Mike, WA3TTS, recently submitted details via email of an approach to mitigating AM broadcast overload in an E-probe antenna that he was successful in utilizing. He explains:
“My W1VD version E-probe, which I recently moved up 25ft in a tree with 1 meter whip, was being hammered by 3 local 50KW stations at 730, 1020, and 1080 kHz. I thought I would try some simple front end rejection and wanted a quick test solution that would not load down the HI-Z front end of the E-probe too much, so I settled on 330uh 1/2w molded choke (low Q about 20) and a 68pf NPO cap out of my stash.
There was a noticeable drop of the overall AM band levels on the E-probe output on my spectrum analyzer, so I gave it a test outdoors. The e-probe was within about 6-dB of my EWE antenna for main lobe signal comparisons on 630m WSPR2 and I was also able to capture 16 spots of WH2XND overnight with the E-probe into a 9th order 300kHz Butterworth LPF driving my somewhat highly modified Heath HD-1420 VLF converter. The only intermod I could detect was a slight amount of modulation products at 290kHz, which is the difference between the 1020 and 730 kHz stations. I used to have pretty severe intermod at 60~70 kHz, which is the difference between the 1020 and 1080 kHz stations. That issue seems to have resolved itself.
If you look at the individual XL and XC values for 330 uH and 68pf, there is some basis for these selection values. Since guru Jay, W1VD, used 68pf in his E-probe gate coupling circuit, it seemed like a good starting point. Crunching some XL and XC values shows why:
So you can see resonance near 1060 kHz as the +2198 XL ohms is largely canceled by the -2208 XC ohms. At 475 kHz, signals see about 4K ohms impedance, almost 17K ohms impedance for 137 kHz, and about 117K ohm impedance for 20 kHz.
Likely there is a better bias resistance setting with the LC shunt in place, which I should investigate to see if the 2IOP and 3IOP points improve. A quick test was to look at the second harmonic (2040 kHz) of the 1020 kHz station 4 miles away, which was 40 dB down in my somewhat overloaded FT-817 receiver. Third harmonic was just detectable in AM mode at 3060 kHz, maybe 10dB out of the noise…
Interestingly enough, an external 10dB pad in front of the FT-817 yields substantially less intermod than using the internal 10 dB attenuator. I was also able to copy several local NDBs directly with the FT-817 (no LF converter) and E-probe in daylight, but only when using my 300 kHz or 500 kHz LPF between the e-probe and the FT-817. Without the LPFs, the FT-817 was unable to hear any LF signals. The local Latrobe beacon at 210 kHz was the lowest frequency on the FT-817 where an NDB was detected using the LPF filters to manage the RF presented to the FT-817. I will have to try this at night.
It was also encouraging to see the Heath HD-1420 VLF converter work reasonably well with the LC shunt on the W1VD E-probe design. I was hearing MSK at 21.4 24, 24.8, 25.2 and 37.5 (kHz) this morning. The 40.7 kHz MSK station was in some other communication mode, WWVB was loud, and a strong carrier noted on 73.6 kHz. (My main LF/MF converter is a high level DBM design with +23 LO and 2N5109 preamp, and IF diplexer, so I have a basis for what “very good” versus “average” performance level would be,)
Next thing to investigate is using higher Q inductors in the shunt LC, and parallel LC vs shunt traps, circuit shielding, etc. A coax choke a meter or two down from the e-probe may also be interesting to observe the effects in the MF range. One half to 1 millihenry may choke off the AM BC common mode but, allow LF signals to migrate up the coax to the e-probe….”