Power Supplies, Small and Very Small

Power supplies aren't always high-power units; some are very small, but no less useful for it.
Today I have four representatives of this class of power supply on my “operating table,” but they'll go through the same tests as always.

Sometimes you need a very low-power supply—for example, to power a low-power device, a sensor, an Arduino-like device, or something similar.
Of course, you could use a large power supply, but that would make the device noticeably bigger, so compact, low-power supplies are used instead.

The tests, however, will be standard, as will the review format.

But today I'll start not with the packaging, but with how these power supplies (at least a couple of them) made their way to me.

Just a little about the packaging, so as not to distract from the rest—I'll tuck it under a spoiler.

These power supplies really are tiny. I'll give the dimensions of each board individually as I go through the review; for now, here's a group photo next to the familiar matchbox :)


Let's start with the lowest-power unit.
Product link in the store, price $3.89.
A general note first: the store doesn't provide all the information listed below; I found some of it on other sites, but it seems entirely plausible.

The listed specifications are:
Input voltage — 110 ~ 370 V DC, 85 ~ 264 V AC
Output voltage — 12 V
Output current — 83 mA
Load power — 1 W
Efficiency — 80%
Output voltage regulation accuracy — ±10%
Ripple — no more than 100 mV
Short-circuit and output overload protection with automatic recovery.
Board dimensions — 26 x 24 x 12 mm without leads, 26 x 33 x 12 mm with leads
Distance between the 220 V leads — 5 mm; 12 V leads — 2.5 mm. The distance between input and output isn't a multiple of 2.5 mm, though, and measures 14.3 mm

The board has no fuse or input or output filters; the design is as simple as it gets.
The input capacitor is 2.2 µF (actually 1.9), and the output capacitor is 220 µF (actually 183). That's enough capacitance for normal operation.
PWM controller OB2535, maximum power 5 W.


Almost all the resistors are precision types, the soldering looks fine, and I couldn't spot any issues. There's also a ceramic capacitor in parallel with the output capacitor.


The schematic of this power supply.
As I wrote above, this is the simplest of the four power supplies. It lacks most of the circuitry found in larger PSUs, in order to keep its size down.
This power supply has no familiar feedback loop with an optocoupler, which is perfectly reasonable at such low power levels. But it does actually measure the output voltage, albeit indirectly. The measurement is taken from the IC's supply winding.
The IC can operate in two modes: voltage regulation and current regulation.


The second one is a slightly more powerful power supply.
Product link in the store, priced at $2.72.
While the first one had one of the most common output voltages, this one provides a much rarer 24 V. Judging by the markings, though, a 12 V version is also available.
Specifications:
Input voltage — 110 ~ 370 V DC, 85 ~ 264 V AC
Output voltage — 24 V (12 V 400 mA and 3.3 V 500 mA versions are also available)
Output current — 200 mA
Load power — 4.8 W
Efficiency — 85%
Ripple — no more than 100 mV
Board dimensions — 41 x 15 x 17 mm


Interestingly, the transformer on this board is smaller than the one on the previous board, yet the rated power is noticeably higher.
The PWM controller has a built-in high-voltage transistor. Its part number is THX208. The datasheet rates it at 4 W for an input range of 85 ~ 264 V. Not much, considering the PSU is rated at 4.8 W.
There is no input filter or fuse; a 0805-size jumper is used in place of the fuse. There is no output filter either.
The input capacitor is 4.7 µF (actually 4.2), and the output capacitor is 220 µF (actually 242). The input capacitor is just barely adequate, while the output capacitor is appropriate for the output current.


All the resistors are precision types, at least according to their markings. That's good to see, since using ordinary resistors can cause the output voltage to drift as the board warms up.


This version does have feedback using an optocoupler, along with a proper output-voltage sensing circuit using a TL431 zener diode.


The third one managed to surprise me during the initial visual inspection, but more on that later.
Product link in the store, priced at $3.05.
This PSU has a fairly common output voltage of 5 V. I chose a 5 V PSU for this review because they can be in high demand, as this voltage is used in many applications these days.

Specifications:
Input voltage — AC 85 V — 265 V
Output voltage — 5 V
Output current — 1000 mA
Load power — 5 W
Efficiency — 85%
Output voltage regulation accuracy — ±0.1 V
Ripple — no more than 150 mV
Board dimensions — 52 x 24 x 18 mm


This power supply has no fuse (there's a 0 Ω jumper in its place), but it does have input and output filters and a resistor to limit inrush current.
The power supply uses a AP8012 PWM controller with a built-in high-voltage transistor. This PWM controller is rated for 5 W (for this IC size and input voltage range). It's also running right at its limit, but the tests will show how it performs.
This board already has an interference-suppression capacitor, and it's the proper Y1 class.
The power supply arrived with minor damage: a piece of plastic had broken off the inductor. Since it was in a bag, the postal service was most likely to blame.


