What are the symptoms when the optical modem power adapter is broken? Light decay drifts and the call drops in the middle of the night; the multimeter only has 9.2V left after reading

early October, the light degradation of my optical modem started to drift again. In the morning, it was -22.8, but in the afternoon it jumped to -25.6. In the middle of the night, LOS flashed twice and disconnected the whole family. I had already treated the summer case once, replacing it with a lid with ventilation holes to secure it. This time, the temperature was clearly normal, but the problem resurfaced. After a week of turmoil, the root of the problem wasn't with the fiber optic, but with the black brick hanging on the wall.
first treated the Light Path Disease for a week, using all three old moves but nothing happened
light decay increases, my conditioned reflex is to check the optical path. Following the three self-inspection tips the light decay article, I will go through them one by one: re-untie the pigtail in the weak current box, leaving a bending radius of five centimeters; Wipe the fiber optic end twice with anhydrous alcohol and dust-free sticks; Leng Jiezi opened it up and reinserted it, clicking into place with a click. After finishing each task, I went backstage to watch the light fade.
blame is right here. After adjusting the cable, it immediately returned to -23.1. I thought it was the same old problem, but the next afternoon it rebounded back above -25. I've seen two types of Zhenguang road problems: cold joints with water ingress and aging, which is stable deterioration and keeps getting worse; The bent and squeezed type is something that looks good after straightening the wires. For this kind of drifting back and forth in the morning and evening, the other in the afternoon, none of the two things match.
the two midnight episodes were even scarier—LOS flashed red, and the whole house went offline for about two minutes before recovering on its own. I pulled out a pair of records on my phone, both times around 2 a.m. At night, no one was on the path, and the cabinet temperature was only around thirty degrees. Putting it under the network disconnection pattern none of these categories matched, I could only call 10,000 to report the fault. To put it specifically: unstable light decay, LOS flashing twice at midnight, and optical path self-check.
technician didn't touch the fiber first when I arrived, but measured the power supply first
technician brought an optical power meter, and the order of work was different from what I expected. Side test of the incoming line, it is normal; Tested the tail fiber again with the modem, and it still works fine. Both ends of the optical path were clean, so he turned and unplugged the optical modem, pulling out a multimeter.
The
measurement method is simple: set to the DC voltage setting, place the red probe against the DC connector core, and the black probe against the outer barrel. The first shot was no-load, and the meter showed 11.6V. I thought this was pretty normal. Then he plugged the power back into the optical mode, measured the probe online, and when the number dropped to 9.2V, it stopped. The rated 12V power supply saw the load drop by nearly 30%, so the cat being able to turn on is considered tough.
technician said this is called "normal load collapse under load." The electrolytic capacitors inside the power supply have aged, but without load, they can still hold up the storefront, but once loaded, the flaws are exposed. I swapped in a universal 12V 1A plug from his bag, and with a load capacity of 12.1V, I checked the backend and saw the light degradation—it immediately dropped back to -22.1. It took me less than ten minutes to go from unplugging the power to seeing the numbers.
insufficient voltage, why does light degradation cause the damage
I couldn't figure this out at first: the power supply powers the motherboard, and if the voltage isn't enough, the system might crash and restart. How could it be reflected in light degradation? After researching, I realized that the most delicate part of the modem is the optical module. The laser's luminous power is sensitive to both power supply and temperature. In summer, the heat suppressed the laser's power, causing the backend light degradation to rise. I went through this process in the article the modem heating issue. This time, it's the other end of the same chain: 12V drops to just over 9V, the motherboard can barely maintain thanks to the internal buck circuit, but the laser's luminous power has already shrunk.
your own transmission power is low, the signal from the data center looks even weaker, and the receiving optical power drops to the backend. Therefore, the light decay figure does not necessarily reflect the optical path itself; the emitter state is also packaged into it. Heat dissipation is normal, the cables are clean, and the light decay is unpredictable—it's time to consider whether the cat itself isn't full.
backend actually provides a window for discernment. In the network-side information, besides receiving optical power, there is also a line sending optical power. The health machine line stays stuck between 0.5 and 5dBm for years, untouched — no matter how bad the optical path is, it remains unaffected. If one day you see the transmitted optical power slipping down on its own, it's not just a matter of the road—it's the cat's heart lacking blood supply, and nine times out of ten, it will loop back to the power supply or optical module. This time, I checked old screenshots. During the days of the illness, the transmission power did drop from 2.1 to 1.4. I couldn't understand it at the time, but now everything matches.
