24 V, 300 W power supply: competitor or companion to the “people’s” 24/100 PSU?

Today I have another review of another power supply. I realize I may have already worn you out with these, but several readers asked me about this one, and I was interested in it myself, so I bought it.
The review will mostly consist of photos, with only a little text. In addition to the tests, I’ll also describe a simple method for testing a power supply relatively safely under thermal conditions close to the limit.


So, I chose a rather interesting power supply rated for up to 300 W, with passive cooling. That’s what caught my interest, especially since I’ll soon be reviewing the device it was designed for.

Specifications:
Power supply model: WX-DC2440
Protection features: overvoltage, overcurrent, short circuit
AC input: AC170-260V
AC frequency: 50 Hz / 60 Hz
Output voltage: DC24V
Output current: 12.5A-15A
Output power: 300 W
Overall dimensions: 146 x 70 x 39 mm; the tallest component is the output diode heatsink.
Distance between mounting holes: 137 x 61.5 mm

Both the seller and the forwarding agent took great care with the packaging. They packed it better than you’d pack a vase from the Qing dynasty.


I’m sure a lot of you will say, “It reminds me of something...”


And yes, you’re right: it’s a larger version of the “people’s” 24/12 V, 100 W power supply, which I’ve reviewed before.
12 V
24 V


Next to the one under review, it looks tiny.


Dimensions from the product page.


Just a few photos from different angles.



A top view. This isn’t my photo, but the product-page photo is actually identical to the real unit.


The input has a proper mains filter and the same terminal block as the smaller version. The only thing missing here, I think, is a varistor.
The fuse isn’t visible in the photo, but it is there, installed between the terminal block and the heatsink.
The input capacitor is 220 µF, 450 V, but for a power supply designed for a “narrow input range,” it should be closer to 300 µF.


The high-voltage transistors are paired: two 20N60C3s. The smaller version uses the exact same transistor.


Quite a few components are tucked between the heatsink and the transformer.


The interwinding capacitor is the correct Y type (marked Y1). On less powerful models, I’ve sometimes seen an ordinary high-voltage capacitor instead (just to the right in the photo), which is wrong.


The transformer is genuinely large; the core measures 46 x 40 x 13 mm. At first, it might look like it’s been impregnated, but although there is varnish inside, most of it is oddly on the outside. By the way, the power supply had a distinct smell of varnish when I unpacked it.


The output uses a pair of 20200 diode assemblies (20 A, 200 V). Interestingly, the diode assemblies and high-voltage transistors are mounted with mica insulators rather than ordinary “rubber” pads, but I couldn’t find any thermal paste.


The output storage inductor, on the other hand, looks a bit small to me: 27 x 15 x 11.5 mm. It’s wound with three strands of 1 mm wire, 9 turns. The marking is AS106-125A. I couldn’t find a datasheet right away; there’s only a brief description from sellers, and another one. If you replace it, bear in mind that it’s not ferrite, but sendust.
The power supply uses a single-switch forward topology, which is why it has a storage inductor despite having just one high-voltage switch. This is due to the relatively high power rating; the old power supply used a single-switch flyback topology.


The output has three 2200 µF, 35 V capacitors from a manufacturer I don’t recognize. There’s also one tiny ceramic capacitor on the output :)


I measured the input and output capacitors. The input capacitor is marked Nippon, but “I have my doubts”; the output capacitors are close to their stated capacity.


The output terminal block gave me a bit of a laugh. It’s the same one used in that “people’s” power supply, except there it handled a noticeably lower current, so it looks a little odd here. A 560 Ω load resistor is installed between the connector and the heatsink. It dissipates about 1 W, so it gets quite hot.


The board is double-sided, but the components are mounted on one side only; the underside is empty. While inspecting the traces, I noticed that the feedback divider circuit is connected fairly correctly, at least the part I could make out.


You probably noticed the small daughterboard soldered perpendicular to the main board. It carries the PWM controller and some of its surrounding components. During my initial visual inspection, I also noticed that the board was coated with something.


And that’s exactly what it turned out to be: the board is covered in some kind of potting compound. At first I thought it was ordinary sealant or silicone, but it turned out to be nothing of the sort—the coating is very hard. I scraped as much as I could off the PWM controller, but still couldn’t see the markings. Alcohol didn’t help either; the coating is indifferent to it (meaning it doesn’t care about my alcohol), though acetone dissolves it a little. That still didn’t help me, though.
Before cleaning on the left, after on the right.


Some of the snubber components and other parts are under the transformer.

Only after soldering the board back in did I think that perhaps it was pressed against the transformer for a reason—quite possibly the PWM controller has overtemperature protection, although that’s pretty rare in Chinese power supplies....



Time to move on to the tests. First, we’ll need an electronic load, a multimeter, an oscilloscope, and, of course, our “patient.” There’s no output voltage adjustment here, but the voltage is set pretty well as it is.


The first test measures ripple amplitude.
The test was performed at 6 points: 0-20-40-60-80-100% load. A 20% load is 2.5 A of output current, or about 60 W of power.

The ripple amplitude seems a bit high to me—up to 150-200 mV peak-to-peak. But considering that it’s specified at roughly the same level for branded power supplies, and that there’s no output filter here at all, it may be acceptable.
The bigger problem, though, isn’t the ripple amplitude but its shape. If it were made up of “spikes,” even a small inductor would cut them almost completely. But the ripple is triangular, which is worse.


