface of the shield is a "third wire" which is connected to the half of the dipole attached to the shield at the feedpoint. So the antenna now consists of one wire connected to the inner conductor and two wires connected to the shield. The end result is that current is flowing on the outer surface of the shield. The common mode current level will depend on many factors such as operating frequency, the antenna geometry, length of the coax and how grounding is done at the transmitter end.
A Maxwell or W2DU ferrite choke effectively places an inductor in series with the outer surface of the shield. This inductor will have a complex impedance (resistive R and reactive X) that is determined by the ferrite mix, physical parameters of the core and the number of turns. The more resistive the better but the value of X can work for or against reducing common mode current. When considering the inductor as a "loading coil" most readers will know that this can make a radiating element have increased current flow (and hence better radiation at a specific frequency) because the inductive reactance can cancel capacitive reactance. So when measuring the Maxwell one needs to aware that measuring |Z| or S21 on the bench and making assumptions can lead to false conclusions. Making S11 measurements of R and X on a NanoVNA can be accurate to 5% or so up to several thousand ohms as I have shown in other measurement posts in this group.
The bottom line is that S11 or S21 measurements can give some insights into choke parameters but the only real test of effectiveness is to build a simple RF current meter that can slope up and down the coax and measure the actual common mode current.
Roger Need
Roger Need