Mean Well SDR-120-24, the first of six

Not long ago, I published a review of a high-quality power supply from the well-known company Mean Well. Today it’s the turn of another "test subject." In this case, I have an industrial model designed for mounting on a DIN rail on my workbench.
Let’s take a look, take it apart, and test it—business as usual.


To explain why this is “the first of six,” and what I’m talking about: one of my regular readers sent me this power supply specifically for testing. It was bought on Taobao, but since Mean Well products are available from many sellers, it doesn’t really matter where you buy one, as long as it isn’t an obvious counterfeit—the results will be the same.
So, I wasn’t sent just one power supply, but, as you’ve probably guessed, six different ones. Today is “episode one.”
I ran a poll on my channel asking which power supply to test first. When testing began, the SDR-120-24 had the most votes, which surprised me a little. But since the poll is still open, the choice is now down to five models.


So, we’re looking at a power supply from an industrial series. That’s not just because of its form factor, but also its output voltage: this model outputs 24 V, so it’s widely used in industrial automation systems, although power supplies with other output voltages, such as 48 V, are used there too.

This model is available with output voltages of 12, 24, and 48 V, with continuous output currents of 10, 5, and 2.5 A, respectively.


The packaging is standard for Mean Well power supplies: a cardboard box with a double seal and a window that lets you see the model name without opening the package.


There are no extras in the box—no manuals, warranty cards, wires, or anything like that. Just the power supply.


The power supply has an aluminum case that also serves as a heatsink. This model is an extension of the DR series, but with a smaller case.


Rather than list all the dimensions separately, I’ll include the drawing from the product description.


Cooling is passive, which is also important in harsh operating conditions. The top and bottom of the case are perforated for ventilation.


This model is one of the lowest-power units in this case style. There are models rated at nearly 1000 W, and they also use passive cooling.


The output terminals and the power status relay terminals are at the top of the case. The relay can signal other devices that the power supply is operating normally.
Below them are a trimmer and an LED indicating that the power is OK.

The mains and ground terminals are at the bottom. Interestingly, the specified tightening torque isn’t just given—it differs for the various terminals. It also says to use copper wire only.


The terminals use a “lift” mechanism, where the entire plate rises, rather than a clamping mechanism. This also improves reliability.


The mounting mechanism is metal, with a spring-loaded plastic latch. In other words, first pull back the latch, then mount the power supply on the rail, and release the latch. You can also mount it without pulling the latch back first—it snaps into place just fine.


As mentioned above, this model is an extension of the popular DR series, so it makes sense to compare the appearance of the two series.
The photo below shows two power supplies with the same power rating and output voltage, but from different series—the old and the new.


The new power supply is noticeably narrower, but it requires more installation depth, which is worth keeping in mind when replacing one.


Time to take a look inside. To do this, first remove the DIN-rail mounting bracket (three countersunk screws), then undo the two regular screws underneath and the two short countersunk screws on either side of the top of the case.


Remove the perforated cover.



All screw joints are secured with paint.


Undo the remaining screws, spread the cover slightly, and remove the circuit board. I wasn't too keen on the fact that thermal paste doesn't cover the entire surface of the heat spreader plate.


The block diagram of the device looks like this:
1. The input power goes to the active power factor correction stage.
2. The power supply inverter, using a flyback topology, is connected after the PFC stage.
3. Next comes the output rectifier, filter, and output voltage and fault monitoring circuit.
4. In addition to the feedback circuit, there's an output overvoltage protection circuit that disables the inverter.


The components listed above in the power supply under review.


The input filter is quite complex. In addition to the usual two-winding choke, it has separate chokes on all three lines (live, neutral, and ground).


The protection components include a fuse and a varistor rated for 470 V peak.
There are also X2- and Y1-rated capacitors.


Further along the circuit there's another two-winding choke, followed by a diode bridge with a heatsink, and then another choke.
After all that comes the active PFC stage; its inductor is visible in the photo.


For ordinary consumers, power factor correction is, at best, simply better protection against input voltage fluctuations. For industrial companies, it means lower reactive power—a bonus, since they have to pay for it separately, monitor its level, and pay a penalty if it exceeds the limit.


