From time to time one hears or sees advice that the station ground conductor from the equipment to a ground electrode system picks up noise and degrades station receive performance.
The solution offered is to shield the ground conductor. This is to mean to use a coaxial cable for the station ground conductor, connecting both the shield and inner conductor to the ground electrode system, and connecting the inner conductor only to the equipment ground, leaving the shield at that end not connected.
The popular explanation is that noise voltages induced in the shield are conducted to ground rather than being induced on the main ground conductor.
This construction forms a short circuit coaxial stub that inserts a frequency dependent series impedance in the ground conductor at the point where the end of the shield is left disconnected. The series impedance can be easily found using TLLC to calculate the impedance looking into the length of coax with an very low impedance at the far end (say 1e-6). This renders the ground conductor less effective, and totally ineffective in some scenarios.
For example, one proponent of this system insisted that this fixed his noise problems on his favourite band (20m) when he used 15ft of LMR600 in this way to connect his first floor shack to the external ground system. Using TLLC, 15ft of LMR600 with a s/c at the other end at 14.275MHz has an impedance of 10kΩ which is inserted in series in the ground conductor. Little wonder that things changed, he had inserted a 10kΩ resistance in the ground lead! At other frequencies, the impedance would be different, but by luck rather than design, this chap had completely nobbled his RF ground in the middle of his favourite band. He might have achieved much the same outcome by removing the ground conductor.
Shielding a ground conductor to prevent noise pickup does not make sense, it does not work the way it is often explained, and the way that it does work is to render the ground conductor less effective, even ineffective.
The station earth conductor should be as short as possible, substantial cross section, and the least inductance possible. Least inductance comes from large diameter round or flat strap, least bends, no sharp bends, plain conductor or stranded but not braided. Least inductance is important to good lightning performance.
A variation on the scheme above is described in the article Grounding Systems in the Ham Shack - Paradigms, Facts
and Fallacies by DU1ANV in which a capacitor is added,
connected between the
shield at the open end of the coax and the centre conductor, see the
figure above. (The article references a HamUniverse article
The explanation offered,
The grounding device utilizes a coaxial line where the ground wire is enclosed by a shield, such as RG-8 transmission line, to prevent the buildup of high voltage standing wave near the station equipment. This ground line is not length sensitive and can be used at any length without concern. It will keep out RF away from the shack.
is not consistent with conventional circuit theory which suggests that it will not 'work' as described, meaning if it appeared to 'work' it did not 'work' for the reasons given.
The configuration is easily analysed as discussed earlier in this article, though the impedance of the capacitor needs to be paralleled to that looking into the S/C coaxial stub. We need to make some assumptions, lets assume frequency is 3.5MHz, the coax is 4m long (enough to reach a first story room vertically, and lets choose a capacitor midway of 0.01µF. Using TLLC to calculate the impedance looking into the length of coax with an very low impedance at the far end (say 1e-6), Zin=0.6+j25. Now, shunting that with a 0.01µF capacitor with an equivalent series resistance of 2Ω gives a series impedance of 0.5-j0.044Ω. What purpose does this have that shorting the coax won't achieve, and if you short that end of the coax, why use the coax as it is a poorer conductor thant stranded heavy earth wire?
The vague nature of the elements of this 'design', eg not sensitive to coax length, capacitor specified over 1000:1 range, are a warning that it is nonsense.
The ground connection will be most effective if the capacitor is
replaced with a S/C, and even more effective if the coax is replaced
with a substantial flat copper strap or stranded copper conductor.
© Copyright: Owen Duffy 1995, 2017. All rights reserved. Disclaimer.