Hunting RFI in an Apartment: Finding the Noise You Cannot Unplug
A method for locating and killing receive noise in a condo or apartment: how to find the source, what you can fix, and what you have to null out instead.
This article contains affiliate links. I may earn a small commission at no extra cost to you.
Most people in restricted spaces spend all their effort on the transmit side. Bigger wire, better tuner, a loop instead of a dipole. Then they get on 40m at night and the S-meter sits at S7 with the antenna disconnected from anything useful, and nothing they do to the antenna moves the needle.
That is a receive problem, and in an apartment or condo it is usually the binding constraint. A compromise antenna running 100 watts into a noise floor of S2 will out-perform a full-size dipole sitting in an S8 hash field. Raising your antenna gain by 3 dB is hard. Dropping your noise floor by 10 dB is often an afternoon of work with a $30 radio and a box of ferrites.
The Noise Floor Is Not One Thing
"Noise" in an urban building comes from three distinct mechanisms, and they need different treatments.
Conducted noise rides on the mains wiring and gets into your radio through the power supply or through common-mode current on the coax shield. Switching power supplies, LED drivers, and battery chargers are the usual sources. This is the category you can actually fix, because ferrite chokes work on it.
Radiated noise comes off a device or a length of wiring acting as an antenna. Plasma TVs, powerline networking adapters, some solar inverters, and long unshielded runs of DC wiring behind drywall. You can fix your own. You mostly cannot fix your neighbor's.
Broadband ambient noise is the aggregate of every device in a hundred-unit building. There is no single source to find. It is the floor beneath the floor, and it is why urban HF operators run noise-cancelling receivers or directional loops instead of chasing individual culprits.
The order of work is: find out which category dominates, fix what is yours, then decide whether the remainder is worth a receive antenna.
The Portable Radio Sniff Test
Before buying anything, establish whether the noise is inside your unit or outside it. The tool is a cheap portable AM radio, tuned to a quiet spot on the broadcast band with no station on it. AM broadcast is 530-1700 kHz. It is not your operating band, but the noise sources that wreck 40m and 80m almost all radiate strongly there too, and a pocket radio with a ferrite bar antenna is directional: null the bar toward a source and the buzz drops out.
A Tecsun PL-330 is worth having anyway, and the ferrite bar makes it a better direction finder than anything you can do with your HF rig indoors. Any pocket AM radio with a ferrite bar works for the sniff test; the PL-330 adds SSB and shortwave coverage so you can listen to the noise on the bands you actually operate.
The procedure that actually works:
Kill the main breaker to your unit. Sit in the dark with the portable radio. If the buzz is gone, the source is inside your walls and you own the problem. If the buzz is unchanged, it is your neighbor, the building, or the utility, and no amount of ferrite in your shack will fix it at the source.
If it is inside: bring breakers back one at a time and listen. When the buzz returns, you have the circuit. Then unplug everything on that circuit and reintroduce devices one by one. Walking the radio close to each device with the volume up finds the emitter in seconds because the field strength falls off fast with distance.
What Usually Turns Out To Be the Culprit
In my experience the same handful of things come up over and over in small living spaces, because those spaces are dense with cheap switching supplies.
| Source | Signature | Fix |
|---|---|---|
| LED bulbs and dimmers | Broadband buzz, drops out with the light switch | Replace with a different brand, or a non-dimmable bulb |
| USB phone/tablet chargers | Rough hash, changes when device charges | Ferrite on the cable, or a better-quality charger |
| Powerline network adapters | Dense hash across all of HF, constant | Remove them entirely. There is no fix |
| Cable/DSL modem power brick | Steady whine on specific frequencies | Ferrite on the DC lead, close to the brick |
| Aquarium or HVAC control boards | Periodic tick, cycles with a thermostat | Ferrite on the mains lead |
| Laptop charger | Hash that stops when the laptop runs on battery | Ferrite, or run on battery while operating |
Powerline networking deserves its own note. Those adapters deliberately inject wideband RF onto house wiring across the HF spectrum. There is no filtering strategy that saves you. If you have them, replace them with Wi-Fi or run Ethernet. If your neighbor has them, you are in a much harder conversation.
