What do your smartphones about car USB charging think

Surely, many people noticed that some car chargers charge smartphones more slowly than network chargers. This is especially true of cheap models. In this case, you can poke a tester with a calibrated load in them and make sure that they issue their volts and amps regularly.



There are two reasons for this. But first, a few words about charging "lithium-ion", the logic of the controllers in smartphones and lies of Chinese manufacturers of charge.

Theory


Charging the "lithium-ion" process is complex and multistage. Usually it consists of two phases: at the first (main) stage, the controller delivers a constant current with a gradually increasing voltage and, after recruiting approximately 80% of the charge, switches to the pre-charge mode. In it, the battery is supplied with the maximum regulated voltage, which for modern smartphones is 4.4 V (and which cannot be exceeded even by 0.1 V), with an adjustable current. Who cares about the details on charge, discharge, permissible current and resource cycles depending on time, temperature and other parameters, you can familiarize yourself with this article, in the second part of which all this is described in detail.

Manages all these processes controller. Sometimes controllers are integrated into SoC, but external solutions are more common.





The controllers monitor several process parameters simultaneously, including the current, voltage of the source and batteries, as well as the temperature of the battery cells. But most importantly - they are able to work with one main unknown, which is the power source. Therefore, before charging, the controller performs a series of tests in order to understand the capabilities of the connected memory device and not to overload it. After checking the memory, the controller sets the current limit, which it deems acceptable and safe in each case.

The main parameter to be tested at the stage of testing the memory, apparently, is the voltage drop when a specific load is connected to the current. If the voltage of a 5-volt source has fallen below 4.8 V, the charger reduces the load by one step and again monitors the voltage. According to my observations, smartphones with a maximum regulated charge current of 1 A controllers use the following steps: 0.4 A, 0.5 A, 0.7 A and 1 A. At the same time, it should be understood that the power consumption of a modern smartphone while playing in the middle The complexity of a 3D action is approximately 2.5 Wh, which would correspond to 0.5 A when interpolated to a 5-volt power supply. Similar power consumption will be when the smartphone is in navigator mode, which is what he most often does in the car. And this means that if the smartphone is actively used,

Tests


To the question of testing, I was just pushed by the situation with two charges that were in my arsenal on one of the long journeys. The first one honestly gave out 0.7 A at 5 V, but the smartphone working in the navigator mode in a few hours not only failed to replenish the charge of its battery, but also completely exhausted it. The second test of the tests gave confident 2 A on any of the connectors, but with her the situation was completely the same. In the end, I ordered three more models on Aliexpress and checked the work of all five with static loads of 1 and 2 A, as well as with five smartphones that were at hand at that time.



Tests with statics revealed two interesting moments: the Chinese are brazenly cheating with the stated characteristics of cheap models (no one here doubted). In addition, their already low rates are reduced to completely unacceptable values ​​after a couple of minutes as the memory electronics heat up.

In order not to overload the article, I decided to remove all tests with static loads per frame, leaving for each charge only the values ​​of the maximum stable current and voltage on one port before and after warming up. But if someone wants to see all the measurements of statics with loads of 1 and 2 A, they can be found here . The only caveat - under my hands, I had a load module with a maximum current of 2 A. And if the port was able to issue more, we can only guess by the voltage level.

The testing process itself was extremely simple: a regular charge was connected to a 12-volt fully charged 7 Ah battery, and five test smartphones were sequentially connected to it, the batteries of which were charged 25-40%. For each smartphone recorded current and voltage, which were chosen by him for charging, while the charger was still cold.



There were five smartphones. Two old men: Samsung Galaxy Nexus (GT-i9250) with a maximum current consumption of 1 A and Zopo ZP-100 (maximum current ~ 700 mA). Plus, three brand new: Huawei P9 Light, Apple iPhone SE and ZTE Axon 7 mini. The first two are limited to a current of 1 A, while ZTE uses Qualcomm Quick Charge 2.0 technology and from a 5-volt source can take up to two amperes.

Here is a general table of results. The first line shows the parameters that give the chargers in the first couple of minutes. In the second line, those that are installed after some time. If the cell is OK, then in the charge-smartphone bundle, the maximum possible mode for it was used. Red highlights the cases when the smartphone included a weaker mode than the charge could initially give.
SA 2211
HZ-008
XKY-013
with screen
"barrel"
Gives out at the start
0.7 A, 4.95 V
0.7 A, 4.95 V
0.98 A, 4.98 V
1.86 A, 4.8 V
2.01 A, 5.19 V
Maximum stable mode
0.56 A, 4.81 V
0.56 A, 4.83 V
0.65 A, 4.89 V
indicators do not fall
indicators do not fall
Galaxy nexus
OK *
Ok
0.72 A, 4.9 V
0.76 A, 5.01 V **
0.42 A, 5.3 V
ZP-100
Ok
Ok
Ok
Ok
Ok
P9 Light
Ok
Ok
Ok
Ok
Ok
iPhone SE
0.5 A, 4.98 V
Ok
Ok
Ok
Ok
Axon 7 mini
Ok
0.5 A, 5.01 V
0.72 A, 5.03 V
OK ***
Ok
Minimum price, rub.
thirty
28
55
163
160

* - approximately every fifth time the smartphone installed a minimum current mode of 0.4 A.
** - on the second connector, a stable result is not more than 0.4 A at 5.08 V.
*** - on the second connector is more than 1.25 I didn’t try to take a smartphone, while the first one steadily extracted 1.6 A.

Test Summary


1. The Chinese shamelessly lie about the characteristics of their charges, since for all of them they promised currents up to 2 A (except for HZ-008).
2. There are “capricious” smartphones that do not like all charging.
3. Controllers for simple and cheap smartphones (and other devices) are the least demanding of charging (see ZP-100).
4. If the voltage at the source drops below 4.8 V, the charging controller reduces the load current by one step. In most cases, the adjustment is done only once before the start of the charging process.
5. QuickCharge 2.0 keeps track of the source parameters all the time, not just before the process begins.
6. If you buy an inexpensive charge, there is always a chance that it may not “like” your smartphone, even if it regularly provides the declared amps and volts.

What the autopsy showed


What is the reason for such selectivity? If we leave behind the scenes the fantastic nihilism of the Galaxy Nexus, which never took more than 0.7 A from any charge, the rest of the “complaints” concern the quality of the sources in terms of current and voltage. At the same time, some smartphones recognize problem charges before current and voltage begin to sink on them. Best of all, Axon 7 mini does it.



In all models of charging installed PWMs, which are responsible for lowering the voltage from 12 to 5 volts. They are fed into the company by the simplest LC filter for smoothing pulsations, and they also add one capacitor to the input, as required by manufacturers of PWMs.

In SA 2211 and HZ-008, the same PWM AD84064Q is installed, which, according to specifications, should not produce a stable current of more than 0.8 A, which we see in practice. In addition, when heated, the performance drops significantly. And what kind of PWM is installed in black charging, which showed the best results in statics and the total current of more than 3 A - remains a mystery. The chip had no marking at all.



At the same time, the Galaxy Nexus refused to accept more than 0.4 A from it. It can be assumed that the reason for this behavior is poorly smoothed pulsations. However, the LC filter in it seems to be implemented much better than in the same SA 2211. But it is very difficult to test its operation without special tools, the same oscilloscope. Perhaps I will buy some inexpensive model from the Chinese and will conduct a more detailed study.

PSI almost forgot, with the network charges about the same story. In my practice, I came across a situation where different smartphones took a different current from the same charge. But being immediately connected to their own, they used it to the maximum.

Also popular now: