A 500 W charger—so this is what you’re like, Abibas
To get ahead of myself, it turned out to be better than I expected :)
A reminder: to keep up with new posts on the blog, I recommend following me on Instagram. I’ll post notifications about new posts 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
The power supply arrived packed so well, you could probably play football with it :)
Inside was a cardboard box containing the unit itself. I was pleased to see it was even wrapped in plastic, so it definitely hadn’t been opened after arriving from China.
But the first thing I noticed, especially given that the unit was wrapped in plastic, was the dust I found as soon as I removed the wrapping. A bit odd.

The form factor will be familiar to anyone who’s ever chosen a power supply: a standard aluminum box that also serves as a heatsink.

1. Interestingly, these "Chinese" power supplies have a terminal cover, while the Mean Well units I showed recently don’t...
2. The model is listed as S-500-14.2, which means 14.2 V, 35 A, or 500 W. Those specs alone are enough to tell you this isn’t just a power supply, but a charger.
3. The two trimmers are another clue that this is a specialized device: current adjustment on the left, voltage on the right.
4. There’s also a 115/230 switch, but it’s hidden under a sticker, which is a good thing.
The seller offers these power supplies with output voltages of 12, 14.2, 24, 29.2, 36, 48, 54.6, 58.8, 60, 67.2, 71.4, 72, 75.6, 80, 88, and even 110 V!

And here’s the "special feature" I mentioned.
The unit is trying to pass itself off as a product from a well-known company, so let me explain.
People often ask me: there are Mean Well units on AliExpress, so how do you avoid buying a fake?
It’s simple: Mean Well products aren’t counterfeited, at least I’ve never come across a fake. But similar-looking labels are used, and an inattentive buyer could easily mistake one for the real thing.
On the left is one example of this imitation; on the right is the correct logo. There may be minor variations—for example, some listings write Meanwell, MeanWell, or Mean Well—but only one version is correct.
This is because there are laws against counterfeiting, and believe it or not, they’re enforced in China. A complete copy of a product is one thing; something that merely looks similar is another. Yes, this mostly applies to products from Chinese companies.
Yes, I know Mean Well is from Taiwan, but how many people in China know that Ukraine and Russia are different countries?
In short, if you see a logo that exactly matches the one shown on the right, it’s genuine. If there’s any difference at all, it’s an imitation.

The dimensions and weight are nothing special.

As always, all the interesting stuff is inside.

The layout is probably similar to 99% of power supplies like this. It’s hard to come up with anything new here.

1. The input has an almost complete filter; the only thing missing, perhaps, is a varistor. The fuse is also a bit undersized.
2. The circuit design is very similar to that of ATX power supplies. There’s an auxiliary power supply that powers the main inverter and the fan. This design allows for a wide adjustment range, while the PWM circuit is galvanically connected to the output. Isolation is provided by the auxiliary supply and the inverter’s gate-drive transformers.
3. The output transistors and their galvanic isolation transformer.
4. The output diode assemblies and the output smoothing inductor.

Another thing I noticed was dust inside the case. That left me with the strong impression that this was either a used power supply or one that had been repaired. Since it was sealed, that’s how it must have been when the seller had it too.

I’ve reviewed a couple of pretty good 600 W and 800 W power supplies, and if you look closely, you’ll see they have a lot in common: the same auxiliary power supply, MOSFETs at the inverter output, and, in fact, almost identical circuit designs. But what I really liked—and what this one doesn’t have—are the extra heatsinks on the transistors and diode assemblies.
Review of the 600 W model
Review of the 800 W model

Let’s check the voltage. For the first startup, just to be safe, I’ll use a lamp in series.
Here you can clearly see the advantage of a power supply with an auxiliary power source: the output voltage can be adjusted from practically zero to the full voltage. The maximum setting is a little over 14.6 V, which is good—you can’t accidentally turn it up too far, since 14.6 V is the full charge voltage for a 4S LiFePO4 battery.
The adjustment is smooth and linear, so you can set the voltage with sufficient accuracy.

