The Surface Problem: The Router Died. Now What?
In my role coordinating emergency replacements for industrial networks, I've handled more same-day turnarounds than I can count. The phone call sounds the same every time: a site goes dark, a router stops responding, and the client needs an answer before the morning shift. The first question is always the same — 'when can you get me a replacement router?' The better question is quieter: 'why did the old one die?'
When a link drops, swapping the router is the right move. It restores service. But if the replacement goes into the same enclosure, connected to the same power source, the replacement is just a countdown timer. That's the part most people don't want to hear.
The Real Problem: Why It Dies Before It Should
The RMA reports I've seen usually say something polite like 'power supply failure.' That's true but incomplete. The power supply fails because of small components inside it. Electrolytic capacitors are the chronic weak point. They contain chemistry that changes with time and temperature. That chemistry is the hidden reason many industrial routers don't reach their expected lifespan.
Here's something vendors won't put on the spec sheet: capacitor life is tied to temperature. A rise of 10°C can roughly halve electrolytic capacitor lifespan. That might be acceptable if the router sits in a cool room. But it's a disaster when the same router lives inside a sealed metal enclosure next to a motor drive.
Capacitors: The Chemistry Inside Your Router
Electrolytic capacitors smooth out power supply ripple. As they age, their capacitance drops and their equivalent series resistance rises. The result is a power rail that looks fine on a voltmeter but collapses under load. The router starts experiencing random resets, intermittent reboots, or link drops that seem to come from nowhere. By the time the router won't power up at all, the capacitor is usually far past its usable range.
That's why testing matters.
How to test a capacitor with a multimeter
If you're troubleshooting a router and you suspect capacitor trouble, here's the practical version. Disconnect the capacitor from the board. Discharge it carefully. Set your multimeter to capacitance mode. Measure across the leads, and compare the reading to the value printed on the capacitor body. A reading more than 20% below rated is a strong warning sign. A reading near zero means it's finished.
One caveat: this test doesn't measure ESR. A capacitor can look fine in capacitance mode and still be bad under load. For intermittent faults, an ESR meter is a better tool. But if you don't have one, a multimeter in capacitance mode still catches the most obvious failures.
Enclosures: The Place Thermal Problems Hide
The enclosure is the last thing a network engineer thinks about and the first thing that should worry them. A router rated for industrial temperatures can only survive if the air around it stays within specifications. Inside a black metal enclosure sitting in the sun, the internal air temperature can be 20°C or more above ambient. I've seen enclosures mounted directly above heat-producing machinery, sealed without vents, with cables crammed in so tightly that no air can move. The router inside those enclosures wasn't failing because of the router. It was cooking.
Good enclosures are thermal systems, not just boxes. Vents, convection paths, and sometimes fans matter. If you can't touch the router's surface with your hand for a few seconds, the capacitors are already in a bad place.
The Semiconductor Is Only Part of the Chain
There's a reason Semtech-based hardware is popular in the IoT space. The LoRa radio chips, including the semtech sx1262, are designed for low power and excellent sensitivity. The datasheet lists performance figures for sensitivity and current draw across different spreading factors and bandwidths. But the chip's excellent specifications only matter if the board around it is clean. A noisy DC-DC converter or a badly matched antenna can make the chip look like a cheap radio.
The same logic applies at the router level. I've deployed Semtech routers in remote industrial sites where a 5G or LTE connection was the only practical link. The routers themselves were reliable. The failures that happened later weren't caused by the router or the LoRa transceiver. They were caused by the environment — a bad power source, an overheating enclosure, or an ungrounded antenna mount.
Because a router is a system of systems. The RF section doesn't matter if the power supply fails. The power supply doesn't matter if the enclosure can't shed heat. The enclosure doesn't matter if the capacitors have already aged past their limits. So when you're evaluating a vendor, don't ask only what chip they use. Ask about the board layout, the thermal testing, and the capacitor brand. Those details are exactly what show up five years later.
The Cost of Ignoring This
In May 2023, I was called to a cold-chain operation where the monitoring link had been dark for four hours. The site stored food products with strict temperature requirements. The replacement router was delivered in under six hours. The router itself cost less than $300. The lost inventory was over $18,000. The penalty clause in the client's supply contract added another $5,000. The original installation had used a cheap enclosure and no one had ever checked the capacitors in the power supply.
That's not an isolated story. I've seen this pattern repeat more times than I'd like. The expensive emergency could have been avoided with a decent multimeter and a better enclosure design. That's the uncomfortable math. The cheapest parts of the system caused the most expensive failure. And the more remote the router, the more expensive the failure becomes, because someone has to drive there and touch the hardware.
The Short Fix
I'll keep this short because by now you already know the direction.
First, learn how to test a capacitor with a multimeter. It takes two minutes, and it catches a whole class of predictable failures before they turn into emergency replacements.
Second, treat the enclosure as part of the router. If it's too hot to touch for more than a few seconds, you're baking the electronics. Improve airflow, add vents or a fan, or choose a larger enclosure.
Third, when you spec new hardware, look at the entire system. Semtech routers are a good example of hardware designed as a system, with radios, interfaces, and protection circuits tested together. But even the best router won't survive a bad environment.
What was best practice in 2020 may not apply in 2025. The fundamentals of electronics haven't changed: heat kills components, and capacitors dry out faster when you push them beyond their thermal limits. But the execution has to be better. A router is not just a box with an Ethernet port. It's a set of small decisions about capacitors, enclosures, and whether anyone on the team knows how to test the parts that fail first.
