QRN: How to chase down a solid S9 noise floor on 40m only

Question

I am new to this, having just set up my station. I have "very bad" QRN on 40m: S9. Here is the setup:

• High density suburb with flats and houses
• Yaesu FT450D with fan dipole at 6m-9m above ground (sloping), this is lower than a building with flats to the North East. Wires for 40/20/10m, nicely resonant, checked with NanoVNA.
• 1:1 Balun (ie choke) at feedpoint. 15 turns on FT240-52. No evidence of RF in shack.
• 22m RG58 run back to station which is on first floor
• Station earth dropping about 3m from 1st floor to 1m driven ground stake (I realise this may not be effective, as run is too long)

These are the things I have tried:

• 40m QRN is consistently > S8 mostly S9 during the day. Grainy white noise.
• 20m and 80m are only S3.
• Noise drops slightly at night but only 1S unit.
• Turning off all power in house with radio on battery reduces it to ~S7.5 so about 1 S-unit. Still a long way from level on 20m and 80m.
• I have tested calibration of S-meter on all bands with signal a generator and a 40db Attenuator. It's within spec acc to the service manual.
• There are some "VDSL" (UK broadband) training tones giving me a "birdie" every 4kHz. But this is only on some days and appears not related to the background S9 "grainy white noise floor".
• Our own broadband line runs almost parallel to the dipole (I have put as much angle as I can but it's < 20 degrees). But turning our router off (as above) doesn't remove the "main" source.
• I cannot detect the "VDSL guard bands" at 5.2, 8.5, 12Mhz
• Noise appears solid across all of 40m, ~6Mhz to about 8Mhz
• I can make contacts on 40m but only with "booming" stations.

Edit: additional info from discussion in comments below To test for radiated vs conducted noise I also performed these tests:

• I have a second FT240-52 toroid with 10 turns of RG58 at the radio end about 10 inches before the PL259. It was spare so I thought I would add it. Made no difference.
• The radio running on battery without antenna is totally quiet on all bands.
• With linear 12V power supply it is also totally fine without antenna.
• As soon as I plug in the antenna I get S9 noise on 40m.
• The same even with a "random 5m wire thrown out the window and shoved into the centre of antenna connector on the back of the radio."

Also: I was taking HTs for a walk with my son today, and we walked past one of those green (UK colours) Telecom boxes about 100m (?) from my dipole. The interference coming off that thing was enough to break through the squelch of our Baofeng HTs tuned to 2m. On FM! Similar cabinets around the neighbourhood showed no sign of interfering with the radios, even with the squelch fully off. Could be a lead?

What is my logical next step?

Run around the neighbourhood with a portable HF radio and a directional loop antenna? Requires some investment? How about a NanoSA, which might be a useful tool anyway?

If I find source, what are options: 1. work with neighbours 2. use a phasing noise canceller?

Any tips much appreciated

Update: I am in the process of building this loop: https://ka7oei.blogspot.com/2021/02/rfi-radio-frequency-interference.html The pre-amp is built and tested. Have made a wooden frame for the loop. Just needs putting together. I am hoping it will help me understand if I have one or many (major) noise sources. There is an 11kV distribution transformer about 40m from dipole, which would be a great candidate!

Update#2: Loop is built and seems to be working. I can receive distinct signals on the tinySA (see above link) and the loop is definitely very directional - able to completely null strong signals. What does it mean? I don't know yet. There is a contest on, so 40m band is drenched in "wanted" signals. :-) -- I may need another amplifier stage to look at the "noise" in more detail, but I suspected that, and left room for it on the circuit board. TBC... If anyone has experience in using such a loop with spectrum analyser please comment or add answer. I have a feeling operating the loop in a structured way to get meaningful results is going to be key.

Update#3: First test results with loop - hard to interpret: This is the Loop:

And the preamp:

And the tinySA fed from my dipole from 3-12Mhz, clearly showing the raised floor that I am also seeing on the S-meter.

And the tinySA showing the signal from the amplified loop standing under the dipole.

The loop is clearly working and showing signals. And these can be nulled by rotating, but it does not show the "raised floor". Why? I am not sure how to interpret this.

