/       /       /    Four slim 12 V power supplies
Support the project on Patreon


Four slim 12 V power supplies

Go to store
It just so happened that quite a few different power supplies ended up on my operating table. And as you can imagine, they didn’t get there for no reason. Since there’s a reason, why not take a look at what they’re like?

A reminder: to keep up with new topics on the blog, I recommend following me on Instagram. I’ll post notifications about every new topic there, and might also write about something interesting - link to account

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

The reason these and other power supplies came my way was that a good friend asked me to help sort through the huge variety of Chinese products, from large factories and smaller ones alike. In short, if we’re going to sell Chinese power supplies here, let’s sell ones that are better made and last longer than the rest of what’s available in our stores. I think it’s the right thing to do, and I’m always happy to get involved in something like this. I also liked the person’s approach.

Recently, I’ve tested several dozen different power supplies, and I can say that price isn’t always the deciding factor. In other words, buying a more expensive unit doesn’t mean it’ll definitely be better. You can find flaws everywhere, even in Mean Well units. But flaws come in different kinds.
Component-related - when the manufacturer uses a decent circuit and PCB layout but cuts costs on the components, their quality or quantity
Design-related - miscalculations in the design, calculations, and so on
Construction-related - for example, inadequate cooling. The components and circuit may be decent, but poor thermal design reduces the power output.
Safety - one of the key criteria. For example, a unit may be well made and work reasonably well, but the person who laid out the circuit board didn’t leave enough clearance between the primary and secondary sides. This can happen deliberately to save money, or accidentally, due to an error.

Since it’s possible either to influence the manufacturer and get them to install the right parts in the right places, or simply to choose another manufacturer, why not pick products that are better than the rest? That’s essentially what we’re doing: finding power supplies with the best balance of price and quality. For many buyers, a power supply is just a little box with some numbers on it and a price tag. Few people think about what’s inside.


So, here come the power supply reviews. Since there are quite a few of them, I’ve grouped several together in each review, and I’ll start with the slim models.

First up are four slim power supplies, all rated for 12 V. The only real difference is their power rating: 100, 150, 200, and 300 W.


The units look very similar: aluminum cases, 53 mm wide and 23 mm thick. The only difference is the length, which is 187, 220, 273, and 286 mm, respectively, depending on the power rating.
12 V, 100 W, 8.3 A
12 V, 150 W, 12.5 A
12 V, 200 W, 16.7 A
12 V, 300 W, 25 A



The connections are identical on all of them: a covered terminal block, with mains and ground on one side and two pairs of terminals for each output polarity on the other.
The output voltage is fixed, with no adjustment.



The insides are also very similar, at least in terms of layout. One thing worth noting right away: in power supplies like these, the components are usually arranged so that the capacitors, being the most temperature-sensitive components, are in the coolest spots. The layout is also quite spacious, which helps keep temperatures down.
The circuit boards and components in all the power supplies are coated with protective varnish. Some are coated more heavily than others, but either way, it’s a plus for the manufacturer.



The following equipment was used for testing:
1. An electronic load based on the Atorch CL24. I’ve already sold one, and the other decided to put in some work rather than sit around waiting for its new owner :)
2. Oscilloscope.
3. Multimeter, my usual UT185A
4. Thermal camera



I’ll start with the 100 and 150 W models. Unlike the more powerful ones, they use a flyback topology, which gives them a wider input voltage range. However, while preparing this review, I noticed that the 100 W version is specified for 190–240 V. I don’t know why; I can’t see any reason this power supply shouldn’t work from 90 V rather than 190 V. That said, it’s not an issue for our mains supply, just a strange specification.

MN-100W12V power supply, 12 V, 100 W, 8.3 A.



It has an input mains filter and a thermistor to limit inrush current. The capacitors connected to the case are Y-class, which is very good.


At the output, there are two diodes connected in parallel, four 1000 µF, 16 V capacitors, and an inductor to reduce high-frequency ripple. One thing to note right away: in this circuit, it's better to use 25 V capacitors.


I really like the transformer, especially the fact that a busbar is used for the secondary winding.


