It is normal for the light modem to be hot to the touch. In summer, the light intensity decreases and the Internet speed drops. Three measures for heat dissipation have been tested.

During the high temperature wave in mid-July, my home network had a strange problem.: The internet speed dropped from 940MB to around 760MB every afternoon, and then returned by itself in the evening. The modem light is all green, and restarting the router has no effect. The night I squatted in front of the weak current box to check, I put a bare cat shell on the back of my hand - it was hot to my hands, even hotter than a router. Enter 192.168.2.1 When I looked backstage, I saw that the light had declined, which was 3 dB faster than during the physical examination in the spring.

Overheating of the optical modem is not the same as overheating of the router

I wrote a special article about router overheating. The symptoms are that when the machine capacity is high, the CPU cannot handle it, WiFi drops out, and the background spins. The treatment method is to move the router, put it up, and install a fan. optical modem are different. The most delicate thing in the optical modem is not the main chip, but the laser that is always emitting light - your data relies on this beam of light to reach the optical splitter in the corridor and then enter the computer room. If you want to know which router set, you can read this article: Is it normal for the router to get hot and scalding your hands? , let’s talk about optical modem specifically.

Lasers are particularly sensitive to temperature. This is a physical characteristic. As the temperature rises, the luminous power drops. If the light is weaker, the signal received at the other end of the computer room will be weaker, and the "received optical power" value you see in the background will become larger - that is, the light attenuation will be greater. So if the router is overheated, it will talk to you about lag, and when the optical modem is overheated, it will talk to you about light failure. Light decay is the thermometer of optical modem.

Let’s look at another hard reference: The nominal operating temperature of optical modem is generally written as 0 to 45 degrees. I looked through the manual of Huawei HG8145V, and it clearly states that if the operating temperature exceeds 55 degrees, the heat dissipation environment must be checked. But in a closed weak current box in summer, 45 degrees is not difficult to encounter - the three-piece optical modem, router, and switch in the box heat together, and the box is exposed to the sun, and it becomes a small sauna.

The time line of my house that week, the light decay follows the box temperature

My weak current box is in the balcony wall. It was cleaned up last year and the optical modem was moved to a lower position in the box. But the wall of the box was exposed to the sun in the afternoon. During the week of continuous high temperatures in July, I hung a kitchen probe thermometer in the box and recorded the box temperature at two o'clock every afternoon. I went backstage to read the light attenuation and speed measurements for six days.

The rules are very strict: The higher the box temperature, the greater the light attenuation and the uglier the speed in the afternoon. July 18th was the most typical day when the box temperature was 49 degrees. In the afternoon, the light attenuation climbed to -24.6 and the speed was measured at 760. ； At ten o'clock in the evening, the box temperature dropped back to 41 degrees, and nothing happened. The light decay returned to -23.2 by itself, and the speed returned to 890. The leather cord has not been moved or the spectrometer has been moved during this day. What has changed is the temperature. This trend of "bad afternoon, good evening" is the signature of temperature-based light decay.

Why does the light attenuation increase when the temperature is high?

Talk about principles in vernacular. The normal transmitting optical power of the laser in the optical modem is between 0.5 and 5dBm, and it is powered on and illuminates 24/7. Lasers are sensitive to temperature since they were born in the womb: Every time the ambient temperature goes up, the luminous power drops a little, and the wavelength drifts a little. When the light is weak, the number read by the receiving end of the computer room will be poor, which is reflected in the fact that the light attenuation number in your background becomes larger.

As the light decays, the bit error rate increases. If there are too many bit errors, the data packet will need to be retransmitted, and the speed will drop - this is why my speed and box temperature went in opposite directions that week. When the light attenuation difference exceeds -28, it means it is not slowing down, but there are too many bit errors to directly drop the line and the LOS red light flashes. At that time, self-care is basically useless, so report it for repair honestly.

Another thing that adds fuel to the fire: ITMS pushed the configuration in the middle of the night. At two or three o'clock in the morning, ITMS issues the configuration and the Timecat CPU will be busy for a while. It is even hotter, which happens to be the worst time for heat dissipation at night in summer. I occasionally got disconnected in the early hours of that week, and it has something to do with it. I recorded a monthly account of the disconnection pattern, and I wrote this article based on the four patterns.: Why does the telecom optical modem always disconnect? .

