When a router keeps dropping WiFi, the three most common causes—in order of frequency—are outdated firmware, overheating, and channel congestion on the 2.4 GHz band. Update the firmware first, move the router to a ventilated shelf, and set the wireless channel manually; those three steps clear the majority of intermittent disconnection issues. If drops persist after that, work through the checklist below in order.
Confirm the router is actually the problem
Before touching any settings, run a continuous ping from a wired device to your router's default gateway. On Windows, find the gateway first:
Look for the Default Gateway line under your active Ethernet adapter. It will be an address in the 192.168.0.0/16, 10.0.0.0/8, or 172.16.0.0/12 range. Do not assume 192.168.1.1—use whatever your system actually reports. Then start a continuous ping:
Replace that address with your real gateway. On macOS and Linux, ping runs continuously without any extra flag. Let it run for ten minutes and watch for Request timed out or Destination host unreachable lines. If they appear in bursts and then recover, the router is crashing or its radio is cycling. If the wired ping stays clean but WiFi devices drop, the problem is in the wireless radio or DHCP, not the WAN uplink.
Also check what IP address your WiFi devices are receiving. If any device shows a 169.254.x.x address, it received no DHCP lease—meaning the router either crashed or its address pool ran out. Self-assigned addresses in that range cannot reach the internet or the rest of your LAN.
Cause 1: Outdated firmware
Firmware updates fix wireless driver bugs, memory leaks that trigger spontaneous reboots, and stability regressions introduced in earlier releases. If the router has never been updated since purchase, this is the single most likely cause of recurring drops.
Log into the router's admin panel at the Default Gateway address you found above. The firmware update option is typically under Administration, Advanced, or System depending on the brand:
- TP-Link: tplinkwifi.net → Advanced → System → Firmware Upgrade
- ASUS: asusrouter.com → Administration → Firmware Upgrade
- NETGEAR Orbi: orbilogin.com → Advanced → Administration → Firmware Update
- NETGEAR routers: routerlogin.net → Advanced → Administration → Firmware Update
- Linksys: linksyssmartwifi.com → Connectivity → Router Firmware Update
- Xiaomi: miwifi.com → System Status → System Update
- Other brands: look for words like "update," "upgrade," or "maintenance" in the admin panel menus—the label varies by firmware version.
If the router runs OpenWrt, update through System → Software or use sysupgrade from the command line. On DD-WRT, use Administration → Firmware Upgrade. After flashing either third-party firmware, do a full factory reset and reconfigure from scratch—leftover nvram values from old firmware cause instability on many chipsets.
Enable automatic firmware updates if your router supports them. TP-Link, ASUS, and NETGEAR added auto-update toggles in firmware released after 2020—look under Administration or Advanced Settings. On routers without auto-update, schedule a monthly manual check.
Cause 2: Overheating
Consumer routers have no cooling fan. Enclosed in a cabinet, stacked directly on a modem, or placed in direct sunlight, the processor throttles and the wireless radio intermittently shuts off. The router may remain reachable by Ethernet but drops every WiFi client for 30 to 60 seconds at a time.
Signs: drops occur more frequently after hours of heavy use, in summer, or in the afternoon. Touch the top of the router—if holding your hand on it is uncomfortable, heat is a factor.
Fix: place the router on an open shelf with at least 15 cm clearance on all sides and above. Do not stack it on the modem. If a cabinet is required, install a small USB fan directing airflow across the chassis. Some ASUS and TP-Link models display CPU temperature under System Info or Device Info in the admin panel—check it before and after relocating to confirm the change made a difference.
Cause 3: 2.4 GHz channel congestion
The 2.4 GHz band has only three non-overlapping channels: 1, 6, and 11. In a dense apartment building, every nearby router occupies one of these, and the overlapping signals cause enough retransmissions that devices see it as a disconnection.
Diagnose the congestion on Windows:
On macOS, hold Option and click the WiFi menu bar icon, then choose Open Wireless Diagnostics → Window → Scan. Count how many networks are on the same channel as yours.
Fix: in the router admin panel, go to Wireless settings for the 2.4 GHz radio. Change the channel from Auto to whichever of channels 1, 6, or 11 has the fewest competing networks. Write down your choice—Auto channel selection can switch at 2 am and send you back to debugging the next morning.
Cause 4: 5 GHz DFS channel drops
DFS (Dynamic Frequency Selection) channels—roughly channels 52 through 144 on the 5 GHz band—are shared with weather radar and terminal radar systems. When the router's radio detects a radar pulse it is legally required to vacate the channel within 10 seconds and avoid it for 30 minutes. This produces an abrupt, complete WiFi drop that appears entirely random.
Fix: set the 5 GHz channel manually to a non-DFS option. Channels 36, 40, 44, 48, 149, 153, 157, 161, and 165 are generally non-DFS in Canada and the United States for most consumer hardware—your router's channel dropdown may label DFS channels explicitly, so check there to confirm. In the admin panel: Wireless → 5 GHz → Channel. Set the channel width to 80 MHz for a practical balance of speed and stability.
Cause 5: DHCP pool exhausted or too many devices
Every device leases an IP address from the router's DHCP server. If the pool is too small or lease times are very short, new associations fail to receive an address and the device falls back to 169.254.x.x (APIPA), which cannot communicate with anything on the network.
Check in the admin panel under LAN → DHCP Server. Look at the start and end of the pool range. A range of 192.168.1.100 to 192.168.1.150 provides only 51 addresses. A modern home with smart TVs, phones, tablets, laptops, smart speakers, and IoT sensors can easily exceed that. Widen the pool—for example, 192.168.1.50 through 192.168.1.254 provides 205 addresses. Also increase the lease time from 1 hour to 24 hours for stable always-on devices to reduce address churn.