At the input there's a capacitor from a manufacturer I don't recognize, rated at 150 µF. Its actual capacitance is just under 110 µF. I'll come back to this later, but I can already say that this capacitance is more than sufficient for this power supply at our mains voltage. With a slight reduction in power, it can also operate at the lower end of the range, 115 V.


I also like the cooling design. Heat from the components is transferred to the aluminum case using fairly large thermal pads. I would also have added a thermal path from the transformer—that would have been even better.



I ran the first tests with a ZKEtech EBC-A10H electronic load. As it turned out, though, this power supply handles even 10 A without any problems, both cold and hot. I'd also note the good output voltage stability.


I also checked the efficiency right away. Unfortunately, after I accidentally fried my old wattmeter, I can only estimate the input power: the new one is much slower and less accurate, even though it looks exactly like the old one. What a shame :(
It also turned out that the power supply works just fine at a load current of 11 A; beyond that, the overload protection simply trips.

Efficiency was measured over the 1–11 A range and turned out to be nothing special, peaking at just 84.5%.


Ripple was measured with the probe connected directly to the terminals. I even forgot to enable the oscilloscope's 20 MHz filter, so the spikes are more visible.
Even so, the ripple level is very low—and in reality it's even lower. I also made the test more demanding by checking ripple up to 10 A rather than 8.3 A.
The screenshots show the results with no load and at 3.3, 6.6, and 10 A. Even under overload, the main ripple is no more than 50 mV, or 150 mV including the spikes.



Low-frequency ripple: no complaints here either.


The warm-up test was run at 6, 7, 8.3, and 9 A. Total test time was 1 hour 16 minutes.


The initial voltage was 12.189 V. It rises slightly under load, then rises further as the unit warms up. The last two photos show the power supply hot, first under a 10 A load and then with no load. The total variation between the cold and hot power supply was just 35 mV—an excellent result.


Temperature rise at a load current of 7 A. The only thing worth noting is that the output diodes don't heat evenly.


Temperature rise at 9 A, almost 110% of the rated maximum.
Nothing particularly concerning here either, although you can again see that one of the diodes runs hotter. I suspect this is because it's close to the snubber resistors.

I'm often asked what component temperatures are acceptable. In fact, it depends on the operating conditions. If the power supply will be used outdoors, where temperatures can fluctuate, and you're testing it at room temperature, that's one thing. If it will be used in the same room where you're testing it, that's another.
The components themselves can tolerate fairly high temperatures. Semiconductors work quite normally up to 115–125 degrees, but the bigger concern is that they can heat up nearby components, such as capacitors. Transformers and inductors are best kept below 100–105 degrees; personally, I'd keep them below 100 degrees altogether.
The rest is simple: check the measured temperature, check the ambient air temperature, then add to the component temperature the difference between the air temperature during the test and the highest temperature at which the power supply may operate. That gives you the full temperature.

For example, say a transistor reaches 100 degrees at room temperature, and we set 120 as the maximum acceptable temperature. Add the 20-degree difference (120–100) to the room temperature of 25 degrees, and we get 45 degrees as the ambient temperature at which overheating would begin. Then decide what to do based on the situation.
This is, of course, a major simplification, but I think it's clear enough. A lot also depends on the type of load.




2. CTG-150W-12V power supply, 12 V, 150 W, 12.5 A - Product link.

This power supply is very similar to the previous one: it has the same topology, roughly the same component layout, and generally looks like a slightly larger version of the previous model.



The primary side also has a filter and a thermistor, but the capacitor here is a decent branded one, which is good.


The output diodes are a bit beefier, and there are more capacitors: five versus four in the 100 W version. But the capacitors are still rated for 16 V.


Put the two power supplies side by side and you can see that, for the most part, this is simply a scaled-up version: use fancier, larger components and you get more power. The transformer is wound with something more like Litz wire rather than a busbar, which is also good. Approved.


The case design is identical. The insulating pads are a little smaller, but I still like it, as well as the fact that a screw near the output diodes presses the board against the case right between the diodes.