Ferrite: Mix, Turns, and Placement
Ferrite is the main tool for conducted noise, and most people get it wrong in the same three ways.
Mix matters. For HF (1.8 to 30 MHz), use mix 31. Mix 43 peaks above 30 MHz and gives noticeably less common-mode impedance across the lower HF bands. The generic snap-on cores sold in packs of 20 on Amazon are almost always mix 61 or an unspecified mix, which is close to useless for 80m and 40m. Buy known material from Palomar Engineers or DX Engineering and know what you are putting on the cable.
Turns matter more than core count. Impedance scales with the square of the number of turns through the core. Five turns through one large mix-31 core beats five separate cores with a single pass each. If the cable is thin enough to wrap, wrap it. A DC power lead through a 2.4-inch mix-31 toroid five to seven times gives real common-mode impedance on 80m and 40m.
Placement matters. Put the choke at the noise source, not at your radio. A choke at the radio stops noise that has already coupled onto everything else in the room. A choke tight against the offending power brick keeps the noise on the brick side of the boundary. Same principle as choking a feedline at the feed point rather than at the rig.
A working starter set: a handful of Fair-Rite 0431164181 split cores for the DC and USB cables you cannot unplug from both ends, plus two or three FT240-31 toroids for the cables you can actually wrap. Expect to spend an evening moving them around and listening.
Your Coax Is Also a Receive Antenna
If your antenna is indoors or on a balcony, the coax shield runs through the same electrically noisy space as everything else. Common-mode current on that shield gets into the receiver as signal, and no amount of ferrite on power cables addresses it.
The fix is a 1:1 current choke at the antenna feed point. This is the same component discussed in the RF grounding and counterpoise article, and it does double duty: it stops transmit RF from coming back into the shack, and it stops shack noise from radiating out to the antenna and back down as apparent signal. If you have already built one for transmit reasons, you have already solved half the receive problem.
Wound-in-place chokes from Balun Designs or Palomar Engineers are the reliable route. Twelve turns of RG-400 through a large mix-31 core also works if you want to build it.
When the Noise Is Not Yours
If the breaker test says the source is outside your unit, you have three honest options.
Talk to the neighbor. This works more often than people expect, especially if the culprit is a failing power supply and you frame it as "your thing is broken" rather than "your thing is interfering with my hobby." A failing LED driver or an arcing dimmer usually dies within months anyway.
Null it. A small receive loop with a rotatable null can drop a single dominant noise source by 20 dB or more. This is the same physics that makes the magnetic loop useful for transmit in tight spaces, covered in the magnetic loop article. For receive-only use, the loop can be much smaller and much cheaper because efficiency does not matter. A one-meter shielded loop on a rotator base in a window is enough to steer a null.
Phase it out. A noise-cancelling unit like the MFJ-1026 takes your main antenna and a small sense antenna, adjusts amplitude and phase on the sense channel, and subtracts. On a single dominant source it can be dramatic. On broadband ambient noise from a hundred sources it does almost nothing, because there is no single phase relationship to cancel. Buy it only after you have confirmed you have one dominant source, not a field.
What I Would Do First
Start with the breaker test. It costs nothing and it tells you whether to spend money on ferrite or on a receive antenna. Most people skip it and buy ferrite kits for a problem that turns out to be the building.
If it is inside: mix-31 cores on every switching supply in the room, five turns minimum where the cable allows, placed at the supply. Then a 1:1 current choke at the antenna feed point. That combination typically buys 10 to 15 dB of noise floor in a small apartment, which is worth more than any antenna change available to you in that space.
If it is outside: skip the ferrite entirely and put the money toward a small receive loop you can rotate. A null steered at the dominant source does what no filter can.
The measurement to keep is simple. Note your S-meter reading on 40m at the same hour before and after. If the antenna did not change and the meter dropped two S-units, you gained roughly 12 dB of signal-to-noise for free. On a compromise antenna that is the difference between hearing a station and working it. The indoor QRP reality article covers what that headroom is actually worth once you have it.