Thanks! I understand that common mode currents are an issue with off-center-fed dipoles. But my question is specifically about the transformer continuing to pass common-mode currents if you eliminate the galvanic connection between primary and secondary:
Of course there is some leakage capacitance between primary and secondary…. But if you eliminate the galvanic connection and provide a proper counterpoise, I do not understand why significant common mode current would flow into the feedline
If you galvanically separate the primary gnd from the secondary gnd then there is only the parasitic capacitance between the windings left to conduct CM current. With a proper counterpoise (common rule of thumb is a counterpoise length of 0.05 times the wavelength), CM current through this parasitic capacitance should be minimal.
I have built my SOTA EFHW transformer boxes with galvanically isolated grounds and added a switch to bridge them to be able to experiment. I have found that with a counterpoise as mentioned above, and the switch in the “open” position (i.e. isolated grounds) I have no CM problems at all, with 40 W from my HF-90. So, this has become my default “modus operandi” on 40m and 20m.
73,
Wouter Jan PE4WJ
Who knows, perhaps there’s a catch to this explanation in the fact that RF common-mode currents always take the path of least resistance to ground, which galvanic isolation cannot significantly prevent.
Since common-mode currents are “ordinary” RF currents and the “ordinary” laws of RF wave propagation apply (frequency-dependent), your feed line might happen to have an advantageous length at 20 and 40 meters (current vs. voltage on the outside of the shielding, VF of the shielding material ~90-97%)?
I use endfed antennas at 100W in my curent QTH - I have the feedline as counterpoise but use a separate coax balun to effectively isolate the RF from the shack.
For portable, I use a separate counterpoise, regardless of what coax I am using
Rick