Both this and the previous model are controlled by a small PWM controller, which, judging by the pinout, is most likely the common OB2263MP. But in the 150 W model, the PWM controller drives the transistor through an emitter follower rather than directly, which should improve the waveform at its gate.
The output uses a typical TL431 circuit, and you can also clearly see that the board is coated with protective varnish.


The input capacitor is specified as 400 V, 120 µF, made by Aishi. Its actual capacitance is 107 µF, which is perfectly fine since it falls within the allowed tolerance.
The output capacitors have a combined capacitance of 5700 µF.

Earlier I said I’d talk about the capacitors. The previous model was marked 150 µF, while this one says 120 µF, but in reality their capacitances are practically identical. How does that happen?
It’s simple: capacitors have a physical size—or, more precisely, a volume—and that’s what determines their capacitance. If you take two capacitors with the same rated voltage, they’ll most likely have roughly the same capacitance (except for obviously half-empty counterfeits). Yes, you can increase capacitance a little by using more expensive dielectrics, or reduce it—for example, in LowESR series—but the difference isn’t huge. Capacitors of the same size will have roughly the same capacitance.
But the laws of physics don’t apply to the Chinese, so they can easily print 150, 220, or even 100500 µF instead of 120. That doesn’t actually change anything. I usually look at the actual capacitance and whether it’s sufficient for the particular power supply.

This power supply is rated for 150 W, and the capacitor measures 107 µF. That’s fine for 220 V mains, but not enough for 115 V. No, the power supply will still work, but low-frequency ripple will increase and efficiency will drop, causing the high-voltage transistor to run a little hotter. It can also affect the output power, so I recommend reducing that as well when using a lower input voltage.

Interestingly, the previous power supply had one and a half times more capacitance per watt of output power, but its specified input range was 190–240 V, while this one is rated for 90–240 V. It would have made more sense the other way around :)



The power supply works fine at its rated 12.5 A, and the voltage starts to drop at around 16.2 A.


Efficiency was measured over the 1–16 A range and is about 86–87%, slightly higher than the previous model.


Ripple was measured with no load and at 5, 10, and 15 A. I’ll note that it’s low—even at 15 A, which is about 120% of the maximum, it was no more than 75–100 mV.


I measured low-frequency ripple (by accident, as it turned out) at 4, 8, 12, and 15 A. 100 Hz ripple only started to appear at the maximum current; the other operating points were fine. But again, 15 A is 120% of the rated maximum, meaning the power supply was delivering 180 W instead of the specified 150 W.


With no load, the output is almost exactly 12 V.


At 7 A, the voltage is practically unchanged. At 13 A, with the power supply hot, it dropped by about 0.1 V.
After running at 13 A long enough to warm up, it dropped further to 11.76 V, and after removing the load it was 11.8 V.
So, voltage regulation is good as the load changes, but the output voltage clearly depends on temperature. It’s not critical; I’d just like to see it do a little better.


Temperature readings at 7 A, a little over half the maximum. I'd say everything is cool, except perhaps the thermistor, which is hot—but that's perfectly normal for it.


At 13 A (the stated maximum is 12.5 A), temperatures did rise, of course, but even then the input transistor is at about 78 degrees, the transformer at 81, and the output diodes reach 100, because the focus has shifted to the snubber resistors.
Overall, this one definitely passes, and there's even a decent margin for operation with reduced input voltage, although that isn't specified at all.


3. CTG-200W-12V, 12 V, 200 W, 16.7 A power supply - product link

This one is more interesting, in my opinion. The construction is similar, although the circuit design is actually quite different from the previous models.



This power supply uses an asymmetrical half-bridge topology. In my opinion, that's the right choice. Its main disadvantage compared with a flyback converter is the narrower input voltage range, which is why power supplies of this type usually have a 115/230 switch on the input. There's no switch here, and there's only one capacitor; switchable models always have two 200–250 V capacitors.

Mean Well's 200 W and 350 W LRS-series power supplies use exactly the same topology, and I'll come back to them later.


There are also two output diodes, but while they simply operated in parallel in the previous models, here they take turns. That said, this isn't relevant to the review. There's no input filter here, which I find a bit disappointing—they could have added one.


