Mean Well HDR-15-24 power supply

Once again, I’m reviewing a power supply—and once again, it’s a 24 V model. But this time, it’s not just a branded power supply: it belongs to the HDR series, which I’ve written about before. And this time, I’m reviewing the smallest model in the series.


First, a little about the HDR series. These power supplies are designed for DIN-rail mounting, and their dimensions make them suitable for standard electrical panels: their height and depth are the same as those of ordinary modular circuit breakers.
The series includes four power ratings—15, 30, 60, and 100 W—and each is available with different output voltages, so you can choose the one that suits you best.

I’ve already reviewed the HDR-60 and HDR-100, and today I’m looking at the smallest model, the 15 W HDR-15.


The packaging is like that of the other models, with a small “window” showing the model name. The side has a wiring diagram for connecting the different models to the mains.


There are no instructions, warranty papers, or anything else in the box—just the power supply itself.


If you’re interested, you can read the datasheet. I’ll just include the relevant excerpts from it throughout the review.

This series includes five versions with output voltages of 5, 12, 15, 24, and 48 V. Mine is the 24 V version.


The power supply is 17.5 mm wide, which is one module. Its height is the same as a standard circuit breaker, but it’s a little less deep, so the front panel will most likely sit flush with the panel cover when installed.


Here are the dimensions in a more familiar format.


1, 2. The input and output terminals are recessed into the case, with the input terminals at the bottom. There are plenty of ventilation slots on both the top and bottom.
3. The terminals are the “elevator” type and look to be of decent quality.
4. The DIN-rail clip isn’t spring-loaded, so it has two fixed positions.
5. The top has all the information needed for wiring, as well as a status LED and a trimmer for adjusting the output voltage. The specified adjustment range is 21.6–29 V.
6. All the additional information is printed on the side.


The power supply is very easy to open: just press in a couple of clips on the sides. Once disassembled, it consists of three parts: the case, the bottom section with the mounting clip, and the circuit board itself.


While taking it apart, I found a curious “souvenir”: a cut-off corner of some kind of packet. It could only have gotten inside during manufacturing... And judging by the looks of it, the material is something like metallized film, which could have caused a short circuit.


As you’d expect, the components are packed in very tightly. It’s not easy to fit a quality power supply that meets all the relevant standards into a case this small.


The circuit board has two layers and components on both sides. The bottom side has the PWM controller, bridge rectifier, and other small components.


1. PWM controller U28740. It features automatic frequency adjustment to reduce EMI, current-limit operation, overvoltage and undervoltage protection, and overload protection.
2. There’s a rectifier diode on the output, feedback based on a TL431, and a TVS diode connected in parallel with the output terminals for overvoltage protection.


1. The input has a mains filter with a common-mode choke and an X capacitor.
2. They didn’t forget the fuse either; it’s mounted on the side board under an insulator.
3. The input capacitor is rated at 27 µF. Judging by its appearance, it’s probably made by Rubycon. The thermistor is visible to the right.
4. There’s a varistor too, rated at 470 V.
5. The high-voltage transistor is an SVF7N65C, with a ferrite bead on one of its leads.
6. The output has a pair of capacitors: a 220 µF Nippon Chemi-Con and a small 100 µF Rubycon from the ZLH series. Both are rated at 35 V. A common-mode choke is installed between the capacitors.


In this case, one of the blog’s readers kindly shared the schematic. Many thanks to them—I think it’ll come in handy if the unit ever needs repairing.


I put everything back together and plugged it into the mains. The output is 24.22 V, and the power consumption is below the wattmeter’s sensitivity threshold.
The output voltage adjustment range is 20.8–29.9 V, compared with the specified 21.6–29 V—so it even exceeds the stated range. Adjustment isn't very precise because the trimmer is a standard type, but in this case it isn't really needed.


Testing presented a few minor difficulties because my electronic load is a bit too coarse for this power supply. When selecting the current manually, the smallest step is 10 mA, but in automatic mode it's 100 mA, which reduced the accuracy of the tests.

During the automatic test, the power supply shut down at around 700–800 mA, then restarted automatically.


When adjusting the current manually, I found that the power supply has a mode where it starts reducing the voltage instead of shutting off completely.
The maximum current was 780 mA; after that, the voltage began to drop, and the power supply went into protection at 910 mA.


The power supply's stated efficiency is 86%; in practice, it was 86–87%.


I measured the output ripple. The specifications claim up to 150 mV peak-to-peak. I measured it by connecting the probe directly, and got no more than 50 mV, even under the worst conditions.
Below are the waveforms with no load and at currents of 250, 500, and 750 mA. You can also see that there are no short spikes, the so-called “needles.”


The same conditions, but with a slower timebase. Ripple at 100 Hz is practically nonexistent; only high-frequency noise is visible.


I decided to simplify the thermal test: I loaded the power supply to 700 mA, above its stated maximum of 630 mA, and left it running for an hour. The graph ends abruptly because I accidentally shorted the output terminals while taking thermal images. Once I cleared the short, the power supply started up as if nothing had happened.


The unit was lying on its side, which impairs cooling because the ventilation slots are almost completely blocked. However, it wasn't installed in an electrical cabinet, so the two effects more or less canceled each other out.
The case temperature near the transformer was 63 degrees, the transformer itself reached 81.5, and the hottest spot on the back was 67.5 degrees.

You could say the temperature is right on the limit, but bear in mind that the unit was running at a 10% overload, so I think it's fine.


Output voltage stability and its dependence on temperature.
1. Unit warmed up, load current 700 mA
2. Unit warmed up, no load
3. After cooling down.

The result: at 110% load, the voltage drops by 40–50 mV, but is almost unaffected by temperature—the difference was just 5 mV. The datasheet specifies 1% regulation accuracy; in reality, I measured about 0.2%.


And of course, a comparison with its “big brothers.”
1. The unit under review
2. HDR-60
3. HDR-100

The only model missing from the collection is the HDR-30.


After I published the review, the well-known Elektroschaman got in touch and told me that the HDR series also includes a 150 W model. He even shared some photos, for which I'm very grateful.
The case is 6 units wide.


He also shared photos of the internals, not just the exterior. Here's one of them.


Conclusions.
What is there to add? MW has once again confirmed that it makes high-quality power supplies. The only thing I can fault it for is the debris inside the case, which genuinely surprised me.
Otherwise, the power supply performed well. I have no complaints about the temperature, build quality or components, regulation accuracy, or current headroom.


That's all for today. I hope you found the review useful.