TMP-30124C Power Supply from Traco Power

Today I’m reviewing a rather unusual power supply. Though, to be more precise, the power supply itself isn’t so much unusual as relatively rare “around here,” partly because of its price. One of my regular readers sent it to add to my collection, and I’d like to thank him for that.


In general, power supplies from this company are positioned as offering good value for money. But as we know, everything is relative, and that’s especially true of the idea of an “optimal price,” since it means something different to everyone. Some will say “that’s astronomical,” while others will say, “That’s practically free—gotta get one.”

This power supply belongs to the TMP series and has overload and overtemperature protection. Versions with one, two, and three outputs are available, and it comes with a stated 3-year warranty.


It comes in a neat little white box, with the model number visible through the window.


This is the first power supply I’ve received that came with an instruction manual.


The variety of models listed in the datasheet is truly impressive: different power ratings, voltages, and numbers of outputs. There are also models with wire terminals and versions with pins for mounting on a circuit board (the former have a C suffix).
This review covers the 30124C model: 30 W, a single 24 V output, and wire terminals.


All the basic information is printed on the label. Interestingly, the datasheet specifies an input frequency range of 47–440 Hz, while the label says 50/60 Hz.


By the way, the 440 Hz upper limit mentioned above applies only to the 4–30 W models; the more powerful models are rated for up to 60 Hz.
I was also a little surprised to see the recommended rating for an external fuse specified in the datasheet.


There’s a two-pin terminal block on the input, with no provision for grounding. The output has a five-pin terminal block, shared across all versions of this power supply.


Both terminal blocks use a lift-style clamp: the wire sits in a square opening, and tightening the screw raises the clamp to press down on it. In my opinion, this is one of the most reliable designs, since it doesn’t tend to push the wire out of the terminal block.


I definitely wouldn’t call this power supply compact, especially given its relatively low capacity.
There’s only one screw visible on the bottom. Also, the bottom isn’t flat; it bulges slightly around that screw.


Here are the terminal dimensions and pin assignments. Since this review covers a single-output power supply, only pins 3 and 5 are used.


One important advantage of branded equipment is that you can often buy various useful add-ons for it, such as an adapter for mounting the power supply on a DIN rail. The downside is that these accessories are often quite expensive.


I probably don’t need to say how curious I was about the insides. Unfortunately, once you remove the cover, all you see is black potting compound. To be fair, I knew that before the power supply arrived: the person who sent it to me had taken a photo with the cover off. But seeing it in a photo is one thing; getting to poke around yourself is quite another.


Inside, there are two boards along the sides with terminal blocks. Essentially, this is a module designed for mounting on a circuit board, adapted for standalone use—something similar could be seen in my review of one of Mean Well’s power supplies.


To get better access to the components, I desoldered the boards, of course. The large copper areas and plated-through holes made them extremely reluctant to come off.


Only after desoldering the boards did I notice that the one with the output terminal has an LED :)
You can see the LED through a small hole in the case, but I hadn’t noticed it at first. You can also see that the board has plenty of footprints for SMD capacitors, but they’re unpopulated.


The inside is filled with relatively soft potting compound. In principle, I could have dug it out, as I once did with a compact power supply, even though that one had hard potting compound. But if you look closely, you can see all the main components—well, except for the PWM controller, perhaps. There’s not much point in redrawing the circuit either, given how uncommon power supplies like this are among hobbyists and how difficult they are to repair because of that same potting compound.


Anyway, let’s go in order.
First, the input stage. You can clearly see the thermistor, input capacitor, X capacitor, common-mode choke and bridge rectifier. Judging by its size, the capacitor is about 68 µF. Later, I looked through Nichicon’s catalogs for similar capacitors and found that at 16 × 31.5 mm and 400 V, its capacitance is unlikely to be more than 47 µF. It depends on the capacitor series, but the series marking isn’t visible here.


The high-voltage transistor’s heatsink is visible at the top, and the output diode assembly can be seen to the left. Since the output current is only 1.25 A, it doesn’t need a heatsink.


The output has five 220 µF, 50 V capacitors and an inductor to reduce ripple. They’re KZH series capacitors, a newer series designed for use in switching power supplies. The transformer is separated by a plastic insert. When I desoldered the board, the insert stuck to it.

Overall, the components are neatly laid out, with no unnecessary frills, and the quality looks good. I didn’t find a varistor, but if you could call that a “matter of taste,” I couldn’t find a fuse either... At this point, I remembered the line in the datasheet about an external fuse.
Sure, the fuse could be hidden somewhere inside, but no matter how carefully I looked, I couldn’t find it. Strange...

Even without taking it apart, you can tell this is a standard single-switch flyback power supply.


Datasheets for branded equipment can be surprising. They include not only the usual information, specifications and dimensions, but also additional details that can be just as important—for example, the maximum capacitive load. For the power supply under review, it’s 1500 µF.