1. There are two transistors at the input, with a pair of diodes next to them. These are characteristic external signs of an asymmetrical half-bridge.
2. The transformer looks like the one in the 150 W model. You might ask how that works: they look alike, but one is rated at 150 W and this one at 200 W. The thing is, a flyback power supply's transformer has an air gap, which reduces its power somewhat. No gap is needed here, so the core is used more efficiently. Like the previous model, this one is wound with litz wire, which is the right way to do it.
3. Another significant difference: there are two optocouplers here, one for the main circuit and one for the output overvoltage protection circuit. In this respect, it's also similar to the Mean Well LRS series.
4. The output has four 1000 µF, 16 V capacitors, and this arrangement makes sense here because they operate under less demanding conditions than in a flyback power supply. The voltage requirements are also lower. Speaking of Mean Well again, the 200 W and 350 W LRS models also use 16 V capacitors, though the brand is much better there :) Another sign of an asymmetrical half-bridge, albeit an indirect one, is the large output inductor.

The board is also coated with varnish, and overall, this is one of those power supplies that's a pleasure to hold.



The capacitor is the same as in the previous model, with an actual capacitance of 108 µF, which is a bit low for this power rating.
The output capacitance is 4500 µF, which is fine.



The construction is also similar to the previous models. The board insulation from the case is done the same way too, and I like it.



The power supply can deliver up to 19.2 A; at 19.3 A, the voltage starts to drop. When hot, this threshold will be slightly lower because the shunt resistance increases a little as it warms up, causing the protection to trip sooner.



This power supply's efficiency is about 86.5%. Unusually, it's a perfectly straight line over a wide range. I even double-checked that I hadn't made a mistake in my Excel formulas. I tested it from 1 to 20 A, though at 20 A the voltage sagged because the load exceeded the maximum rating.



Ripple with no load and at 5, 11, and 17 A. I have no complaints at all: even at 17 A, the ripple is only about 100 mV over the main waveform and 160 mV including the spikes.


The same measurement at low frequency. I wouldn't call it critical; the input capacitor is a bit small, after all, but it's perfectly acceptable. What interested me more was the lower part of the waveform, where the ripple was almost independent of the load power.


With no load, the output is about 12.15 V.


At half load, it drops to 12.14 V. At full load, even with the power supply warmed up, it has barely changed.
After warming it up under a full 17 A load, we get almost the same result as before. Remove the load, and the output is 12.162 V. Simply perfect: a 10 mV difference between the hot and cold power supplies, and about 20 mV between 0 and 100% load.




These are the temperatures after warming up at half load.
The input transistors and transformer are almost cold, and the output diodes are pretty much the same. The only thing that concerns me is how hot the PWM controller gets.


After warming up at full load, we got the following temperatures.
The input transistors are definitely below 90, since the focus shifted to the PWM controller. Perfect.
The transformer is at 80 degrees. Excellent.
One of the output diodes is at 118, which is rather high. So at room temperature it's fine, but there's essentially no headroom.


4. CTG-300W-12V, 12 V 300 W 25 A power supply - product link

At first glance, I thought this was the same power supply as the previous one, just with heatsinks added. But it turned out that's not quite the case, although they do have a lot in common.


The input section looks identical, except for the addition of a heatsink. Even the input capacitor is the same, though there are a few minor changes.


The output section also looks very similar to the 200 W model.


1. KA3845 PWM controller. The upper transistor is driven through an isolation transformer, while the lower one is driven directly.
2. The output has a pair of 3300 µF 16 V capacitors, so the total capacitance has been increased in proportion to the power rating.



The input capacitor is the same as in the previous power supplies, about 105 µF. In my opinion, that's far too little for this power rating. The output has the specified 6600 µF, which is fine, although lots of smaller capacitors work better than fewer, larger ones.


Put the two units side by side and you can see that the 300 W model is slightly longer, but otherwise it's almost identical to its lower-powered sibling.


The transformers are almost completely identical, right down to the markings; they differ by just one character.


Both versions even have some odd little wire sticking out at the side, which is further proof of how similar they are.


