24 V 10 A power supply—how can they get away with this?

The second power supply I ordered for review—and, honestly, it managed to surprise me. Actually, no: it wasn’t the power supply itself that surprised me; it’s relatively decent. What surprised me was the sheer audacity, but you really have to see it for yourself.

To stay up to date with new posts on the blog, I recommend following me on Instagram. I’ll post notifications about every new article there, and maybe write about something interesting too: link to my account

You can also subscribe to updates and new articles on the Telegram channel - https://t.me/KirichBlog

So, along with the previous power supply, I ordered a second one. It looked quite nice and was rated at 24 V and a whopping 10 A. Sure, I stopped believing in Santa Claus, pink ponies, and the like a long time ago, but what arrived exceeded all expectations.
The power supply currently costs $22. I think it was a little cheaper when I ordered it, but not by much. Either way, I wouldn’t call that cheap.

No complaints about the packaging: neat and reasonably secure.


On the outside, it’s an ordinary, fairly large brick, measuring 170x72x40 mm.


It has a flat-pin plug and a DC5525 connector with a split center contact.


The label claims the power supply outputs 24 V at up to 10 A.


I realized something was wrong right away, and it came down to the weight. Anyone who’s ever held a laptop power supply with roughly this much power will have noticed how heavy it is—it’s basically a compact brick. This one, on the other hand, was light.
I even dug out a power supply from a PoE switch. It was about the same size and weight, but there was one catch: that PoE power supply is rated for half the power.


Before testing, I opened it up—and found a pleasant surprise: the case is held together with clips, so it’s easy to take apart and might come in handy for something later.


The board inside is clearly used, but the mains connector is what made me laugh.


The LED clearly wasn’t part of the original design, so someone added it later. And judging by the soldering, the person who did it was a level-80 soldering wizard.


All in all, it comes apart pretty easily, if you can get past the double-sided foam tape.


The board is from a reputable manufacturer, and I have absolutely no doubt about that. It’s all controlled by an L5991 PWM controller—a rather interesting chip that even has a synchronization input for running multiple controllers in parallel. That’s unusual.
The transformer is clearly too small for the claimed power, as are the heatsinks. Judging by eye, it’s good for 70 W at most.


There’s a proper input filter, though it could really use a varistor. Interestingly, the two common-mode chokes are simply connected in series, with no capacitor between them.


Even the transformer’s insulation tape shows this is a quality product. There are chokes on the output too: a common-mode choke and another to reduce the main ripple.


1. High-voltage transistor.
2. Current-sense resistor. It appears to be the original one, which is important. I’ll explain why later.
3. Output diode assembly.
4. Name-brand capacitors: Rubycon, 1000 µF, 35 V. Another sign that they’re original is that the potting compound holding the components in place has visibly hardened.


There’s a little board tucked between the high-voltage capacitor and the heatsink. I think it’s either thermal protection or a separate output overvoltage protection circuit powered from an auxiliary winding. I didn’t remove it.


There’s a shield on the bottom, but I was more interested in the markings on it: model name PS-180 and a manufacturing date of week 33, 2004. In other words, this power supply is almost 20 years old!


I took the time to look up what this power supply was, and almost immediately found an Epson printer power supply rated at... 50 W—or, more precisely, 24 V, 2.1 A. Good grief.


There’s nothing particularly interesting under the shield.


Actually, no—remember I mentioned the current-sense resistor? Well, the Chinese often like to replace it, or even more often, simply add another resistor in parallel with the original, thereby raising the protection trip threshold.
In this case, the resistor is original, and nothing is connected in parallel with it. The danger was that by changing the protection threshold, you could simply burn out the power supply while expecting it to deliver what the label says. Yes, sometimes you really can raise this threshold, but we're talking about 20–30%. Although I've seen a unit where the current (or, more precisely, the power—though those are related) was raised from 150 to 380 W, and it worked just fine. But that's more the exception than the rule.


1. The input capacitor is also a Rubycon. Based on the measurement, it was 180 µF; the voltage marking was visible: 400 V.
2. The output capacitors measured quite a bit less than 2000 µF, although that may be because I measured them while they were still on the board. But judging by the date code, I’m more inclined to think they’ve simply worn out.


I switched it on, and the unit started up without any problems. The only thing is, startup takes a couple of seconds.
Next, I connected it to an electronic load and started testing.


To find the maximum power, I set the limit to 3 A more or less at random—and it turned out to be right on target: the power supply managed 2.9 A, then went into protection at 3 A. Voltage regulation is decent, though not perfect.


While I was at it, I measured efficiency at load currents of 0.5, 1, 1.5, 2, 2.5, and 2.9 A. At 2.9 A, though, I had to take the measurement quickly—the power supply goes into protection almost immediately.
The efficiency is pretty mediocre, if you ask me.


Next came the heat-soak test. The power supply worked fine at 2 A, but trying to load it to 2.6 A after it had warmed up caused it to shut down. I lowered the current to 2.5, 2.4, 2.3, and 2.2 A, with the same result. It only kept running at 2.1 A.
The board wasn’t in its case; it was lying on the table with the transformer facing up and half the case placed over it. So it’ll be even worse inside the case.


1. Temperature after 20 minutes at a load current of 1 A.
2. After 20 minutes at 2 A.
3, 4. After some more time at 2.1 A.

The temperatures aren’t great either. The hottest component was the transformer, which reached about 90 °C. In a closed case it would be closer to 100 °C, so yes, 50 W is the maximum for this unit.


High-frequency ripple with no load and at 1, 2, and 2.5 A. I expected it to be lower; even the 200 W unit from my previous review had significantly less ripple.


And here’s the low-frequency ripple. You can see that with no load, the controller switches to “green” mode and generates bursts of pulses. Otherwise, well, I expected better here too.


After inspecting and testing it, I had the obvious question: who are all these people giving this little box 5 stars?


By the way, I browsed AliExpress and saw that power supplies of this size are usually rated for up to 120 W, which is much more realistic—especially since this one doesn’t even deliver 120 W: link


But you can also find listings where you can order a power supply rated for the same 24 V and 10 A, at a similar price.


This one has a lot of orders, and there’s a better option. The reviews include photos, though the highest rating I found was 24 V, 8 A.
Judging by its internals, it’s unlikely to deliver that much, but in my opinion it’s considerably more powerful than the one reviewed here. The capacitors are clearly a cheaper grade, though, and the board itself is probably not from a reputable manufacturer.


And here’s the $41 option. It has a different case and costs more, which of course doesn’t mean it really delivers 240 W, but still. Unfortunately, it has few orders and reviews.


And this is what a genuine 24 V, 5 A power supply looks like. There’s a more detailed description in the engraver review.
Note that it’s 5 A, not 10 A. Now try scaling all of this up by a factor of two.


What can I say? Sometimes there’s no limit to the audacity, and this review is proof. Take a 50 W board, stuff it into a different case, solder on a little LED, and slap on a label claiming 240 W—that’s five (five!) times more. It’s outrageous.
Yes, the board is from a reputable manufacturer, but damn... It's almost 20 years old, it's tired, it wants to retire...

Now I don't even know what to do with this power supply. The little box is nice, of course; maybe I'll build something in it, or bring this board back to life a bit, set it up for 19–20 V, and use it instead of this dismal power supply that's powering my monitor right now. Or maybe make a charger? Anyway, I don't know...

That's all for now. What do you think about all this?