But something else surprised me. I've reviewed a ton of different power supplies, but this is the first time I've seen an input varistor in one (maybe the second time; I'm not sure), especially in such a tiny power supply. You don't find one in powerful, more expensive power supplies, but they put one in here. If only they'd added a fuse too :(
The input capacitor is rated at 4.7 µF (measured 4.2), and the two output capacitors are 1000 µF, 10 V each (measured 2 × 1095). There's also an output interference-suppression inductor.


The circuit board. As with the previous power supplies, the manufacturer has used precision resistors here too. Nice :)
The soldering is generally good, and the board is clean.


There's nothing new in the circuit: classic design, with a filter, PWM controller, and TL431 on the output.


And now the fourth power supply.
Product link in the store, priced at $4.17.
This power supply is a bit of an outlier, since its power rating and dimensions are noticeably larger than those of the previous ones. But I've been asked about power supplies with these specs several times, so I decided to include it in the review as well.

First, the specifications:
Input voltage — AC 85 V — 265 V
Output voltage — 5 V
Output current — 2000 mA (2500 mA peak)
Load power — 10 W (11 W max)
Efficiency — 85%
Output voltage regulation accuracy — ±0.1 V
Board dimensions — 60 × 31 × 20 mm


This is the first board in the review with a proper fuse.
It also has input and output interference-suppression inductors and a thermistor to limit inrush current.


This board has a more powerful diode, as well as a Y1-class interference-suppression capacitor (the marking isn't visible in the photo).
The input capacitor is rated at 15 µF (measured 15.2), and the output capacitors have a combined capacitance of 2000 µF (measured 2110). The capacitance is as required.
This power supply uses a low-power PWM controller with an external MOSFET, partly because its power rating is higher than that of the previous ones.


As with the previous power supplies, precision resistors are used. For some reason, though, there are soldering marks near the output connector, even though the board is otherwise clean and tidy.


Interestingly, there’s a spot in the output circuit for an additional resistor connected in parallel with the lower feedback-divider resistor. Installing a resistor there can raise the output voltage.
I couldn’t identify the PWM controller, but it’s most likely a 63D12, whose closest equivalent is FAN6862


The circuit is very similar to one of the power supplies I reviewed earlier—almost identical, with only some component values differing.


That’s the visual inspection done; now it’s time to move on to the tests.
This time I’ll use a simple electronic load, since I don’t see any point in using a powerful one. Besides, it’s quite noisy, and these tests were going to take a while.
I’ll test the power supplies in the same order I described above, but the test procedure will be a little different from the one I used in previous reviews.
Since these power supplies are small, the procedure was as follows:
Test at no load (or, more precisely, at 20 mA), then a 15-minute test at 50% load, temperature measurements, a test at 100% load, and more temperature measurements.
Then I’ll increase the load until one of the limits is reached (overload, overheating, or power supply failure).
All the results will then be compiled in a single table.


The tests are complete. Now for a table of the results, but first, here's a list of the reasons each power supply test was stopped:
1. The power supply went into protection at 250 mA, shutting off the output.
2. The power supply's output voltage dropped below the allowed limit.
3. The test was stopped because the PWM controller got too hot.
4. The test was stopped because the output diode got too hot.


Now we can draw some conclusions.
First power supply.
The design is very simple, with no fuse or filters, but a power supply that can handle three times its rated load while maintaining such excellent output voltage regulation deserves respect. A fuse could be added, although since this power supply was clearly designed to be used as part of some device, a fuse is often already present on the main board.

Second power supply,
It meets its stated specifications, but has no power headroom. At 1.3 times its rated load, it goes into protection, even though it has plenty of thermal headroom. It's also a shame there's no fuse :(

Third power supply.
It works great under normal conditions, and has the lowest ripple of all the power supplies tested. However, I don't recommend using it above 1 A (though nobody promised more than that). The downsides are the lack of a fuse and poorer output voltage regulation.

Fourth power supply.
Output voltage regulation is pretty good. There is some ripple, but it's within acceptable limits. It has input and output filters, but the output inductor is a bit undersized for a power supply of this rating. It handles the heat fine, but its low inductance results in noticeable output ripple.

Overall, for all the power supplies.
They all passed the tests. Some did better than others, but they all meet their stated specifications.
The first power supply's performance was a surprise: rated at 1 W, it can deliver 3 W without any trouble. Was this one really made in China?
I was also surprised to find proper EMI suppression capacitors in the 5 V power supplies, and a varistor in the 5 V, 1 A unit. They don't even put those in higher-power supplies, and yet here they are…

That's about it. As always, I look forward to your questions, corrections, and additions in the comments. I hope you found the review useful.

The product was provided by the store for this review.