Low-profile Mean Well LRS-100-48 power supply

I’ve reviewed an LRS-series power supply before, but it was a higher-power model with active cooling. This time, I got my hands on its little "brother," so today I’ll introduce it to you.
Inspection, tests, conclusions—as usual.


How I came to have this power supply is a very simple story: several units were bought for PoE surveillance switches, and I bought one as a spare. I’ve been asked about this series of power supplies several times, but until now they’d only passed through my hands. This time I kept one for myself, so I decided to write a review.

The link in the title goes to the store where I bought this power supply. I chose it simply based on price, although at the moment it’s even a little cheaper at another store. These power supplies are also available elsewhere, including on AliExpress: link.

It comes in the familiar Mean Well box with a window that lets you see the model name.


A datasheet with the technical specifications is included; I’ll quote some of it throughout the review.


Inside, there’s just the power supply itself—no instructions or accessories.


The LRS-100-xx series includes 7 models with output voltages of 3.3, 5, 12, 15, 24, 36, and 48 V. The model under review is therefore the highest-voltage one. However, the series’ rated 100 W output is available only on power supplies with an output voltage of 12 V or higher.


The aluminum case also serves as a heatsink. The top is covered by a perforated panel, with holes so large that the power supply is practically open-frame.


There are two mounting holes on the bottom. The power supply can also be mounted through the side panel, which has three holes.


The dimensions and mounting-hole positions are shown in the drawing from the product description.


For comparison, here are the dimensions of the power supply under review and the RS-150-12, the most powerful model in the RS series. Despite having only one and a half times the power, the LRS model is considerably smaller—and, just as importantly, thinner.


Connections are made via a screw terminal block; there’s no protective cover. On the right are the power indicator LED and a trimmer for adjusting the output voltage. The LED is barely visible, which feels a little unusual.
The top has a label with the main specifications and a link to the manufacturer’s website.


The cover is held in place by a single screw. Once you remove it, you can slide the cover off and get to the internals.


Overall, everything is neat and even rather spacious, which helps with cooling.


Although these power supplies have been in production for quite a while, the board is marked with a design date of September 16, 2019, and the puzzling word “Victor.”


A few words about the components.
1. Input filter: there’s a thermistor, fuse, EMI suppression capacitors, and a common-mode choke. The choke is wound with a busbar rather than wire; I’ve shown one like this before in a review of another Mean Well model.
2. Input bridge rectifier. A glass-bodied suppressor (presumably) is visible on the right.
3. 150 µF, 400 V high-voltage filter capacitor made by Lelon.
4. The number of wires coming from the transformer on the high-voltage side surprised me a little. Below are two optocouplers: one handles feedback, and the other protects against output overvoltage.
5. There are three output capacitors, rated at 120 µF, 63 V. There’s room for three more next to them, but those spots are empty.
6. The output inductor has no core, which is a little unusual.


After unscrewing the power components and the screw that secures the board to the case, we can remove the board. There’s insulation between the board and the case.


Now we can see which power components are used.
1. High-voltage power transistor SVS11N65F — 650 V, 0.37 Ω
2. Output diode assembly 20LC30N - 300 V, 20 A

Both components are in fully insulated packages and are also covered with silicone caps.


The board is clearly designed to be universal across the entire series of power supplies. Three resistors for discharging the X capacitor are visible at the bottom left, but there is also a spot nearby for a dedicated chip that performs the same function with lower power consumption.


There are also unused spots elsewhere on the board.
The PWM controller is an MW03A. I couldn't find a datasheet, but it was listed for sale on AliExpress, though the listings have since closed. At first I thought it was a Mean Well chip, by analogy with Astec chips, but probably not.
There's nothing original or unusual here, really: just a standard PWM controller for a single-switch flyback power supply.


The output section consists of large copper areas coated with solder.


1. The feedback circuit is about as classic as it gets: a basic TL431 and a voltage divider. On the left, you can see that a ceramic capacitor has also been placed in parallel with the output capacitors, though there's only one.
2. Output overvoltage protection. More interestingly, the board is so universal that it can also use a synchronous rectifier. To do this, you simply solder the jumpers differently, replace the diode assembly with MOSFETs, and solder in the chip that controls them.


I didn't redraw the power supply schematic because it's utterly basic; there's simply nothing unusual about it. It's a standard flyback power supply.
The datasheet includes a block diagram, which is quite enough to understand how it works.


All right, that's enough looking it over. Let's try powering it up.
The output was initially set to just over 48 V. The power supply clearly has a green mode, since it draws only about 0.2 W from the mains with no load.
The output voltage can be adjusted from 41 to 54.6 V.


The overcurrent protection threshold is about 2.85 A, and it resets automatically after tripping.


The power supply can deliver around 2.6–2.7 A continuously.
Load regulation is excellent: across loads from 0 to 125%, the voltage changed by only about 4–5 mV.


The description claims efficiency of up to 91%; I measured 91.5%, with the 91% threshold reached at 50–60 W and above. The horizontal axis on the graph is marked in increments of 210 mA, or 10% of rated power. The test ran up to 2.73 A, or 130%.


Next came a thermal test and an assessment of how much the output voltage drifted as the unit warmed up. During the test, I eventually removed the cover, since it wasn't obstructing cooling anyway, and measuring the temperature with it on wouldn't have been very accurate.


The test had three stages: 20 minutes at 50% load, another 20 minutes at 100%, and finally 10 minutes at 125%.
1. Start of test, no-load voltage
2. Start of test, voltage at 50% load.
3. End of test, voltage at 125% load.
4. End of test, no-load voltage.

The power supply has a slight positive voltage-versus-temperature coefficient. At the same time, there's virtually no voltage dependence on load, as the last two photos clearly show.


The transformer was the hottest component; the others were significantly cooler. The transformer temperature was almost unchanged at both 100% and 125% load, and didn't even reach 70 degrees.


In principle, you could reduce the current-sense shunt resistance and comfortably use it at a load of 120–130 W, but there are a few caveats.
1. It was tested open on a bench. In real-world conditions, especially outdoors in summer, things could be considerably worse.
2. This model has the highest efficiency in the entire series, so the others will run hotter.
3. Keep in mind the information in the description, which specifies the temperature ranges and their limits depending on output voltage and other parameters.


Output ripple.
1–4. No load, then at currents of 1.05, 2.1, and 2.7 A, corresponding to 0, 50, 100, and 130% power.
5, 6. 100 Hz ripple at 50% and 100% load.

With a specified limit of 200 mV, the power supply's 75–80 mV is comfortably within spec.


Conclusion.
There isn't much to say here other than that the power supply meets the specifications listed in the description. I can say that I've been using power supplies like these for several years, and they work great.
I use them to power PoE switches. A 100 W unit is enough for an eight-port switch, while a 65 W power supply will do for a four-port model. That's assuming all ports are in use and the switch is installed outdoors, where the equipment enclosure can heat up in the sun. Cameras typically draw around 5–6 W, so the power supplies have plenty of headroom. In a cooler location, you can easily use a 65 W unit for an eight-port switch and a 35 W unit for a four-port switch.

In short, I definitely recommend it.

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