A small 12 V, 3 A power supply

I spotted this power supply on the ChinaBay website and wanted to take a closer look. Since they ended up providing it to me for free, it would have been sacrilege not to experiment with it. Anyway, more after the jump.

So, I ordered this power supply. It crawled around somewhere in the Chinese postal system for a while, but eventually made its way to my desk.

It arrived in a standard yellow envelope. The usual blah-blah-blah; photos under the spoiler.

The specifications are printed on the end.


The specifications confused me a little. Usually, either the full input voltage range is listed, or, if there's a 110/220 V option, there's a switch and the input rectifier circuit can be switched to voltage-doubler mode. There was no switch here. We'll take a closer look inside later.

It's fairly compact.




At one end are the 220 V input terminals, the ground terminal, and the 12 V output terminals. There's also an LED indicating that output voltage is present, and a trimmer for adjusting the output voltage.


After opening it up, I found the power supply's circuit board.
The board has a full input filter, a 33 µF, 400 V capacitor (quite reasonable for the rated power), a high-voltage section built around a self-oscillating circuit (when I ordered it, I was hoping for a standard UC3842), and an output filter consisting of two 470 µF, 25 V capacitors and an inductor. The output filter capacitance is a bit low; I'd have doubled it.


The power transistor is a 5N60D — www.icemostech.com/ice/superjunction/ICE5N60D20140221Rev2.pdf in a TO-220 package.
The output diode is an STPS20H100CT — datasheet.octopart.com/STPS20H100CT-STMicroelectronics-datasheet-10413158.pdf, also in a TO-220 package.
The regulation and feedback circuit is built around a TL431.

The back of the board.


Nothing unusual: average-quality soldering, the flux has been washed off, and everything looks fairly neat.
But the marking on the board caught my eye (there's one on the top side too).
SM-24W. Maybe it was originally a 24-watt power supply, then they decided that wasn't enough and changed it to 36?
The experiments will tell.

First power-up: nothing blew up, which is a good start.


I loaded the power supply with a pair of classic, indestructible Soviet 10 Ω resistors in parallel.
Current was about 2.5 A.


I measured the voltage after the wires leading to the resistors, so it had dropped a little.
I left it like that, went to have some tea and a smoke, and waited for it to blow up.
It didn't. It barely even got warm—around 40°C, maybe 45°C. I didn't measure it; it just felt slightly warm.

I added another 0.22 A load (I couldn't find anything suitable nearby), but nothing changed.


I decided not to stop there and connected another 10 Ω resistor to the output.
The voltage dropped to 10.05 V, but the power supply kept working stubbornly.


At that point I started feeling sorry for the designers of this power supply. They'd managed to simplify it so much and still make it work, so I decided to call it quits on the standard tests.
I have to say, I was skeptical about this power supply, mainly because of its circuit design. I'm used to working with more expensive power supplies that have a PWM controller, current monitoring, and so on. But the tests showed that this approach can work just fine too.

Next, I decided to move on to the less conventional part of the testing and see if I could get it to do what I wanted it for. Regular readers of my reviews know that I like not only to show a product, but also to put it to use. I won't disappoint you this time either.

Once the board was assembled, I installed it inside the power supply.
The connection points are marked on the board, and you can see where the negative trace has been cut (above the number 3).


I wrapped the board in tape and tucked it into a reasonably clear spot.


After that (though it's actually better to do this before wrapping it in tape), I set the power supply's output voltage to 13.8 V. This is the voltage that will be maintained across the battery; it's usually set to 13.8–13.85.


Here's the assembled and adjusted unit.


I connected a small load and a battery. The charging current is 0.39 A (it may drop a little as things warm up).


I disconnected the power supply from the mains, and the load kept running. The multimeter shows the load current plus the relay's current consumption plus the current consumed by the measurement circuits.


My friend needed a UPS for a current of 0.8–1 A, so I loaded it a little more heavily.


Then I connected the 220 V mains. One multimeter shows the voltage across the load (it will rise further because the battery isn't charged); the other shows the charging current, which has dropped a little as things warmed up.


Overall, I think the modification was a success. This power supply can power small loads of up to 1–1.5 A. I wouldn't push it any further, since it's operating outside its intended mode. If you use a 15 V power supply, you can increase the current, but you must always take the battery charging current into account (it's set by resistor R1). A resistance of 1.6 Ω gives a charging current of about 0.4 A; the lower the resistance, the higher the current, and vice versa.
You can modify more powerful power supplies this way, too; you only need to replace the relay and the high-power diode on the board.
If anyone disagrees with the set charging current, charge termination voltage, or automatic shutoff, these can all be changed easily. If needed, I'll explain how.

This power supply has been running for a week now without any problems (my friend soldered the battery cable terminals on himself). I hope this modification will be of interest to someone. Questions and comments about the circuit, design, and circuit board are welcome.
Yes, I know there are Mean Well power supplies with a UPS function (and I use them successfully), but they cost around 30 bucks. Here, we have an inexpensive power supply (you may even have one already) and a set of parts costing less than a dollar.


Summary.
Pros
It works.
The specifications are quite close to the advertised ones.
The build quality is perfectly acceptable.

Cons.
The terminal block is rather awkward; even a tinned 0.75 mm² wire is hard to fit in.
A varistor after the fuse wouldn't hurt, but I'm nitpicking now. At least they included an input filter.
The output capacitors are pretty poor; I'd replace them, but they'll do just fine as long as you don't load it to the max.

In my opinion, this is a perfectly decent power supply, suitable for powering LED strips, video cameras, electric locks, and the like.
This product was provided to me free of charge for testing and review by the chinabuye store.