While reviewing the connected clients list, check for devices you do not recognise. Neighbours on your network consume pool addresses and bandwidth. If you see unknowns, switch to WPA3 or WPA2-AES with a strong passphrase immediately.
Cause 6: Failing power adapter
An ageing or underpowered power supply causes voltage sag when the router's processor spikes under load. The result is a brief full reboot: the power LED flickers, all clients drop, and devices reconnect 60 to 90 seconds later. This is more likely on routers more than four years old or when a third-party adapter is in use.
Fix: replace the adapter with an original or OEM-equivalent unit matching the exact voltage and current rating printed on the router's chassis label. Third-party adapters rated below spec trigger identical symptoms. If a new adapter does not help and the router is more than five years old, failed internal capacitors produce the same behaviour—replacement of the router itself is the practical fix at that point.
Cause 7: ISP line instability
If the continuous ping to your gateway is clean but pings to external addresses drop, the problem is upstream of the router. Run a second ping to an external IP in a separate window:
On macOS and Linux:
If the gateway ping succeeds and the 8.8.8.8 ping drops intermittently, the WAN connection is unstable. On a cable modem, log into the modem's own status page—on many cable modems this is at 192.168.100.1, but the address varies by model and ISP, so check the label on the modem or your ISP's support documentation before assuming. Look for downstream power levels, SNR values, and uncorrectable codeword errors. A healthy cable modem's uncorrectable error count should be near zero or not actively climbing. Errors that increase in real time indicate a signal problem on the coaxial line or a failing splitter.
Also watch the modem's physical indicator lights during a drop window. A flashing or amber DS (downstream) LED is a clear signal to call the ISP and request a line quality check from the outside.
Cause 8: IP address conflicts
When two devices share the same IP address—typically because a static IP was manually set on one device inside the DHCP pool range—both lose connectivity intermittently as they overwrite each other's ARP entries.
Detect conflicts on Windows:
Look for two different MAC addresses mapped to the same IP. Fix by using DHCP reservations in the router admin panel—binding a specific MAC address to a fixed IP—and ensuring any manually assigned static IPs on devices use addresses outside the DHCP pool range.
How to read the router log
Most routers maintain a timestamped system log recording reboots, DHCP failures, WAN disconnections, and channel switches. Find it under Administration → Log, System → System Log, or a similar location—the exact label varies by brand and firmware version. Export or screenshot the log immediately after a drop while recent entries are still present. Look for lines containing words like reboot, kernel panic, WAN disconnected, dhcp lease failed, or channel switch. The timestamp and event type together usually identify which cause from the list above is responsible.
On Linux clients, journalctl -u NetworkManager -n 100 shows recent network manager events with timestamps you can cross-reference against the router log. On macOS, open Console.app and filter for WiFi, or run:
Platform-specific verification commands
Windows
netsh wlan show interfaces reports current signal strength (RSSI) and the channel the adapter is using. A signal below -75 dBm is marginal; below -80 dBm expect frequent drops regardless of router settings. Move the device closer or add a mesh satellite node.
macOS
Linux
iwconfig wlan0 shows link quality and signal level in real time. resolvectl status confirms which DNS server is active after a reconnect—useful for verifying that full network restoration has occurred, not just that the radio has re-associated.
Android and iOS
Neither platform includes a built-in continuous ping tool. Install a network utility app from the relevant app store. On Android 10 and later, Settings → Network → WiFi → current network name shows the gateway address. On iOS 16 and later, Settings → WiFi → tap the (i) next to your network name to see IP address, subnet mask, and the router address. If the router field shows 169.254.x.x, DHCP failed specifically on this device—toggle WiFi off and back on, or forget the network and rejoin it.
How to confirm the fix worked
After applying any change, run the continuous ping to the gateway for at least 30 minutes under normal usage load. Zero timeouts during that window is a strong indicator. Keep a concurrent ping to 8.8.8.8 running to confirm end-to-end stability as well. Check again the following evening—thermal problems worsen after the router has run all day, and interference problems peak in the evening when neighbours are under maximum load. Two consecutive clean evenings is a reliable success signal before closing the investigation.
Common misdiagnoses
- Blaming the device, not the router: if two unrelated devices drop at the same moment, the router is the common failure point, not either device individually.
- Rebooting instead of fixing: a reboot clears a memory leak temporarily, but the leak returns within hours or days. Identify and fix the root cause rather than managing symptoms.
- Switching DNS servers: changing to a public DNS resolver fixes name resolution failures but does not stop WiFi radio drops. If the ping to the gateway is dropping, DNS is not the problem. See Google's Public DNS documentation for the scenarios where a DNS server change is actually warranted.
- Blaming the ISP first: more than half of reported home internet outages originate inside the home router. Eliminate local causes before calling the ISP—it saves time for both parties.
A 2026 note on IPv6, DoH, and DNSSEC
If your ISP provides native IPv6 and the router renews the prefix delegation frequently, connected devices can experience a brief connectivity gap when their global IPv6 addresses change. Check the router's IPv6 WAN status—typically under Advanced → IPv6—and confirm the preferred and valid lifetime values are set to at least 3600 seconds to reduce unnecessary churn.
Windows 11, current Android releases, and iOS devices now default to DNS over HTTPS (DoH), which routes encrypted DNS queries directly to a public resolver and bypasses the router's own DNS server entirely. If you are troubleshooting name resolution during a drop and the router's DNS log shows no activity from an affected device, the client is using DoH and the router is out of the loop. This does not cause WiFi disconnections but can complicate diagnosis. Similarly, DNSSEC validation failures at the resolver level cause queries for specific broken domains to fail silently—but this affects name resolution only, not the wireless radio or DHCP.