Hi Bruce,Very interesting, thank you for your comments and observations. The mix 31 cores are certainly worthy of experimentation in this realm.
One spot test that can be quite instructive is to touch the shield of an antenna feedline to the input of the radio. This gives - in noise terms - a very clear idea of what potential difference exists between the radio's chassis and the feedline shielding. i.e. it gives a fair idea of the worst case for common mode noise that can be expected at the frequency of interest. One can also get a similar result connecting only the feeder centre conductor to the radio - noise voltage adds in much the same way. At "PVC Towers" the noise is very significant from LF through much of HF. I think something similar occurs when you make the change to connections on the Zepp.
Careful use of isolating+matching transformers makes a world of difference. I use small 73 mix binocular cores at HF e.g. for connecting a dipole used for 22m monitoring.
I also use a LF converter, this model: (wb9kzy.com/lfconv.htm). It works well, and performance overtakes my receivers for tuning below 100 kHz. Some care is needed to avoid overload from broadcast band signals, per the comments in my first message. The converter is powered by a 9V battery and mounted in a small diecast project case. It can be used in conjuncton with the Belka receiver for a pretty compact portable setup. I think there would be room to add circuitry for a homebrew active-whip - so definitely food for thought there.
73s and Best regards
Ed K06BLM