Wi-Fi Dead Zones Explained: Why One Room Always Has No Signal

One room in the house just does not work. The phone still shows a bar or two, pages hang halfway, calls break up — and every other room is fine. That room is a dead zone, and it has one of four causes.

Only two of them are fixed by an extender. Working out which one you have takes ten minutes, costs nothing, and decides whether the box you buy fixes the room or sits in a drawer.

What a dead zone actually is#

A dead zone is almost never a place with no signal. It is a place where your signal has fallen close enough to the background radio noise that the link can no longer carry useful data.

Two numbers decide it:

• The signal level arriving at your device, measured in dBm. It drops as the signal spreads out and as it crosses walls, floors and furniture.
• The noise floor — the background radio energy in that room, also in dBm. In a quiet detached house, the 5 GHz noise floor typically sits somewhere around -95 to -90 dBm. In a dense apartment building on 2.4 GHz it can sit at -85 dBm or worse, because every neighbor's network is part of your noise.

What matters is the gap between the two. That gap is called SNR, and as a rule of thumb anything under 15 dB is trouble however strong the signal looks, 20 dB is the minimum for reliable throughput, and above 25 dB is comfortable for everything a home does. The numbers behind it are in what each dBm level actually supports.

So there are two completely different ways to build a dead zone, and from the couch they look identical:

1. Push the signal down — distance, walls, a router in a closet.
2. Push the noise up — a congested channel, a microwave, fourteen neighboring networks stacked on the same frequencies.

As a practical floor: below about -80 dBm most devices stay associated but do very little; below -85 dBm they drop and rejoin in a loop. A room reading -78 dBm is not dead, it is starved — and starved rooms are the ones an extender helps most.

The four causes, ranked#

This is the rough order in which each turns out to be the culprit in an ordinary home — a rule of thumb from how the physics stacks up, not a survey.

1. Distance plus what is in the way#

The most common cause by a wide margin, and the one an extender was designed for. Open air is cheap: doubling the distance in a clear room costs about 6 dB. Walls are expensive. A drywall stud partition typically costs 3 to 6 dB at 5 GHz, an interior wall stuffed with wiring and plumbing 5 to 8 dB, and brick or reinforced concrete three to four times either of those — the material-by-material numbers are in how walls affect Wi-Fi.

Add them up and the arithmetic gets ugly fast. The bedroom above the garage is not one wall away, it is a floor plus two walls away on a diagonal, which is why it fails while a room twice as far down a straight hallway is fine.

The fix: move the router toward the middle of the coverage you need. If you cannot, add a second radio on the near side of the expensive wall — an extender, a mesh node, or a wired access point.

2. The router is in the wrong place#

Almost every home puts the router where the cable enters the building, which is almost always a corner. A corner router spends half its coverage on the yard and the street.

The tells are physical, not technical: on the floor, in a media cabinet, in the closet with the boiler, or on an exterior wall at the far end of the house. Moving it to a central, waist-height-or-higher, open position can be worth 6 to 12 dB across the far half of the house — in the same league as a well-placed extender, for nothing. Do this before anything else; the specifics are in where to put your router.

3. Noise and congestion#

Raise the noise floor by 10 dB and you have deleted 10 dB of usable range without touching your signal at all. This is the cause people almost never suspect, and it dominates in apartments.

The signature is time. A geometry problem is constant — the room is bad at 3 a.m. too. A noise problem has a schedule: the evening, or every time the microwave runs. If the room is fine at breakfast and unusable at dinner, stop shopping for an extender and read what is interfering with your Wi-Fi.

The fix: change channel, stay at 20 MHz on 2.4 GHz and drop 5 GHz from 80 MHz to 40 MHz in a dense building, and move whatever you can onto 5 or 6 GHz where the neighborhood is quieter. An extender here is actively counterproductive — it is one more transmitter on the same crowded channel.

