Compact QC 3.0 charger / power supply

Standalone devices that support Quick Charge are becoming increasingly common, so I decided to take a closer look at what "feeds" these "pets"—a QC 3.0 charger (power supply). It’s very compact, but you can read all about it in the review: as usual, I’ll take a look, run some tests, take it apart, and share my conclusions.


First, a brief digression. Strictly speaking, all these "chargers" are power supplies, since they lack the features of actual battery chargers. But since most users call them chargers, I’ll use that term too for clarity.

Also, a quick reminder of what QC, or Quick Charge, is.
Devices connected to a USB port usually use 5 V, which makes sense since that’s the standard voltage. Mobile device batteries are gradually getting larger, and users want to charge them faster—which also makes sense. So chargers capable of delivering up to 1, 2, or even 3 A started appearing. But eventually they ran up against the laws of physics: as current increases, more voltage is lost in the cables to your devices. USB connectors also take a beating, and when you add the widespread use of cheap, thin wires in those cables, things get really bad—literally, you can turn out the lights.

Manufacturers scratched their heads and decided that 5 V clearly wasn’t enough, so they came up with chargers using other voltages, such as 9, 12, or even 20 V. And that’s not even mentioning that they couldn’t agree among themselves, so some companies came up with their own version of the "bicycle". But the principle is the same: raise the voltage, lower the current, and it becomes more efficient to deliver power to your device. Alternatively, you can keep the current the same, but increasing the voltage also increases the power delivered.
One of the most common fast-charging protocols is Quick Charge, with two widely used versions: QC 2.0 and 3.0.


A charger that supports fast charging is essentially a regular power supply with a small chip added. The chip is controlled by signals from your smartphone, tablet, power bank, etc., and adjusts the power supply’s feedback loop to tell it to output the required voltage.
For example, if a smartphone can charge at 9 V, it first connects in standard mode. If everything is OK, it sets a specific combination of voltages on the USB connector’s data lines, and the power supply outputs the required voltage—if it can, of course.
The difference between QC 2.0 and 3.0 is a small one: QC 2.0 can output 5, 9, 12, or 20 V, while QC 3.0 does the same thing but in 0.2 V increments: 5, 5.2, 5.4, 5.6... 19.6, 19.8, 20.0.
QC 3.0 actually supports voltages below 5 V too, but they’re used much less often.


So, by and large, the difference between 2.0 and 3.0 isn’t all that significant, and 2.0 is perfectly adequate in most cases.

Anyway, I came across a small charger with QC support. It arrived in a small bag, with no instructions, colorful box, or anything else.


A little box like this—I’d even call it cute.


There’s a single red USB port on the short end of the case. The long side has a label listing these specifications:
Input voltage: 100–240 V
Output voltage:
5 V — 3 A (15 W)
9 V — 2 A (18 W)
12 V — 1.6 A (19.2 W)

You may have noticed that the maximum output power increases with voltage. This is common: the lower the voltage and the higher the current, the harder the power supply has to work. So higher voltage usually means higher efficiency, and you can get more out of the power supply.


It’s a little bigger than a matchbox, mainly because of the case thickness.


I won’t beat around the bush—let’s move on to the tests.
I used an electronic load and a USB tester for the tests. The tester let me tell the charger which voltage to output.
At first, I wanted to show all the standards it supports, but it turned out that when detecting them automatically, the charger tends to reboot. I won't blame it, since during auto-detection the tester tries to sweep the full range in QC 3.0 mode, starting at around 3 V. The experiment showed that the charger can't go below 4.2–4.2 V, so it ends up shutting down in protection mode.

With no load and the output set to 5 and 9 V, the actual voltages are 5.24 and 9.26 V, which is right within the standard's limits. That's a plus.


As the load current increases from 0 to 3 A at an output voltage of 5 V, the voltage drops by just 60 mV. In my view, that's an excellent result, especially considering that most of the drop probably occurred across the USB connectors.


At 9 V, the voltage change was even smaller: 40 mV across the full 0–2 A range.


After setting the output to 12 V, the no-load voltage was 12.3 V. At the maximum current of 1.6 A, it dropped by just 30 mV.


During testing, I found that the charger also supports Samsung AFC. At least, I can select that mode with the tester and it outputs 9 V, but I can't test it with a phone yet.


I can give the charger a solid 5 for output voltage stability, but its thermal performance is a different story—I'd even say it's rather disappointing.
Below are 9 thermal images taken at the maximum rated current, grouped as follows:
1, 2, 3 — 5 V, after 10, 15, and 20 minutes.
4, 5, 6 — 9 V, after 10, 15, and 20 minutes.
7, 8, 9 — 12 V, after 10, 15, and 20 minutes.

My goal was to see how the temperature changed over time. Twenty minutes at each voltage was enough, since the device is fairly small and therefore doesn't have much mass.
You can see a noticeable temperature rise between 10 and 15 minutes, but only a very small increase—around 1–2 degrees—between 15 and 20 minutes. This indicates that the device has reached a stable thermal state.

But a case temperature of 70 degrees is clearly too high; it burns your hand even through the plastic case.


To figure out what was going on, I had to open the case. You can do this fairly carefully by first running something like a thin steel ruler between the black and white parts of the case. Start at the power plug and USB connector, then pry the black part upward. The adhesive will give way and the cover will come off.


The case is universal, with two cutouts for USB connectors.