We take out the board and find the same design, with heat transferred to the case. I have to say, it works pretty well: the case gets noticeably hot, which means the heat is being transferred.
The additional heatsinks, though, probably aren't as effective, since heat reaches them through plastic and across a relatively small area of the component cases. So yes, they improve things overall, but not by much. The heat transfer through the bottom works better.


There are some changes under the heatsinks.
The changes in the primary circuit are purely cosmetic. For example, the shunt value has changed: it was 0.75 + 0.47, and is now three 0.75 Ω resistors. That means it went from 0.29 Ω to 0.25 Ω.


On the secondary side, though, two large diodes have been replaced with four smaller ones. In my opinion, that's the right call, since it should dissipate heat a little better. And, like the 200 W version, it has overvoltage protection on the output.



With no load, the power supply outputs about 12.18 V, roughly the same as the previous one.
But the current is a problem: I got a maximum of about 21.5 A, despite the rated 25 A. Though that's not surprising: the 200 W model delivered 19.3 A with a 0.28 Ω shunt, and this one delivers 21.5 A with a 0.25 Ω shunt. In other words, the manufacturer has deliberately limited the output power to about 250 W.

As the person who sent me these power supplies said, the Chinese sellers also told him that this unit's power output is no more than 80% of the rated value. But it's one thing for a power supply to deliver 300 W when rated for 300 W, but not be able to sustain it for long. It's quite another to see a 300 W power supply set to 250 W. They should have rated it at 250 W, then. Why mislead people? We saw that the previous power supply worked fine at 200 W; with improved cooling, this one should definitely be able to sustain 250 W.



As for efficiency, the results are about the same as for the 200 W model. The only difference is that the upper current limit here was 22 A instead of 20 A. At 22 A, the power supply's voltage dropped, since its maximum output is just over 21 A.


The ripple is higher here, though. I tested it with no load and at currents of 7, 15, and 22 A. At maximum power, the ripple was up to 250–300 mV. I think this is because of the different output capacitors; as I explained, lots of small capacitors usually work better than a few large ones.


At low frequencies, the picture is somewhat different; you can also see the 100 Hz ripple getting much worse at maximum power.


With no load, the output is about 12.2 V.


With an 11 A load, it was 12.202 V; after warming up under full load, it had barely changed.
But after warming up under a 21 A load, the power supply started lowering its output voltage, which was somewhat predictable. Once I removed the load while the power supply was still hot, the output was 12.216 V.
I have no complaints about voltage regulation, either as the load changes (if we ignore the protection kicking in) or when the load changes suddenly. It's simply perfect. The 200 W model had no problems with this either, though.


So, how about the temperatures?
The transistors and diodes are under heatsinks, so I had to stick small pieces of electrical tape on them. Measuring the heatsink temperature with a thermal camera without changing the settings would give inaccurate results.
At 11 A, there are no issues at all; everything is great.



At 21 A, things are broadly just as good. The transformer reached 88°C, still well below its maximum. The transistors and diodes are under heatsinks, so I can only estimate what's going on there, but based on tests of the previous power supply, I can say that everything is most likely fine.
The hottest component was the output inductor, at about 105°C. Its maximum temperature is a little higher than the transformer's, but that's still too hot for my liking. It clearly needs to be larger and wound with thicker wire.

So, we can say this is a power supply capable of delivering a genuine 230–250 W.




But let's get back to the capacitors.
Earlier I mentioned Mean Well power supplies. They have two popular models rated at 200 and 350 W, and their specs are as follows:
200 W, two 330 µF capacitors in series, giving 165 µF. -
350 W, two 560 µF capacitors in series, giving 280 µF

Here, the 200 W and (nominally) 250 W power supplies have just 105 µF of actual capacitance, which is 1.5 to 2 times less than required.
That's why I decided to see what would happen if I increased the capacitance. I used another 100 µF capacitor, soldered in parallel with the original one.

Along the way, I learned a couple of new things:
1. The voltage on the capacitor stays there for a very long time; even after an hour, it's still around 200 V.
2. My multimeter has protection in capacitance measurement mode, so when I tried to connect a capacitor charged to 200 V, it simply switched to Discharge mode and only then displayed a reading.