4. The device, not the house#

Underrated, and cheap to rule out. A 2.4 GHz-only smart plug at the edge of coverage, a TV with its antenna against a metal cabinet, or a phone clinging to the router it associated with two rooms ago.

The tell is a mismatch: your phone gets 200 Mbps standing in the doorway and the TV three feet away buffers constantly. That is not a coverage problem. Move the device, get it out of the metal enclosure, or wire it.

Prove it is coverage before you buy anything#

Run these in order. Each step eliminates a whole category, and you can stop as soon as one of them explains the symptom.

1. Plug a laptop into the router with Ethernet and run a speed test. This is your ceiling; everything wireless is a fraction of it.
2. If that wired number is already far below what you pay for, stop. No extender fixes an ISP or modem problem. Work through the ISP versus Wi-Fi test instead.
3. Stand next to the router and repeat the test over Wi-Fi. If it collapses within arm's reach of the router, the problem is the router, the band or congestion — not coverage.
4. Walk to the dead room and run it again on the same device. Write down both the speed and the signal reading.
5. Compare the two signal readings. A drop of 30 dB or more between the router and the dead room is a coverage problem. A drop of only 10 to 15 dB with terrible speed anyway is a noise or capacity problem.
6. Test a second device standing in the same spot. If device B is fine and device A is not, you have a device problem, not a house problem.
7. Repeat the dead-room test at a different time of day. Big swings between morning and evening point to congestion or ISP contention, not geometry.
8. Pause everything else in the house — the 4K stream, the console download — and retest. If the room recovers, you were out of capacity, not signal.

When an extender is the wrong answer#

An extender repeats what it hears, so it can only solve a signal-level problem, and only when there is somewhere decent to put it.

• The room fails on a schedule. That is noise. Adding a repeater to a congested channel makes the floor slightly worse for everyone.
• The wired test at the router is already slow. Nothing downstream can create bandwidth that never arrived.
• There is nowhere in between with a good link. If every spot between the router and the dead room already reads worse than about -67 dBm, the extender has nothing clean to repeat. Run a cable, or use a mesh kit with a dedicated backhaul radio.
• The dead room needs real throughput. A single-radio extender roughly halves the speed a device would otherwise see. If the room is a home office pushing uploads all day, perfect placement still may not clear the bar.
• The space is detached. A garage or garden room across 40 feet of open ground and two exterior walls is an outdoor access point or a buried cable, not an indoor plug-in extender.
• Only one device is affected. Fix the device.

If none of those apply, you have a plain path-loss problem, and an extender is a genuinely good, cheap answer — as long as it goes in the right place, which is closer to the router than you would guess.

The fix, in order of cost#

Work down this list and stop when the room works.

1. Free: move the router; raise it; get it out of the cabinet; change the channel; move the failing device.
2. Cheap: one extender, placed where the router still reads about -60 to -67 dBm.
3. Better: a wired access point — an old router in bridge or AP mode on the end of one Ethernet run beats any wireless repeater.
4. Whole-house: a mesh kit, when three or more rooms fail and you want roaming that works, not a patch on one room.

Map it before you spend anything#

All four causes produce the same complaint, which is why so many people buy the wrong fix. Causes one and two are pure geometry, and geometry is the half you can settle on paper. Range Up is a floor-plan simulator for exactly that half: trace the rooms, set the material preset to what the house is built from, stand the router on the spot it really occupies, and every room comes back with a predicted level. That is enough to separate cause one from cause two — a bad room that sits 15 dB down has a different answer from one that sits 35 dB down, and dragging the router across the plan shows you for free whether relocation alone closes the gap.

If it does not, mark that room as the one that matters and let the app weigh outlets against it, then flip between the coverage you have and the coverage a unit there would give you. Nothing here touches your live network — it is arithmetic on your drawing, run on the phone. And the discouraging answer is the useful one: a best position that barely moves the room means path loss is not what is wrong with it, and no purchase changes that. Put it on your phone and draw the floor you are arguing about, then take the answer to the placement rules.