Power bank for 8 18650 cells, or almost

A few power banks for 8×18650 cells passed through my hands, and I ended up bringing one home to my workbench “for experiments”—which is what I’ll be telling you about in today’s mini-review.
I should probably warn you right away that this power bank really has just one notable feature: it can hold a large number of cells. Otherwise, it’s pretty basic and doesn’t really do much.



The review features the green model, although there are four color options: white, pink, green, and blue. I ordered all of them, but only the green one made it into the review :)


Five stars for the seller’s packaging right off the bat: they came in bags of five, each unit individually wrapped, with the main body and cover packed separately. In the photo, I’ve pulled the box out of its wrapping to show how it was packed.


The package includes the bottom half with the circuit board and contacts, the cover, and a small piece of clear plastic for the display window. In my case, all the “windows” were in one of the packages.


The first thing you notice is its rather substantial size: 179×79×21 mm. You definitely wouldn’t call it small.


The negative contacts have springs. They hold the cells securely enough, but there probably isn’t enough room for protected cells—or you’ll have to force them in.


The positive contacts are simply stamped out. They’re mostly fine, but one of them had a crack, although it doesn’t really matter much.


There’s no reverse-polarity protection, so be very careful when installing the cells to avoid any “special effects.”


Also, the case has no dividers to keep the cells in place individually. If you don’t fill it completely, the cells can shift around and rattle inside.
So either install all eight, or fill the empty space with something.


There’s a small LCD on top that shows the charge percentage, operating mode, and either the connected output or an approximate charging current (more on that later).
It’s designed to be viewed from above, since from the side you can see all the indicator segments. It also has the classic eye-searing blue backlight, which is really bright.
The backlight comes on in several situations:
1. When the first cell is installed
2. When a load is connected
3. When a charger is connected
4. When the button is pressed.
5. During an overload (in this case, the backlight flashes).

In every case, the backlight turns off automatically after a few seconds, so it won’t be annoying for long—which is something to be thankful for.


The following connectors are on one end:
1. USB output — 1 A
2. USB output — 2 A
3. 2× microUSB ports for connecting a charger

Although both output ports are powered by the same converter, they aren’t fully in parallel, since each has its own load-connection indicator.
When I ordered it, I thought the two microUSB ports were there to increase the charging current and that it had two internal charging controllers. As it turned out, the ports are simply wired in parallel, although even that arrangement can help.

And of course, there’s a flashlight. What self-respecting Chinese manufacturer would make a device without one? :)
It turns on and off with a double press of the power button. The light is mediocre and comes from two directional beams, so it’s really only for emergencies.


The plastic “glass” comes separately, although I don’t see why they couldn’t have glued it in place at the factory. I haven’t glued it in yet, since the power bank wasn’t for me.


For the same reason, I didn’t fully assemble the case. The clips are quite tight, and it’s even harder to open afterward, so I wouldn’t recommend opening and closing the case unnecessarily.
This type of power bank is handy if you have lots of inexpensive cells or leftover cells from laptop batteries. The currents are low, so even relatively “tired” cells will work. The only requirement is to check the cells for self-discharge. If one of them has high self-discharge, it will drain the whole battery pack. So be careful when selecting cells.


The design is extremely simple: there are two contact strips along the sides of the case, soldered to a small circuit board mounted at one end.
The board has an indicator and a power button. The control logic is also very simple: a short press turns the power bank on, a long press turns it off, and a double press turns the flashlight on or off.
The power bank can also turn itself on when a load is connected to either output. After that, a long press will turn the power off.


I removed a couple of small screws and took out the board. I didn't bother desoldering the contact strips; I took it out “as is.”


The board is, of course, made in the “Chinese style,” although it's actually quite neat.


The voltage converter is on the underside, along with all the connectors and LEDs.


1. An unidentified microcontroller is installed next to the indicator, along with a button, the converter shutdown transistor (bottom right), and another transistor whose function I don't know (top right). Someone in the AliExpress reviews complained that theirs had burned out—literally burned through. In my case, it didn't even get warm, so I suspect that person had a short circuit somewhere on the board.
2. This IC controls power to the USB ports. Interestingly, though, it also got warm during charging tests, even with nothing connected to the ports. It stays cool in standby.
3. A step-up voltage converter, apparently with synchronous rectification, since I couldn't find an external diode. Unfortunately, I couldn't find a datasheet for any of the components used, which is a shame, since I've seen these ICs for sale.
4. The inductor is wound with two wires and doesn't overheat during operation.


I actually traced the board, hoping to figure out which component was the charger, and so far I've concluded that the converter itself also handles charging, since I couldn't find any other candidates. That's why I was curious to see the datasheet for this IC.


The connectors look like they're soldered properly, although from above the soldering looks poor. I reflowed them all during my tests anyway, which improved performance slightly, though not by much.


The connectors on the board are also marked 1A and 2A, although in practice they're absolutely identical; the center contacts are simply shorted together.