The input stage is very basic, although that's typical of most compact devices: there's often simply no room for an input filter. But the lack of a fuse is disappointing; a resistor is used in its place. The input capacitor is 15 µF, which is just about enough for 200–240 V and far too little for 100–200 V.


The transformer is also quite small, especially for a power rating of around 20 W. I'd say it's closer to 12–15 W, at most.
The noise-suppression capacitor is the correct Y type, so no complaints there.


Here's what surprised me. Given the advertised specs, I expected to find two things inside: a synchronous rectifier and solid-state capacitors. The capacitors are debatable—they're not all that common—but 3 A in a case this size can only be rectified properly with a synchronous rectifier.
Well, I thought there'd be at least two diodes, but there's just one: an SR5100 (5 A, 100 V). Yeah, that's disappointing.
The output capacitors are Chang, 2 × 470 µF, 16 V. That's clearly not enough for 3 A.


1. DS3118 PWM controller. I couldn't find a datasheet, but indirect evidence suggests this controller is designed for 16–30 W power supplies.
2. PWM controller power supply circuit and current-sense resistors. The power supply circuit is complicated because the output voltage varies over a wide range, so the PWM controller's supply voltage varies proportionally as well. To keep it operating stably, that voltage has to be regulated too. The circuit in the photo looks like part of a voltage regulator.
3. The transformer winding is pretty rough. I wouldn't have let this pass; I'd have given the person who wound it a slap on the wrist. The primary and secondary windings don't touch, of course, but there's no visible insulation margin between them, let alone any varnish.
4. The output stage uses a QC chip, a TL431 adjustable shunt regulator, and a PC817 optocoupler.


Although the board is made quite well—cleanly and neatly—the 2.5 mm clearance between the primary and secondary sides (around the optocoupler) is too small. It should be at least 5–7 mm.


After inspecting it, I decided to run some additional tests.
First, a load test, this time without accounting for the voltage drop across the connectors.
At 5 V, I varied the current from 0.2 to 3.2 A. The voltage stayed at 5.24 V up to 2.3 A, then began to drop slightly, reaching 5.21 V at 3.2 A. An excellent result.


At 9 V, with the current ranging from 0.2 to 2.2 A, the voltage dropped by just 10 mV, and only at the maximum current of 2 A.


At 12 V, I first tested it over the 0.2–1.8 A range, but then decided to run a second test, this time up to 2 A.
The output voltage stayed stable even under overload. Not bad, if it weren’t for one thing...
The thing is, judging by the thermal images shown earlier, the power supply clearly gets very hot. When I raised the load current to 2 A at 12 V, bringing the output power to almost 25 W, I expected the overload protection to kick in, but it never did. This could cause the power supply to overheat and fail.


Speaking of overheating, I often see reviews where the case temperature is measured during testing, as I did above. But in reality, the case temperature can differ dramatically from the temperature of the components, so I ran this test as well to demonstrate the difference.
The test was run at output settings of 5 V, 3 A and 12 V, 1.6 A. The test duration was the same—20 minutes—and the case was closed. After 20 minutes, I disconnected the load and the charger from the power supply, opened the case, and took thermal images. Of course, it takes a little time to open the case and remove the board, but I don’t think the difference is very large—just a few degrees.

The temperature inside reached around 110–115 °C (12 V on the left, 5 V on the right). The output diode and transformer got the hottest, at roughly the same temperature.


And the final test: measuring output voltage ripple. For this test, the oscilloscope was connected to the power supply’s output capacitor, while the load was connected to the second capacitor.


1, 2 — 5 V at 1.5 and 3 A
3, 4 — 9 V at 1 and 2 A
5, 6 — 12 V at 0.8 and 1.6 A.

The ripple amplitude is practically independent of the output voltage: around 140 mV at 50% load and 200 mV at 100%. That’s generally fine for 9 V and 12 V outputs, especially at 50–75% load, but it’s a bit high for 5 V. The load will work just fine; I’d just prefer to see less ripple.
By the way, I also checked whether the power supply interferes with the smartphone touchscreen. It works perfectly, although that depends on another power supply parameter.


As for the heat, without special measures such as synchronous rectification, any similar charger in such a compact case will get about as hot. For comparison, I placed a good-quality 12 V, 1 A power supply next to it below. It’s noticeably larger. Yes, it has an input filter inside that takes up space on the board, but its power rating is only 12 W.
Mean Well makes a 25 W version in the same form factor as the power supply shown on the left, but it uses a little trick: the top of the case is made of metal-plastic, which carries heat away from the high-voltage transistor and output diode.


Now for the conclusions.
The power supply has excellent output voltage regulation, supports QC 2.0 and 3.0 protocols over the 5–12 V range, and delivers the rated output current. However, it does overheat, and ripple is somewhat high at a 5 V output.
In my opinion, it’s best to use this power supply at load settings of 5 V, 2 A; 9 V, 1.5 A; and 12 V, 1 A. That’s a more realistic assessment.

One drawback is that QC devices usually don’t ask how much power the power supply can deliver. For example, the Besiter BST-K6X power bank, which I also tested with this charger, “asked” for 9 V at around 1.8 A, and the power supply got noticeably hot. I’d also count the overload protection as a drawback: it’s clearly set to a very high power level, well above what this power supply can reliably handle. That’s a very poor design choice.

I’m sure people will mention the Gocomma QC 3.0 charger in the comments. The one reviewed here has the same specifications, and it’s also identical inside. The only difference is that the high-voltage section of the circuit board is laid out slightly differently; even the components are the same.

That’s all for today. I hope you found the review useful.