The voltage doesn’t depend on the load current, at least from zero to 30 A; my load won’t pull any more. And, strictly speaking, even if it could, testing a charger with a load that only has CC mode isn’t valid—you should test it in CV mode. But my load is rated for just 150 W, and on loads like “Atorch,” CV mode is software-controlled and poorly implemented.

But it’s not all smooth sailing. During testing, I found that the output voltage has a negative temperature coefficient. That’s normal, in principle, although on the other hand, if the coefficient were positive, I could warm up the power supply first, set the required voltage, and then, as it cooled down, the voltage would drop, making things easier on the battery.
Though I only noticed later that this power supply has active cooling, and I was running 30 A through it with the fan off :)

To avoid stressing it too much, I did the first test at 20 A, without active cooling. You can see that as it warmed up, the voltage dropped from 14.60 V to 14.515 V. Twenty minutes after removing the load, it had risen to 14.575 V. It probably eventually reached the set 14.60 V, but I didn’t wait.

I ran it like this for about half an hour. My only real complaint was the snubber getting hot, but the 600 W and 800 W models I mentioned above had the same issue.

At 20 A without active cooling, the resistors reached 138 °C. The input thermistors were next on the list, but they’re supposed to get hot.
The output diode assemblies were the third noticeably hot spot, but again, the power supply was running without active cooling.
The transformer and output inductor, on the other hand, were doing great, so I think things will be much better with the fan running.

As I mentioned, my load can’t pull the full current, so I took the easy route for further testing: I used a battery, drained about 60 Ah from it, and used it as the load.

The output current was initially set to 36 A, and everything worked fine. What’s more, the fan didn’t even start during the initial tests, so I wondered whether I’d actually connected it.

It turned out that the current dropped as the unit warmed up, reaching 35.36 A after a relatively short time.
The lowest adjustable setting is about 1.1 A. I couldn’t check the maximum because this isn’t my power supply, and I was genuinely nervous about turning it up. The highest setting I tried was about 43 A.
After it had warmed up a little, I set it to exactly 35 A and moved on to the tests.

The test took about 1 hour and 40 minutes in total. During that time, the charging current quickly dropped to 34.5–34.6 A and stayed there. I didn't wait for charging to finish. The battery has no protection board, so I disconnected it when I saw the voltage start rising rapidly.
A few words about the fan. First, it's noisy. Second, it's very noisy. Third, it runs in on/off mode, with intervals of about half a minute, maybe a little longer, and it's noticeably annoying....

For obvious reasons, I didn't try removing the cover while it was running to measure the temperature. The most I can show is an overall photo and the temperature of the hottest component I could reach with the thermal camera: the output diode assemblies.
Subjectively, the power supply gets warm, which is understandable, and the fan blows out noticeably hot air, but I wouldn't say it's overheating.
What I think it could use is some additional heatsinks. As I showed above, the case is moderately warm around the transistors, but around the diode assemblies it's hot enough to burn your fingers—which makes sense, since they're nearly 80 degrees.

After completing the tests, I took the power supply apart for a closer inspection—and it turned out I was right to do so. Oh, very right.

At first glance, everything looks neat and tidy, with a bit of a Chinese touch, of course, but still.

1. Mains filter. This photo is mainly to show that the board is designed for at least the 450 and 600 W models.
2. Y capacitors
3. Thermistors and the inverter's output capacitor.
4. Bridge rectifier. In operation, I wouldn't say it's doing badly—it gets warm, that's all. But in my opinion, a heatsink wouldn't hurt, especially since there's room for one.

The PWM controller section is pure classic: the familiar TL494. What's notable is that the 600 and 800 W models I showed earlier used the SG3525, which, while not exactly new either, is still newer than the TL494.