Try these three ways to dissipate heat one by one, and you will be able to count the amount of light that fades back.

The root cause of the disease is temperature, so we should lower the temperature. The three moves are arranged in ascending order of cost. After trying each move, go backstage and copy it again.

The first trick is to open the weak current box cover and pay zero yuan. The door of the box was fully opened at night. The temperature of the box dropped from 49 to 43 in the afternoon of the next day, and the light attenuation returned from -24.6 to -23.1. The shortcomings are also obvious: It's still unsightly to leave the box open and dusty, so it can only be used as a diagnostic tool - the light fades significantly after the lid is opened, which means it's definitely a temperature issue.

The second move is to stand up the optical modem and stop stacking it. My router originally rested directly on top of the optical modem, and the two burned each other. Move the router away and place the optical modem on its side (some models have mounting ears that can be hung directly on the wall of the box), leaving a punch of space at the top and bottom. Two days later the light decay was -22.7. This trick is a free long-term benefit and is highly recommended to be done first.

The third trick is to blow the USB small fan towards it. There are more than a dozen USB fans, powered by the charging head, fixed at the mouth of the box and blowing towards the optical modem. After two hours, the light decay reached -21.9 - better than before the high temperature. This trick is experimental in nature and proves that there is nothing wrong with the optical module itself. It is purely the ambient temperature that holds it back.

My final plan now: The box cover was replaced with one with heat dissipation holes (about 30 yuan online, measure the size before buying), the fan was put away, the light attenuation was stable at around -22.5, and the 940 was still running at full capacity during the evening peak. If you find it troublesome to replace the box cover, you can simply move out the weak current box. I have written about several ways and costs of moving it.: Can the optical modem move? .

In both cases, don’t rely on temperature

The first type: The temperature is not high and the light attenuation is also large. With the cover open and the fan blowing, the light attenuation is still above -27, so it's not a matter of temperature - it's probably the optical module that has reached its age. The laser is a consumable part, and long-term high temperature will accelerate its aging. The hotter it gets, the faster it will break, creating a vicious cycle. Don't worry about heat dissipation at this time. Call 10000 to report "weak light". The master will bring an optical power meter to measure it in sections. It's time to change the cat. The optical modem is the property of the operator. It will be replaced free of charge if it is not damaged by man-made damage. Before reporting for repair, take a photo of the SN code on the sticker on the fuselage. The master can ask, what is the purpose of each string of codes on the sticker? I wrote a dictionary: What is written on the sticker of the Telecom optical modem? . My second-hand TEWA-750G was replaced once through the repair process, and no money was taken for the entire process.

The second kind: The light decay is good and bad, and it doesn't match the temperature. The readings are also bad when it's cool in the morning, so don't doubt the cat first. Squat down and take a look at the leather cord - it's pressed by the box door, bent, or tied too tightly, which will cause the light to fade. Cable management is cost-free. Smooth the cables first and then count them. Three tips for grading light failure and self-examination, I wrote them in detail before: -From 8 to -27 healthy, from -27 to -28 marginal, after -28 don’t mess around with yourself, read this article: What should I do if the light attenuation of the telecom optical modem is large? .

Now that you are in the right position, let’s talk about two more habits.

Two habits to finish off. First, the optical modem should not be powered off and restarted when the router is finished. Frequent power on and off will cause damage to the laser rather than maintenance. I have calculated this account.: Should the telecom optical modem be restarted regularly? . Second, keep the base light attenuation recorded on the day of installation and check it once every summer when it is the hottest. An increase of one or two dB a year is normal aging. An increase of four or five dB in a summer means it is alarming you. If the PON light starts flashing and cannot be registered, it is another disease. Read this article: The PON light of optical modem keeps flashing and cannot be registered. . Pay attention to the overall arrangement of the weak current box and the placement of the optical modem. Huawei officials have a counter-intuitive statement and have also written about it.: How to tidy up the messy weak current box .