1, 2. When connected to a charger, the display shows the operating mode (IN) and charge percentage.
3, 4. During charging, the output current indicator is also active, but it serves a different purpose here: it shows the approximate charging current. For the 2A port, a reading of around 1.2 or higher means about 2A; for the 1A port, up to 1.2A.
5, 6. In discharge mode, each port has its own indicator. The charge percentage is, naturally, the same for both. The readings are reasonably accurate, but after a full charge/discharge cycle, the display may show anything right after the batteries are installed.



When a charger is connected to one port, the charging current is about 2A. Using both inputs gives roughly the same total current, but it is split between the two ports depending on the quality of the connected cables.

Charging continues up to 4.21 V, and the power bank shuts off at 3.17–3.20 V when fully loaded. The reason is that the protection transistor drops up to 0.2 V in this situation, leaving about 3.0 V at the converter input. If you reduce the output current, the power bank shuts off when the batteries reach 3.05–3.10 V.


It can't charge and power devices at the same time; in this mode, it only charges. The photo shows 0% charge, but it behaves the same when fully charged.


And, of course, a few tests.
I tested the 2A port first. Why? I don't know.
1. Auto-shutoff kicks in below 150 mA.
2. With a 1A load, the output voltage was 5.01 V, which is perfectly fine.
3. But at 2A, it dropped sharply to 4.42 V.
4. Then it dropped all the way to 3.95 V. The current in the photo is lower than in the previous test because I was “playing around,” raising and lowering the current to see how the power bank responded to the load.
5. I could only get 4.75 V at 1.86A. But keep in mind that I measured the voltage not directly at the output connector, but through the cable to the load; the actual voltage was about 0.04–0.05 V higher.
6. The power bank shuts off above 2.04–2.05A.


The second port surprised me a little. I expected it to perform even worse since it's marked 1A, but things turned out somewhat differently.
1. Auto-shutoff here kicks in at a lower current, around 100 mA.
2. At 1A, the voltage was the same as last time: 5.01 V.
3. At a current of 2.009 A, the output voltage is 4.78 V
4. I showed the reading to two decimal places above for a reason: at just 2.006 A, the output voltage rises to 4.83 V. There seems to be a threshold beyond which the output voltage drops sharply. The 2 A channel behaved the same way; its threshold was simply a little lower.
5. The protection trips at almost 2.3 A, which is 250 mA higher than on the 2 A channel.
6. When the protection trips, the power bank flashes its backlight for a couple of seconds, then completely cuts power to the output. It can only be turned back on with the button.


A more direct comparison.
The 2 A channel is on the left, the 1 A channel on the right. At the same current, the voltage on the 1 A channel is noticeably higher. And that's despite the actual voltage at the connector output being 50 mV higher.


Now let's try loading both channels at once.
A load was connected to the 2 A channel, and a smartphone to the 1 A channel. The cable to the phone was long, and there was also a USB tester in the circuit, so the current was somewhat limited. That doesn't affect the results here, though.
1. The 2 A channel is loaded to 1.2 A, and the 1 A channel to 400 mA. The voltages look perfectly reasonable.
2. I increased the load current to 1.65 A, and the load on the 1 A channel disconnected, causing its voltage to rise.
3. Oddly, I increased the current to 2 A. The voltage dropped as expected, but on the 1 A channel it rose quite unexpectedly, even though the phone still wasn't charging.
4. I loaded the 2 A channel to its maximum, and both voltages dropped as expected. Then I reduced the load on the 2 A channel, and the phone started charging again. I don't yet know why this happens. There's voltage, but the phone isn't charging.


Overall, the output performance isn't as bad as it might seem at first. For comparison, I tested this power bank against one I'd reviewed earlier. The one under review reached the same minimum acceptable voltage at a higher current. By the way, my USB plug is clearly not the best quality; this time I didn't measure the voltage directly at the board contacts.


One last test: current draw at idle and in sleep mode.
At idle, it draws 33 mA with the backlight off and 41 mA with it on. In practice, though, the backlight-off figure is the one that matters, since the backlight turns off quickly.
In sleep mode, the current draw is very low—just 39 µA, or 0.039 mA—and automatic wake-up when a load is connected still works. Given the potentially large battery capacity, this current seems almost laughably small, comparable to the batteries' self-discharge.



I didn't measure efficiency, but under full load (around 1.8–2.0 A) and at the minimum voltage, everything gets warm, though the temperatures are quite manageable—around 70 degrees. By “everything,” I mean the converter IC, the port-disconnect IC, and the battery-disconnect transistor. The inductor reaches about 50 degrees.
At an output power of 9 W, I'd estimate the heat losses at no more than 1.5–2 W, giving an efficiency of around 75–80% at full load with the batteries almost completely discharged. This is, of course, a very rough estimate.
Efficiency will naturally improve as the load current decreases and/or the battery voltage increases.


That wraps up the tests. Here's a brief summary.
In my opinion, this power bank has its place when you don't need high current or QC support, but do need high capacity at a low price—and you have plenty of batteries on hand.
The downsides are the limited output current of just 1.8–2.0 A, automatic shutoff at a fairly high current (100–150 mA), and rather stiff latches.
The upsides are the price, overload protection, low current draw in sleep mode, and automatic wake-up.


I hope you found this review useful. As always, feel free to ask questions in the comments.