The power components are mounted on a double aluminum plate, which transfers heat to the case. An insulating plate is also fitted between the power components and the rest of the electronics.


The PFC stage is followed by a 100 µF capacitor made by Nichicon. A 0.07 Ω resistor in the PFC circuit serves as a current-sense shunt for overload protection.


The power supply transformer is not only varnish-impregnated, but also wound with a multi-strand bundle to improve its performance. I'd venture to say it's litz wire, but it's hard to tell without taking it apart.


The transformer and output rectifier section. In this case, the manufacturer opted for a synchronous rectifier to improve efficiency, so the heatsink carries a transistor rather than a diode assembly: IRFB4321. Its typical on-resistance is 12 mΩ, which at a pulsed current of 15 A gives a voltage drop of just 0.18 V, compared with about 0.5 V for Schottky diodes.


The remaining capacitors are made by Rubicon. The capacitance after the rectifier is 3 × 1000 = 3000 µF, rated at 35 V.


Optocouplers for the feedback loop and output overvoltage protection.


There's a single noise-suppression choke on the output, wound with fairly thick wire. There are also two more capacitors here, rated at 330 and 470 µF. Together with the main capacitors, that brings the total capacitance to almost 4000 µF.
The relay is here too; it signals at the output that the voltage is OK.


Unfortunately, there won't be a schematic of the power supply this time. I could redraw it, of course, but it would be a lot of work and I'd have to partially desolder the circuit board.

I also didn't notice any safety issues. The circuit board has protective slots in all the hazardous areas.


A few words about the components used.
1. The power supply is controlled by a dedicated PWM controller TEA1750T, which combines control of the flyback converter with control of the active PFC.
2. The PFC protection diode is an LT7B33.
3. The synchronous rectifier is controlled by a TEA1761 controller.
4. A 5 mΩ resistor is installed on the board as a current-sense shunt. This does reduce efficiency somewhat, of course, but the resistor is hard to do without: unlike a diode, which does everything “in hardware,” this circuit needs to monitor the current.
5. The output voltage is monitored using a quad operational amplifier, the LM224.
6. There are several zener diodes near the op-amp, which also take part in the overvoltage protection circuit.


To learn more about the power supply—more specifically, its power components—I had to remove the heat spreader. The components are held against it by two plates: a wide one on the left and a narrow one on the right.


1. The PFC transistor is a TK20J50D, rated at 20 A, 500 V, and 0.22 Ω. The PFC diode, a BYV29X, is located to its right.
2. A bimetallic thermostat, seki st-22, rated for 90°C, is installed between the PFC components and the inverter transistors.
3. The converter's high-voltage transistor is a TK13A65U, rated at 13 A, 650 V, and 0.32 Ω.


That’s all for the inspection; we can move on to the tests.
First, I checked the output voltage, adjustment range, and how easy it is to set.
It was initially set to a little over 24 V, although that doesn’t really matter: even if you set it to 25 V, nothing will happen to the equipment.
The adjustment range downward is very small—it doesn’t even reach 23 V. Upward, though, you can set it to almost 29 V.
The adjustment range isn’t very wide, and setting the voltage precisely is quite difficult—though maybe I’m just being picky now :)


Next, I assessed the ripple level, load capacity, and efficiency. Since the power supply has active PFC, I also measured the power factor.
I won’t show the whole process; I’ll just put all the results in a table at the end. I’ll only mention that this test used the familiar test setup many of my readers will know: a multimeter, an electronic load, and a wattmeter.
The wattmeter is a household model, so the efficiency and power factor figures are approximate.


Ripple levels at load currents of 1.2, 2.5, 3.8, 5, 6.5, and 7.5 A.
The specifications state a maximum ripple amplitude of 100 mV. The oscilloscope is set to 50 mV per division. The power supply exceeded the specified limit at 7.5 A; at 6.5 A, the ripple amplitude was 100 mV. Both of these are within the normal operating range, but exceed the rated continuous power, so the power supply passed the test.
I should also note that the power supply produces virtually no ripple or audible noise at no load and under light loads.


