RFI pickup problem with an interesting solution

[ChrisSmolinski]

My skyloop antenna picks up RFI - what appears to be harmonics from various switchmode power supplies. The pickup is not incredibly strong, perhaps -110 dBm during the daytime, when the noise floor is around -120 dBm. At night, it isn't noticeable, since background noise levels swamp it. But it's noticeable on the SDR waterfalls, and is annoying. Since the antenna goes around the yard, with the house in the middle, one thought is that it is picking up junk radiated from inside the house. Another is that the interfering signals are traveling down the coax shield. The skyloop is fed with coax running to a balun.

I tried several things:

First connecting a better ground to the incoming coax where it enters the shack. Not only didn't this help, but it usually made the RFI worse.

Next I tried putting lots of ferrite on the incoming coax. At first I thought I was getting somewhere, after putting about 20 clamp on ferrites on the coax, the RFI was pretty much eliminated. Unfortunately, so were all signals on this part of the spectrum. Apparently I made a notch filter. There was also little effect on other frequencies.

Then someone suggested grounding the coax at the balun. So I drove in a ground rod, and got some wire to connect from the shield connection of the coax at the balun to the ground rod. Again, the RFI got worse if anything, not better.

I was monitoring the results while outside, using SdrDx which can stream the waterfall over wifi. I have an iPad client app I wrote that displays it. So I can tweak the antenna outside while observing the effects, no need to run back and forth between the antenna and shack. 

While doing this, at one point I had the grounding wire attached to the balun, with the balun back up in the air, but the other end of the wire was not attached to the ground rod. And... the RFI was GONE. Or pretty much gone, it was still faintly visible, but had been heavily attenuated.  The wire is about 30 ft long, I have not measured it. This is pretty close to a quarter wavelength at 6900 kHz, where I was observing. However, the RFI is pretty much eliminated over all of HF above about 4 MHz or so (there isn't an exact frequency where it appears again, but it starts to in the lower 4 MHz region).

I am not sure why the addition of the wire helped, but the improvement is substantial. If it helped over a fairly narrow frequency range, where it was a fixed fraction of a wavelength, I might be able to come up with some theories, but it helped over much of the HF band. For reference, the balun is fed with 100 ft of RG6 (TV type) coax.

[rdla4]

Interesting solution. Perhaps the 1/4 wave wire is at the right point to act as a tuned stub of some sort?

[cmradio]

That's what I have on my FM RX antenna to cancel all the multipath rubbish

Peace!

[Lex]

I've had similar experiences using various lengths of ungrounded wire attached to either the receiver ground or to various places on my homebrewed loops made from ordinary TV coax.  I just experiment until it attenuates noise at the desired frequency without attenuating the signal.  No idea why it works.

[redhat]

It could be that it enhances the balance of your antenna system.  Think about it, no balun is perfect.  Putting enough reactance on one side of the antenna can effectively balance it, or perhaps because it is a balanced system, it forms a small receive antenna with enough sensitivity to local RFI to cancel out the trash being picked up by the systems residual unbalance.

Just a thought,

+-RH

[Duffer]

Aren't lengths of ungrounded antenna like that called a counterpoise? http://en.wikipedia.org/wiki/Counterpoise_(ground_system) I'll try it myself, noise is a pain here.

[SW-J]

Next I tried putting lots of ferrite on the incoming coax. At first I thought I was getting somewhere, after putting about 20 clamp on ferrites on the coax, the RFI was pretty much eliminated. Unfortunately, so were all signals on this part of the spectrum.

This part, just above, is downright scary; sounds like ALL your pickup is via induction between coax, the BALUN and some wire in the 'loop' (and not exclusively due to a "propagating" RF wave creating/inducing currents flowing only in the loop!)

Obviously, speaking RF ex cathedra the majority of the signal was not via differential 'signals' (RF energy) appearing at the input to the BALUN from the loop, but rather something much more involved and un-intended (i.e. not per the intended design!)

One or two 'isolation' ferrite transformers in series with the coax cable might have been tried as well, but with probably the same results as the clamp ferrites.

(Of course, I assume we all understand that the 'differential' currents flowing on the inside shield and outside center conductor of the coax (in what is called "TEM", TEM00 for "Transverse ElectroMagnetic" mode) are not in anyway affected by application of ferrites around the outside *shield* of the coax!)

my $1.25 on the subj anyway.