By the way, speaking of the TL494, it's at least 40 years old, according to this Chinese Wikipedia article (I couldn't find it in other languages) -
The auxiliary power supply is built around the more modern Viper22A integrated PWM controller

1. The power stage uses a pair of 23N50 transistors. These are 500 V transistors rated for 23 A, with an on-resistance of 0.21 Ω—more than adequate for this circuit.
2. The snubber circuit and the resistors that get hot.
3. The output uses a pair of MBR40100 diode assemblies, rated for 40 A and 100 V. That seems a bit low here, given that the output current is 35 A.
4. There is also a capacitor between input ground and the output common wire, but it’s an ordinary high-voltage capacitor, which is perfectly fine in this case. Although Mean Well does sometimes use Y capacitors here too.
I want to point out that the board is laid out quite well. For example, on the left in the last photo you can see a mounting hole. The trace doesn’t just go around it; there’s also a protective slot cut into the board.

The current, voltage, and temperature control circuit is built around an LM324 quad op-amp.
To be honest, I haven’t studied the circuit in detail, but I can already say that:
1. Three of the four channels are used.
2. The wires from the trimmer run toward the PWM controller. They’re the ones visible in the photo near the diode assemblies.
3. Most likely, this op-amp simply amplifies the shunt signal, while the PWM controller handles the comparison and regulation.
4. The voltage adjustment trimmer is connected to this op-amp, which is really odd. Yes, this allows linear adjustment from zero to maximum, but in principle the PWM controller itself could do the same thing.
5. If you need to change the fan hysteresis settings, look right next to this op-amp.
I was going to redraw the circuit, but then I realized there’s really no point. A photo is enough for repairs, and if you want to build your own, it’s easier to make your own circuit.
There are five shunt resistors. By the numbers, that works out to 7 A per shunt, but something deep down tells me these are 10 A shunts, and this charger is simply a simplified version of the 50 A model, where each shunt would indeed carry 10 A. The fact that the power supply could deliver more than 43 A also points that way, but as you can imagine, this is just a guess.

The input capacitors are rated at 820 µF, 200 V. When measured, they came in somewhat lower, at 735 µF, giving a total filter capacitance of about 370 µF, which is perfectly adequate for this charger.
The output capacitors are rated at 2200 µF, 35 V, for a total of 6600 µF. The measured value was higher, though the parallel load resistor may have affected the reading.
A reminder about something I’ve noticed: if the capacitors have an obscure or unknown name—in other words, they’re clearly Chinese—the capacitance is probably fine. But if you see a counterfeit or a knockoff of a brand-name part, there’s a good chance the capacitance will be low, often by a factor of 1.5 or even 2.
In the second photo, you can see a small diode (or a TVS diode, though that’s unlikely) between the capacitors and the terminal block. I only noticed it while writing the review. I wondered where it was connected, and it turned out to be across the output. Now I just need to figure out why, because it can’t really protect anything there... Any ideas?

But when I looked at the underside of the board, I realized I was right to take it apart.
Overall, I have no complaints about the board itself: it has protective slots, and the output voltage doesn’t sag with load current, which means the feedback sensing points are in the right places. But in some spots, it’s so filthy...

At this point, I was even more convinced that the power supply had been repaired before.
1, 2. Inverter transistors
3, 4. Diode assemblies.
5. Fan control transistor circuit and current adjustment resistor connector
6. The output inductor isn’t soldered very neatly.
After carefully inspecting the power supply and the areas where work had been done, I came up with my own theory.
The power supply was probably working, and judging by the amount of dust, not for very long. Then it broke, was sent back, and the customer was sent a new one or given a refund. This one then went in for repair.
It was repaired and sat in storage for a while (judging by the dust on the case). Then, when demand for chargers suddenly surged, it was put up for sale.
So what might have happened to it? Let’s try to reason it out.
1. There’s evidence of work around the fan control transistor and connector, so perhaps there was a bad solder joint and the fan stopped working.
2. What happens when the fan stops? The unit overheats. The hottest semiconductors here are the output diodes, so they overheated and one or more failed short-circuit (at least one).
3. And what happens when the output diodes fail short-circuit? Sometimes the inverter transistors blow, which is what we see here.
In my personal opinion, this is the only sequence of events that could have led to these signs of repair.

