Some time ago, the cold wave swept through most of China. When our socialist successors were frozen, it was estimated that many mobile phone users were confused. If your friend suddenly loses contact, don't worry, they just bravely tested what is called "frozen shutdown" in person

On that day, human beings finally recalled the horror once dominated by the lack of power and the humiliation of the phone's power failure: with 20% of power left, there was only one last turn left in the wild car racing, and the WeChat apology to his girlfriend was one word short, Conan was one word short of pointing out the murderer, Taobao hack almost paid, and the phone was turned off!!! (╯‵□′)╯ ┴─┴

Even the "electricity quantity" at the bottom of Maslow's demand theory cannot be guaranteed. What else can we talk about! Even the power display has pitted us. Is there anything more insecure than this in our life? Holding the infinite grief and indignation of being trapped by the power display, we decided to study an important life proposition: is the power display on the mobile phone really accurate?

Normal temperature test

We selected five machines according to the price range for testing, including the 5000 yuan iPhone 6s (yes, the protagonist of the "freeze shutdown" case), the 2000 yuan Meizu MX5 and OPPO R7s, the 1000 yuan Hongmi Note3 and Huawei G7.

Here we introduce our copy tool Fast discharge and the Battery Monitor Widget. The former can discharge the mobile phone with a fixed load through the screen, the flash light is always on, and the CPU high-frequency operation. The latter is one of the most professional power information recording tools. Because the iPhone does not have these two software, it uses slow motion photography and flash for constant output.

Under such a fixed load, the mobile phone will power down at a linear speed. If the recorded power consumption curve is not a straight line, it is obviously that its power statistics are inaccurate.

At room temperature, several mobile phones at the 2000 yuan and 1000 yuan levels do not lose their links. Except for the lowest Huawei G7, which has a certain degree of false power reporting, almost all of them can coincide with the ideal power down curve.

Low temperature discharge test

The baby is cold, but the baby doesn't say

The chemistry teacher told us that low temperature will affect the speed and efficiency of the chemical reaction of the battery, but the electricity statistics should not be erratic or suddenly hang up. What happened to the iPhone at low temperatures? Will other mobile phones also have accidents? We still use experimental data.

The same test method and experimental object are used, but the test environment is changed to a 0 ℃ incubator:

At low temperature, the first half of the iPhone 6s only lost power faster than usual, and the curve often drifted. It took more than half an hour to reach the second half with 1%. However, because the iPhone test machine is new, there is no frozen shutdown. Netizens are frozen off because the error of electricity statistics increases after using mobile phones for a period of time.

On the 2000 yuan machine side, OPPO R7s experienced a "step by step" phenomenon when it was powered down, but its performance was much better than that of other machines. The power consumption was slightly affected by temperature. However, the situation of the MX5 nearby is not very optimistic. The machine's power consumption is falsely reported throughout the whole process. At last, about 15% of the power consumption is directly frozen and shut down.

For the 1000 yuan unit, Hongmi Note3 has the best performance among the five machines, except for a slight power failure in the early stage. However, the G7 next to it has an exaggerated deviation in power consumption. After 25% power consumption, it suddenly died

Why do mobile phones behave differently?

Apple said that the sudden shutdown of the iPhone was actually a low-temperature protection for the battery, but that passionate and unrestrained power curve made Apple's face swollen.

Of course, as young people with knowledge, culture, pursuit and Google, what we mobile phone geeks are more interested in is the reason for the difference in the statistical accuracy of different mobile phone power, so I went to check the data shyly

The whole story starts with the inconspicuous electricity meter chip on the mobile phone. You may still remember that when the Nokia Magic Brick was used, the so-called electricity statistics were only 3~5 grids (as shown in the above figure). Although the accuracy is very low, the equipment charged once a week does not need to pay much attention to the power. Later, with the popularity of real smart phones and the collapse of mobile phone battery life, the technology upgrade of the meter chip ushered in, and we also entered a wonderful era of percentage power display.

From the previous 3-5 grids to the current 100 grids, the accuracy has increased by more than 20 times. The behind the scenes contribution is the coulometer chip. However, mobile phones will have certain usable capacity fluctuations in different temperatures, different ages, and even different discharge currents (tasks with different power consumption, such as playing games and watching movies). The accuracy of electric quantity display is closely related to the electric quantity meter. However, the electric quantity meter, which affects the "sense of security" of the machine, has been forgotten by everyone. As a result, the mobile phone manufacturer did not mark the information of the electric quantity meter at all. As a result, we had to dismantle the machine.

Disassembly analysis

The electricity meter chip of the iPhone 6s was made by Texas Instruments and placed on the circuit board packaged with the battery.

The specific model is quite special. It is the SN27546 customized by Texas Instruments for Apple.

OPPO R7s circuit board excerpt, which also uses the electricity meter of Texas Instruments. The chip is placed on the independent circuit board directly connected to the lithium battery, surrounded by a lot of sticky insulation glue.

The specific model is BQ27541 of Texas Instruments.

Main board diagram of Hongmi Note3

Hongmi Note3 uses CellWise's electricity meter, model CW2015. The chip is very small and placed on the back of the power cable connector.

Meizu MX5 circuit board excerpt, the machine is not equipped with an independent electricity meter chip.

Huawei G7 circuit board excerpt also does not carry an independent electricity meter chip.

Statistical table of electricity meter models

After dismantling the phone to check the marking on the chip, it was found that only 3 of the 5 mobile phones used independent electricity meters. Among them, Apple iPhone 6s, Hongmi Note 3, and OPPO R7s use independent electricity meters, while Meizu MX5 and Huawei G7 do not have independent electricity meter ICs. The power reported by the mobile phone has to be estimated by the CPU itself. The accuracy of the power display is poor, and there is no sky error under low temperatures, and even cold shutdown is inevitable.

Comparison Diagram of Discharge Curve of MX5 Battery before and after Aging

In addition, there is also the problem of aging offset when independent electricity meters are not used. In other words, during normal use, with the aging of the battery, the error of the power display will continue to increase. For this reason, we purposely compared the new MX5 and the MX5 after six months of use. Even at room temperature, the discharge curve of the MX5 that has been used for a period of time has a very touching deviation, which has greatly affected daily use.

In actual use, if it is found that the power display of the mobile phone is getting more and more inaccurate after a long time of use, it is likely that the machine is not equipped with an independent power meter, and this type of machine is likely to have a very exaggerated power display error (don't look around the MX5, it means you). More tragically, the lower the power, the greater the error, but this is the area where we are most sensitive to power

However, the performance of independent electricity meters is also different. The Saiwei electricity meter used in Hongmi Note 3 performs unexpectedly well, while the TI electricity meter used in iPhone 6s and OPPO R7s has a different performance at low temperatures. Over the years since the freeze shutdown event, Apple has still not optimized the low-temperature characteristics of the coulometer chip, which is not very authentic.

The Last Pieces of Thoughts

Although users' awareness is not high, the accuracy of the electricity meter will directly affect our psychological prediction of a mobile phone's endurance. Accurate power indication can make us dare to make an important call when the power is 5%. If the power indication is not accurate, even if it shows 50%, we will be worried about the battery falling off the chain, which is an important topic related to security.

But unfortunately, even if it is not extreme weather, many mobile phones that use non independent electricity meters cannot provide accurate electricity indicators. In addition, if there is a 10% error in the power calculation, 10% of the battery capacity will never be used, and the mobile phone's endurance time will be reduced by 10%. The endurance of the phone that has already died cannot withstand such a toss.

Before we talk about charging for five minutes, talking for two hours, Type C, graphene and other black technologies, do you dare to optimize your home's power display accuracy first?