I ran a load test, but essentially nothing changed. The power supply delivered the same power as before and, just as before, started lowering its output voltage after warming up at 21 A. But this wasn't caused by the capacitor's low capacitance; it was due to the current measurement settings.
The temperatures were essentially the same too. It would be good to check the transistor temperatures, but it's awkward to do that under the heatsink, and repeating the tests without the heatsink would take a long time.


What did change noticeably was the 100 Hz ripple level. Those are the spikes up and down in each division. This was also predictable.

A fair question: if everything works fine as it is, why am I making such a fuss about this poor input capacitor? I'll try to explain.
The power supply would still work even if the capacitor's capacitance were reduced by another factor of two or three, but the input capacitance affects the following:
1. Input voltage headroom: the lower the capacitance, the sooner the power supply will start to "give up" as the input voltage drops.
2. 100 Hz output ripple. This can sometimes be critical, and it also depends on point 1.
3. Capacitor life. The lower its capacitance, the greater the ripple current; the greater the ripple current, the more it heats up and ages.
4. To some extent, reducing the capacitance may cause the input transistor to run hotter, but the difference here is very small, and flyback power supplies are more often affected by this.


For me, points 1 and 3 are the more critical ones, and I really don't understand why the manufacturer didn't just make the case a couple of centimeters longer and install two capacitors in parallel. While they were at it, I would also slightly reduce the shunt value so the power supply could deliver at least 300 W for short periods. And the output inductor—I wrote that it's really only good for a maximum of 240–250 W. Basically, a few small changes could have made this power supply noticeably better. but that's just my personal opinion.





Conclusions.
The hardest part of any article or review is summing up, very briefly, everything discussed above.

In short, my main complaint is about the 300 W power supply, and even then, mostly because of the low input capacitor capacitance and the odd current-limit setting.
Otherwise, the power supplies made a good impression.
The 100 W model has a counterfeit input capacitor, but with the correct capacitance. The power supply has relatively low ripple, delivers its rated output power—and even a little more—and is well designed.
The 150 W model is slightly better than the previous one in terms of circuit design. It handles its rated power without any issues, and ripple is not a concern either.
I liked the 200 W model a little more than the previous ones. It has a sound circuit design and output overvoltage protection. The only issue is that the input capacitor seems a bit small for 200 W.
The 300 W model is the most ambiguous. It seems to have been improved compared with the 200 W version, but the input capacitor is far too small, and the two output capacitors can't reduce ripple enough, probably because of their higher ESR. In other words, the idea is good, but the execution falls short. Used as a 200 W power supply, it would be great.

What I liked about all the models: fairly high build quality, a conformal coating on the circuit board, and quality input capacitors (in the 150, 200, and 300 W models). The capacitor in the 100 W model is a bit of a mystery, but its capacitance is sufficient. I'd also point out the fairly well-designed heat dissipation from the semiconductors. It would be even better if heat were also drawn away from the transformer and inductor.

To keep it really short, I liked the power supplies, although the manufacturer still has room to improve them. And, in my opinion, the prices are quite reasonable.

That's all for today. As always, I'll be glad to hear your comments and questions. I'd also like to thank Vitaly for sending the samples for testing.
Go to store
This page was found when searching for:
1, 3

Questions and answers

Which slim 12 V power supplies are reviewed?

The models are rated at 100, 150, 200, and 300 W. Each is 53 mm wide and 23 mm thick; the case length varies.

What efficiency did the MN-100W12V achieve?

Across a load range of 1–11 A, its peak measured efficiency was 84.5%. It handled 11 A, after which overload protection activated.

What did the review find about the 300 W power supply?

It was the most mixed result: its input capacitor is undersized, and the output capacitors do not reduce ripple enough. The author expects it to perform well when used as a 200 W supply.


This model is included in the summary table: Comparison of power supplies: ripple, efficiency, and temperature rise

The figures from this review are in the measurement database: Power supplies

Calculate manuallyAll calculators

Discounted products


You may be interested in


Discounted products


Support the project

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!

Discuss on the forumHave a question or some experience to share? Discuss the article with the author and readers: the forum is a convenient place to share schematics and photos.
Go to the forum

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