Home » Radio » Radio Frequency Interference » Worst Case Wideband RFI Reduction

Worst Case Wideband RFI Reduction

worst case wideband rfi

All 4 of my test units attacked worst case wideband RFI successfully, but the performance of the cheap X-Phase really blew me away!

I thought we should start my noise reduction comparison tests at the heavy end of things. My worst case wideband RFI comes from poorly installed halogen lights in a nearby neighbor’s kitchen. When doing kitchen renovations, their incompetent electrician ignored the CSA approved 10 foot wiring harness for three lights, and did his own thing with sixty feet of #18 wire. As a result, the voltage drop across this circuit drives the switch-mode power supply crazy.

As a result, I have to live with wideband interference across several bands. Worst case is 11-12 MHz getting swamped with S6 or -100 dBm RFI. As you can see on the left, this includes 50 kHz noise peaks from the SMPS free running oscillator. Embedded in the interference you will also find a rich spectrum of 120 Hz harmonics. Neighborhood RFI doesn’t get much worse than this.

Shown above, clearly labelled, are my results from testing four noise reduction devices over the same hour and frequency band. To make this truly a worst case wideband RFI reduction test, I chose two wideband E-field antennas. I created these by putting my LZ1AQ AAA-1C wideband loops into vertical dipole (voltage gain) mode. My Main antenna is about 250 feet east of the offending kitchen. My Auxiliary antenna is 350 feet east of the noise.

As you know, most RFI from household wiring is vertically polarized. So, my vertically polarized E-field antennas are highly susceptible to this kind of interference. Thus, my description of “worst case”. (Magnetic loops in the same location are less susceptible to local RFI.)

Results were quite surprising.

Worst Case Wideband RFI Reduction Results

Amazingly, my best results came from both the cheapest and most expensive equipment. Best was the high-end DX Engineering NCC-1 Receive Antenna Phasing Controller. It reduced the noise floor by around 28-30 dB across a 600 kHz swath. But second best – and this blew me away – was the $21 RA0SMS X-Phase QRM Eliminator. Noise reduction with this little circuit was around 25 dB.

Not far behind, though were the MFJ-1026 and ANC-4 with similar noise reductions of 15-20 dB. However, these devices tended to peak the reduction across 200 kHz swath, requiring more frequent adjustment.

In all cases, signals were maintained within 6 dB when the units were switched in and out. In all cases, performance was reliable and repeatable.


  1. Mike Lanoway says:

    Hi John,

    First thanks for your hard work on this site. I have been following with great interest your posts and topics around the loop antennas, RFI, etc. I believe this test is the closet to reality you can get. I have RFI like everyone living on a small lot with an active neighborhood.

    I borrowed and tested out the 1026 at our old QTH in Panorama Hills with generally mixed results. Looking for feedback at the last CARA flea market (2018) voting on the floor was 50/50 all comments ending with a try it at your location.

    Although contrary to where my ham radio hobby is going I think I will i invest in one of the “assembled” QRM X units. At my ripeness the surface mount soldering is best left to better equipped younger eyes. Amazon as a unit from China and there are several on eBay.

    I will keep you posted.

    Be safe and play safe.

    Mike in Airdrie

  2. Alan Baker says:

    I read your missive with much interest, as I’ve been trying to use the MFJ-1026 for some time, and TBQH it’s been useless! I also have been using the LZ1AQ double mag-loop with limited success. The noise level in my urban area is around S9 and it’s widespread with one peak due to an errant wall-wart in a neighbour’s house, which she refuses to let me replace! The LZ1AQ does have some effect, as I can see the differences on the P3 waterfall/spectrum display, but although it reduces the noise by a couple of S points, the wanted signal seems to be less, when compared to the main transmitting antenna – an OSCFD.
    I would happily buy the NCC-1, but as for you there doesn’t seem to be a huge advantage over the 1026, it must be something in my setup that I’m doing wrong.
    FWIW for noise reduction and audio tailoring, I’ve been doing better at the AF end of the spectrum with a BHI Parametric unit, rather than at the RF end.
    Any thoughts or advice (apart from move house) would be much appreciated.
    Alan. G4GNX

  3. Peter says:

    Excellent work, John. I have been long wanting to see just how well the NCC1 does in adverse conditions with a good noise antenna. I also have an MFJ-1025 (1026 less noise preamp) and found it was quite good at reducing plasma TV noise from across the street. Fortunately the plasma is gone and so are the neighbours. I now fight LED SMPSs and Ryobi 1 chargers.
    vy 73 VE6KK

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.