Naturally, I cleaned the board, resoldered all the odd-looking and untidy spots, and reinforced the high-current output traces with 2.5 mm² wire. 1–1.5 mm² would have been enough, but 2.5 mm² was all I had on hand.

I also made a couple of small modifications.
1, 2. Snubber resistors. I reversed the position of one resistor so it sits a little farther from the capacitor and the other resistor. This should also spread the heat more evenly across the circuit board. For context, heat is also dissipated through the leads, especially the shortest one. Before the modification, both resistors were heating the same trace on the board; now each will heat its own, so the temperature should drop a little. They will also heat each other less and get better airflow. I expect the temperature to drop by about 10 degrees (but I can't say for sure).
3, 4. While reassembling the unit, I noticed that one of the diode assemblies wasn't making good contact with the case. Not only was the clamp clearly off-center, but part of the diode assembly's case extended beyond the raised section of the housing. I had to take it apart again, heat the assembly's leads, and tilt it slightly toward the neighboring one. The pressure is now more even (though it could be better, but that would be more complicated), and the assembly is now pressed against the case across almost its entire surface. Good thing I took photos before the modification, so we can compare “before” and “after.”

The usual conclusions.
Subjectively, it's a decent charger. It passed my tests and works normally. The current and voltage do drift a little, but not enough to endanger the battery, so you can ignore it. I suspect the problem lies with the trimmer resistors and their temperature coefficients, but I can't say for sure.
Yes, there was a vague feeling that “this thing is about to blow up,” but the charger behaved itself, staying quiet most of the time and, most importantly, doing its job. Just keep in mind that it doesn't disconnect the output when charging is complete; in other words, it uses the most basic CC/CV algorithm.
As for extra features, you can use it as an adjustable power supply—essentially a basic bench supply. With a few modifications, you could even turn it into a proper bench power supply.
What I didn't like? At the very least, the noisy fan and its ridiculous control circuit. If you keep the unit in a living space, it'll drive you crazy—fortunately, charging doesn't go on forever.
Build quality: I can imagine it was fine originally, but it seems to have been repaired, and rather carelessly, which is why it looks the way it does in the photos.
In short, this charger costs about 33 dollars including shipping, and in my opinion it's worth the money. If you're handy with a soldering iron, a few simple tweaks can make it a little better and more reliable. I've covered all my ideas in the review.
That's all from me. I hope you found this useful, and thanks to the charger’s owner for the bonus :)
17, https://www.kirich.blog/obzory/bloki-pitaniya/1151-zaryadnoe-ustroystvo-moschnostyu-500-vt-tak-vot-kakoy-ty-abibas, зарядное устройство из блока питания компьютера на tl494, схема зарядного устройства на tl494, 5
Questions and answers
What are the rated specifications of the S-500-14.2 charger?
It is rated at 14.2 V, 35 A, and 500 W. The author could set its output to slightly above 14.6 V.
How did the S-500-14.2 perform under load?
Output voltage did not depend on load current from zero to 30 A. At 20 A without fan cooling, it dropped from 14.60 to 14.515 V as the charger warmed up.
Can the S-500-14.2 be used as a laboratory power supply?
It can be used as an adjustable power supply, and the author says modifications could turn it into a laboratory supply. Charging uses a basic CC/CV algorithm and does not switch off the output when charging ends.
The figures from this review are in the measurement database: Chargers
Discounted products
You may be interested in
Discounted products
All the content on the site and the advice are free, but we’d be grateful for your support of the project and the channel!
Comments3
Only registered readers can comment. It takes a minute to sign up, and you can log in with the same account you use on the forum.
Log inRegister




Comments are published in their original language.