there's nothing mysterious about power aging either. The shortest-lived component in switching power supplies is the electrolytic capacitor. After being exposed to low-voltage boxes at around forty degrees Celsius for years, the electrolyte gradually dries up, capacity drops, ripple increases, and the load voltage drops accordingly. Routers age year after year like this. I wrote about the same mechanism in my article router lifespan The randomly purchased power supply for the installation isn't good to begin with; it can last five or six years and is considered a normal performance. Mine is just getting old.
three ways to replace the power supply, pass through all three specifications first
confirm it's the power supply, the route becomes more flexible. My route one: When reporting a fault, they directly said, "The optical modem power output is insufficient." The technician came and brought one but didn't charge me. A colleague's family was charged twenty yuan for the same issue. Reports vary from place to place, so when calling 10,000, they first ask if they need to be charged. For Lu Er, buy it yourself. 12V 1A branded products priced between 15 and 25 yuan are sufficient, but avoid those priced at 9.9 yuan with false labels. Before buying, you can do the math: the optical modem is only about 6W on standby, 12V times 1A equals 12W, so the margin is just right, so there's no need for high power. The L3 is borrowed from a 12V power supply from a retired router at home—but before plugging it in, you have to pass through all three checkpoints one by one.
| check items | standard | speak plainly |
| voltage | 12V word for word | 9V Put away all 19V ones; if it's too high, burn out the |
| current | just not too much | 1A can be replaced by 2A, 0.5A can't be driven |
| plug | have matching inner and outer diameters and polarity | compare with the original one; the inner is positive and the outer is negative, so a half-millimeter difference is not reliable |
how big your optical modem should be? No need to guess—the bottom sticker clearly states the power panel. My ZTE F7015T has 12V 1A, and I wrote a special article about this sticker before Just to block a thought: don't try the phone charger. PD fast charging only boosts voltage with a protocol handshake. A regular cable only gets 5V, and if you can't turn it on, it's a waste of time. The details are all covered in the article on router power .
what kind of situation should you suspect the power outlet
The most frustrating thing about
power supply aging is that all symptoms go elsewhere: light degradation is like optical path disease, network disconnection at night is like wiring fault, and restarting improves like firmware issues. After suffering this loss, I have summarized four situations that should raise suspicion and compare them with them:
| symptoms | more like whose problem | first check |
| light decay fluctuates irregularly, wiping the head doesn't work | power supply | load voltage |
| after a power outage and reboot, the next day the | power supply | load voltage reappeared |
| If the cabinet temperature is normal and the casing is not hot, but light degradation is severe | don't focus solely on cooling | first measure the power supply before checking the optical path |
| light degradation will steadily deteriorate and | the optical path or module | fault report and have the technician bring the meter for testing |
two out of four were correct, first checking the power supply and then reporting the fault, saving the technician a visit. If you don't have a multimeter at home, use a substitute: borrow a matching 12V power supply and plug it in overnight. The next day, compare the light decay records—it's more intuitive than anything else. On the night you borrowed the power supply, remember to save screenshots of the light degradation numbers before and after. Whether you want to return or exchange from the seller or call for a technician, you have the right hand and speak your mind. One more reminder: during light decay and unpredictable light failure, LOS flashes and breaks at midnight. I noted in the article on LOS red light that first distinguishes whether it's a light circuit failure or a full system restart. The troubleshooting direction is very different—the rebooted cat still needs to be registered at the data center, which is half a beat late. PON registration the entire process is covered.
Conclusion: Measure before buying, five steps to complete it
this review, my order was: first record the light deterioration baseline, then keep records once in the morning and once in the evening; Excluding heat dissipation and cables, these two are the most common and easiest to arrange; Then measure the power supply, one round each with no-load and load; on-load is the real deal; Once confirmed, report the fault or buy the one with matching specifications; After the replacement, I tested for three consecutive days, only after stabilizing did I finish it.
two casual reminders. Don't try to power off and restart the modem after the router is done. Each time the power goes out, both the power supply and laser get hit by a power cycle. This issue was calculated in the the modem should be restarted. If your home is disconnected, don't just focus on the optical path. I've recorded the triage sequence of power, optical path, and router in the network disconnection inspection ledger. After using the modem for five years or more, before reporting 'major light decay,' I first looked down at the black brick stuck in the wall—it was also a consumable.