Above, I mentioned the limit operating conditions. Here I’ll add a brief clarification of what I mean.
The power supply is rated at 120 W, but can briefly deliver up to 180 W (150% of its maximum rating) for no more than 3 seconds. During testing, I found that this limit is not just on paper. The power supply shuts off automatically after about 8–10 seconds if the load current exceeds 6 A. The protection latches, so you have to briefly disconnect the input power to restore normal operation.

The graph shows how long the 180 W load is applied and the required pause between loads. In my experiments, I actually made the pauses much shorter and didn’t encounter any problems, which suggests that the power supply is reliable.


In addition to checking the high-frequency ripple level, I also tested it at 100 Hz. The oscilloscope traces were captured at load currents of 2.5, 5, 6.5, and 7.5 A. No problems were found here either.


Since the power supply is rated for an input voltage range starting at 100 V, I tested it at that input voltage as well.
In reality, the voltage was around 107–108 V, since the mains voltage was 230 V and I couldn't get it any lower. I don't think this is critical, though.


For this test, I also used a TN-61 transformer as an autotransformer.


I didn't notice any difference in the high-frequency ripple. In fact, I'd say the ripple level even dropped at peak power.


Since the power supply has active PFC, the ripple at 100 Hz remained unchanged as well.


In the comments on the video, someone asked me to check the current waveform in the primary circuit, so I did that too.
I used a 0.47 Ω, 1% wirewound resistor as a shunt.
No load.


Load at 25%, 50%, 100%, and 150%. Note that the scales differ between the oscilloscope traces: the first is 50 mV/div, the second 100 mV/div, and the last two 200 mV/div.



After that, I decided to run a short thermal test, although with the unit open.
Below are six thermal images. The top three were taken after almost an hour at 5 A and an input voltage of 230 V; the bottom three were taken at the same current but with an input voltage of 108 V. You can clearly see that the PFC section gets hotter, especially the input bridge rectifier. In all cases, though, the hottest component is the output transformer.


Next, I wanted to make the test conditions closer to those where this power supply might actually be used. I took a metal enclosure from some kind of alarm system, removed almost all of its internals, and mounted a small piece of DIN rail on the side wall to secure the power supply.


I set the assembly upright, closed the cover, and ran it at 5 A for 1 hour.


Because of the power supply's design, I couldn't properly measure the temperature of individual components, so I had to use a thermal camera. Fortunately, the openings in the enclosure are large, and the most critical components are right next to them.
On the left is the output transformer at 72°C; on the right is the PFC inductor at 65°C.


Then it occurred to me that this was all a bit too easy, since the power supply is rated to deliver its specified power at ambient temperatures up to 60°C.



In the end, I made the test a little tougher by increasing the current to 5.5 A, or 110% of the rated load. I also measured the air temperature inside the enclosure at the end of the test. Unfortunately, the highest temperature I got was 45°C.


In any case, the results are useful and can certainly be put to use. The hottest component is the output transformer, at 81.6°C. If we add the missing 15 degrees, the transformer would reach 97°C. That's high, but still tolerable. Also, let's not forget that the test was run at 110% of the rated output power.


Here's the final table of test results.
The output voltage regulation is good. The power supply doesn't show the usual behavior where the voltage drops under overload: the voltage stays fixed, but the unit shuts down 8–10 seconds after the load current exceeds 6 A.
The specified efficiency was 91%, while I measured a maximum of 89%. The efficiency measurement at a 110 V input is approximate, since I powered the unit through a transformer and only subtracted the transformer's own 4 W power consumption.


Video version of the review.


To sum up, I liked the power supply. It uses quality components and is well built, with a full set of protections against overheating and overload. I should also mention the active PFC and synchronous rectifier. Output ripple is within the specified limits, and output voltage stability is excellent. It also has a relay output indicating the power supply's status. The efficiency was a little lower than specified, but I suspect that was due to the limited accuracy of my wattmeter.

The only thing I wasn't too happy with was the fairly high transformer temperature. In my opinion, the manufacturer could have used one with a slightly higher power rating.

In any case, the power supply passed all the tests and is fully functional. During testing, I heated it up, loaded it with up to 8 A, and shorted the output while it was running, but it keeps working just fine. I think it’ll be hard to kill :)

That’s all from me. I hope you found the review useful.