At 100 Hz, it’s quite tolerable up to 240 W, but at 300 W it’s clear that the input capacitor doesn’t have enough capacitance. That’s exactly what I mentioned at the very beginning of the review.


Now let’s check voltage regulation under different loads and measure efficiency.

Efficiency:
No-load power consumption is about 4 W.
20% load - 80.6%
40% load - 84%
60% load - 84.8%
80% load - 85.5%
100% load - 85.5%

In reality, my wattmeter reads a little high, so the actual efficiency is about 1-2% higher. Even so, the power losses are still significant, which means all that power is dissipated as heat in the power supply. More “advanced” manufacturers try to use synchronous rectification in power supplies like this, but this one uses ordinary Schottky diodes.

As for output voltage stability, though, everything looks very good: the difference between no load and full load was 27 mV.


Here I simply set the electronic load to draw the stated 300 W from the output and checked how much power was being drawn at the input. About 55 W of losses—or even if we call it 50 W, that’s still a lot.


A couple more tests:
1. Measuring component temperatures at different loads
2. Checking how accurately the output voltage is regulated at different temperatures.

Tools needed for the test: an electronic load, a multimeter, an electronic thermometer, and a thermal camera.


In the end, I never made it to the stated 300 W. Each test stage lasted 20 minutes, after which I measured the temperature. Even at 10 A and an output power of 240 W, it was getting very hot, and the smell of overheated varnish was even stronger—it could be smelled throughout the apartment. Still, I decided to raise the output power to about 250 W and let it run for another 10 minutes. Everything ended well in the end, of course, but this is already beyond what the power supply can handle, and if you want more, active cooling is needed.

Two components overheated: the output inductor and the diodes. The diodes are mounted on a rather impressive-looking heatsink, which is actually almost completely useless because it is designed for forced-air cooling. On top of that, there are no ventilation holes underneath it, and part of it is covered by capacitors. I had doubts about the inductor from the start: it would be better to use a larger core and wind it with 4–5 strands instead of three.


A quick tip about a few testing details.
If you're testing a power supply under extreme thermal conditions, a wattmeter can make the test a little safer.
An overheated power supply can become much less efficient as it heats up, mainly because the transformer and inductor cores get hot. To reduce the chances of the test ending badly, I connect the power supply through a wattmeter and carefully monitor its input power after moving on to each new test stage.
If the output voltage and current (and therefore the output power) remain stable, but the input power suddenly starts to rise, that's the first step on the road to the great beyond: the power supply is running away, and you should switch it off immediately. This won't guarantee a 100% success rate, but it greatly reduces the chance of burning it out.
That's exactly how I “saved” a 500 W power supply from this review.

Below are photos of the test at the highest temperatures. I didn't dare raise the power above 250 W.


Temperature after testing at 180 W.


And after 250 W


Output voltage drift as the unit warms up. This power supply has a small negative drift: I first measured the voltage with the unit cold, then warmed it up, removed the load, and measured again. The difference was just -12 mV, which is an excellent result.


The final test was overload capability. Here, the load runs in DCtest mode, gradually increasing the load current to the set value (16 A here) while monitoring whether the output voltage drops below a specified threshold (20 V here).
I got a maximum output current of 14 A, after which the protection kicked in.


The same test in real time. You can see that the power supply holds its voltage until it shuts down, after which the voltage drops almost to zero. It then returns to normal operation because the load was removed after the voltage dropped. The behavior is entirely correct.



But while preparing the review and looking through the photos, I noticed something interesting. The power supply has four 0.39 Ω current-sense shunts, but one of them is broken. It didn't break during testing, though, as you can see in the photo below from the initial photo session. If you look closely, you can also spot it in one of the first photos in this review.
So it turns out the power supply's maximum current isn't 14 A, but about 18.6 A. The product description, however, stated:


I thought about replacing the resistor, but then decided against it. The power supply already runs quite hot, so why make potential problems worse?



What can I say in conclusion? I can't give a definitive verdict: the power supply seems pretty good, and the price is reasonable too, at about twice the price of the “people's choice” model. But some shortcomings make it hard to say that everything is fine.
On the other hand, the power supply performed well up to 240 W, apart from overheating. It also handles 300 W easily, but the 100 Hz ripple increases significantly.
Overall, I think it's definitely worth considering, but it also definitely needs some work. If you plan to use it for short periods under heavy loads, I'd recommend:
1. Increasing the input capacitor's capacitance
2. Replacing the output capacitors with higher-quality ones.
3. Optionally, adding an output noise filter.

If you plan to run it under load for extended periods, in addition to the above:
1. Applying thermal paste under the transistors and diode assemblies
2. Replacing the output inductor.
3. Adding a small fan after all.


That's all. I hope this information was useful. As usual, I'll be glad to answer questions and read your comments.


This review was sponsored by the yoybuy.com forwarding service, which covered the shipping costs. Affiliate link to sign up; new users get a coupon worth 10 off 50.
The power supply costs $12 including shipping to the forwarding service. Shipping from the forwarding service depends on various factors. The power supply weighs 377 grams, according to the order page on the forwarding service's website.