M7BTU

Answer 1

Look around your neighbourhood or Google satellite imagery for new solar panel installations. We recently discovered a solar panel installation with noisy regulators creating noise at a neighbours house.

Also as you mentioned that LED lights are also creating noise. Look for new street lamp installation nearby that is using the new LED replacement lamps.

Answer 2

If the RF noise is that strong, then you should be able to pick it up with a much smaller movable antenna. Even scrap wire or conductive tape on a large cardboard box. That may allow you to build small directional antennas (tiny loop and stub dipole) to determine if the noise is stronger from some specific directions, or a polarized antenna to determine the noise polarization, if any.

Answer 3

I'd suggest trying a different time of day, a different radio, a different antenna, and a different location; not necessarily all at the same time though. There's probably nothing wrong with your antenna, but you might get different results at different heights, or if you raise the height of the ends to the same height as the middle.

Have you tested the noise level at different times of the day? Maybe there is a lot less noise at 4 AM. That would be a clue.

You could ask other local hams who are on 40m. What are their noise levels? Are they using vertically-polarized or horizontally-polarized antennas? Your high noise level might be particular to your neighborhood, and asking several different hams might help answer that question. If a local ham reports that his or her noise level is lower, you might ask if you could test your radio in his or her shack. The problem isn't likely to be the radio, but it's possible, and testing your radio against another radio with the same antenna could help determine whether the radio is the problem or not.

You could try a portable HF radio and a loop antenna, but those can be expensive; you could also try your existing radio powered by a battery connected to a 40m dipole flung up into the trees in a nearby park or wilderness area, if you have such places in your area that allow such things. If your problem is local noise, which it likely is, then your radio and a dipole in a rural or wilderness area would likely be quite an improvement (assuming that you live in an urban area).

Once you have a better idea of the source of the problem, then you can think about potential ways to narrow the problem down further, or ameliorate or solve the it. If your problem is local noise, then there are lots of questions and answers on this site about tracking down and dealing with local RFI. Feel free to ask a new question once you know more.

Answer 4

Start by trying to determine if the noise is "conducted" or "radiated" into your receiver. Basically, "conducted" noise gets into your receiver from the power supply or a ground wire or some other cable. "Radiated" noise gets in via the antenna.

Start by powering your radio with a battery instead of the mains-powered supply. Or borrow a battery-powered shortwave radio, preferably one that accepts an external antenna. (If the alternate radio is quiet, you've just learned something very useful!)

Remove the antenna cable and all other connections to the radio. All bands should be very quiet. (If they aren't, you have a more serious problem than I would know how to deal with!) Begin adding cables to the radio, starting with the mains-powered supply and ending with a short length of coax that is open at the far end. Finally, connect the coax that goes to the antenna.

You may learn from this process that you have one or more major and/or minor culprits. What you learn will help determine the remedial steps to take. Please amend your question, or start a new question, when you have gone through this procedure.

Answer 5

You might consider an RF choke at the interface between your radio and feed line.

The choke at the feed point will help prevent common mode current reflecting from the antenna... but the outside of the shield can still be a source of noise on receive if it's not also choked at the receiver.

In fact, you might test what happens when you put a dummy load at the feed point, and then with/without a choke at the transceiver/feedline.

Edited to add: I think the next thing I'd suggest is to see if a local ham has something similar to the MFJ-1026 noise canceler (using a second antenna, mix the noise back into the line, phase-shifted 180° to blank it out). (am not affiliated w MFJ, and there are probably less expensive products out there).

Answer 6

You've received several good answers here. To Brian's answer suggesting that you should try powering it from a battery, I would add start with your own house before running around the neighborhood with a radio by turning off your main circuit breaker. This will determine whether or not the RFI is coming from something in your house, which is quite likely.

If your noise significantly decreases, then turn on one branch circuit at a time until it appears again. You may have more than one RFI source, so turn on just one breaker at a time.

A nearby amateur mostly solved his noise by unplugging a laptop charger. Any unfiltered switching power supply will do this, even if the device is plugged in and not turned on.

Answer 7

I purchased a West Mountain CLR/DSP 58407-949 noise reduction unit. Seems to help but still requires a strong signal to break through the QRN. I run an FT-450D also.