That’s it for the inspection. Let’s move on to the tests, starting with a simple power-on and output voltage measurement.
The power supply starts up fairly quickly. The output voltage is slightly low, but I don’t see that as a serious issue.


Next came the load tests. I’ll say right away that the power supply holds its voltage very well, even under more than 100% overload. The reading on the right is at 2.7 A, while the rated maximum current is just 1.25 A!


I had to run the automatic test several times: first with a 2 A limit, then 3 A, then 3.5 A. It only hit the protection limit when I set it to 4 A. It turned out I’d just missed it on the previous run: the protection trips at 3.6 A, almost three times the rated current.
On the one hand, that’s bad; on the other, we’ll check that later.


Measuring output ripple.
The specified maximum for this power supply is 1% of the output voltage, or 240 mV peak-to-peak. The oscilloscope was connected directly to the output terminals.


First, the normal operating conditions: no load, then 10, 20 and 30 W. Even at maximum rated power, the peak-to-peak ripple was under 80 mV, including the “spikes”—well below the specified 240 mV.


Now for conditions that aren’t quite normal: 40, 50, 70 and 85 W. The maximum here is almost a threefold overload.
Naturally, the ripple increased. These oscilloscope traces are shown at 50 mV per division, and even at 85 W the peak-to-peak ripple was 300 mV—and that was only due to brief spikes. The main ripple component was noticeably lower.


Here’s the waveform at a 5 ms/div timebase, i.e. at a lower frequency. The measurements are at no load, then 10, 20, 30, 45 and 85 W. The first four traces are at 20 mV/div, the fifth at 50 mV/div, and the sixth required 100 mV/div.
You could say that’s a lot, but ripple is only specified for the first four conditions, up to 30 W. Beyond that, nothing is promised, but the result is still perfectly reasonable.


Let’s get back to the load tests and try measuring efficiency. I’ll say right away that this power supply draws quite a lot of power with no load—about 2 W. With the widespread shift to energy-saving technologies, controllers with a “green” operating mode, and special chips replacing bleeder resistors, this result looks rather unusual.
Modern power supplies of comparable power usually consume less than 1 W in this mode, and more often around 0.3–0.5 W.


Efficiency was measured at load currents from 0.25 to 3.5 A in 0.25 A increments, giving us 14 data points: 0.25, 0.5...3.25, 3.5 A.
Efficiency fell below the stated 80% only at currents of 0.25 and 0.5 A; in all other modes it was around 84–85%.


The entire test was completed in one go, and the power supply handled it without any problems. The test lasted more than 6 minutes in total, most of which was spent operating under overload.


And finally, the temperature test.
By chance, I heated it differently from my usual procedure of gradually increasing the power. Instead, I ran it at 15 W, 30 W, 45 W, and then 45 W again. For some reason, the load didn’t switch to 60 W the last time, and I noticed too late.
Each stage lasted 20 minutes, so the whole test took 1 hour 20 minutes, half of which was spent at one-and-a-half times the rated load.

First, 20 minutes each at 15 and 30 W. Since most of the heat was on top, mainly due to the potting compound, I took the thermal images from above only.


Two 20-minute stages at 45 W. The case around the high-voltage transistor’s heatsink and the output diode assembly had already reached about 92–95 degrees, but the power supply kept running.


As I wrote above, there’s usually nothing good about a power supply being able to withstand heavy overloads, but there are exceptions.
This is a good thing when the power supply has thermal protection and it’s implemented properly.
For the next stage of the temperature test, with the power supply already thoroughly warmed up, I raised the load to 60 W. After 1 minute 15 seconds, it shut down into protection mode.
And it did so correctly: it abruptly switched off the output voltage, and I could turn it back on only after it had cooled down a little and its input capacitor had discharged. So the overtemperature protection is excellent too.
I tested the short-circuit protection by accident, and as you can imagine, there were no problems there either.


Finally, a little about output voltage stability and how it depends on temperature.
1. Voltage under a 150% load with the power supply warmed up.
2. No load, with the power supply still warm.
3. After cooling down. There is a slight temperature dependence: the voltage drops a little as it warms up, but a difference of 0.02 at maximum operating conditions is excellent, in my opinion.


So, what can I say?
It’s an excellent power supply. At the very least, it meets its stated specifications, and at best, it exceeds them. For example, it can withstand not just a 105% overload, but almost three times its rated load. Ripple is also well within the stated range.

There are probably two things I didn’t like: the lack of a fuse and the high power consumption with no load. Though I still have some doubts about the missing fuse—I really did try to find it. I may dig into the potting compound.


That’s all from me. I hope this was useful, and here’s a small wish: may you only come across high-quality power supplies. And of course, a huge thank-you to Dmitry for providing this power supply for testing.