Sign in with
Sign up | Sign in
Don't Be Surprised When Your Cheap PSU Blows Up
By , Patrick Afschar,
1. 1670 Watts For $60?

The PC power supply market is one of the most difficult for enthusiasts and the tech press to address. There are products for every requirement, every performance category, and even for most tastes in design, not to mention the huge price span! In our past power supply roundups, we tested products ranging from a little more than $40 to more than $200. With very few exceptions, most of the units we've evaluated have done what they've claimed; sometimes by a long-shot and other times just barely.

Beyond the solutions from well-known vendors, there is a very large number of cheap, no-name products out there, many of which are really easy to find on auction sites at really attractive prices. To get an idea of what lies behind these purported bargains, we bought three of them using Buy It Now. Obviously, we did not give away our real identities to the seller, ensuring we'd receive the same hardware as everyone else.

We ended up with 500 W and 750 W PSUs from a brand called Sutai and an additional 420 W unit without an identifying sticker on it. Prices ranged from around €13.90 to €25.90, which currently equals $19 to $35, plus shipping. In the end, we paid something around $60 for 1670 W worth of power delivery. On the surface, that sounds like a pretty juicy deal.

Without giving away too much about performance, it's safe to say that the value of what you get when you spend this little on a power is pretty darned minimal. The two Sutai models came without any accessories at all, and were simply wrapped in film. The 420 W PSU at least came in a box with a descriptive label with Power Supply written on it.

None of the test subjects come with anything more than a minimal selection of cables and connectors, lacking any kind of information about the internals, technical specifications, or capabilities. Anything beyond wattage, voltage, and amperage ratings seems to be asking too much.

2. Sutai Double Power P4-500W

Right off the bat, the name Double Power promises more than this lightweight PSU (less than 1 kg) can deliver. Inside the 500 W PSU, there is a circuit board that takes up about 70% of the unit's base, with a generous 140 mm fan on top. The unit does not have a power switch.

The length of the cables and number of connectors is meager. Aside from the motherboard and CPU cables, there are just one SATA, three Molex, and one floppy connector. Using a graphics card with an auxiliary power connection requires a Molex adapter. The length of the heavily bent and uncovered cable is around 30 cm, or 12 inches.

Already during warm-up, the Sutai Double Power P4-500W gives away a serious design flaw: the airflow from the fan does not blow over the PSU components and PCB, but simply past the side of it. The air consequently flows (without doing much cooling) straight from the fan to the cooling vents on the back panel. Our airflow temperature test supports that impression, as the escaping air has nearly the same temperature as the air in the test lab.

Although this weakness alone concerned us, we decided to put the Sutai Double Power P4-500W through our tests anyway. The very first test, reviewing efficiency at 100% load, did not go very well. Even at just 300 W the PSU repeatedly shut itself down.

We managed to gather some maximum load values for the different voltage rails. The 3.3 V rail, for example, has a claimed maximum load of up to 32 A, but anything above 10 A causes this PSU to shut down, despite other rails hardly being used at all. We could not even begin to test the two 12 V rails. Rated at 26 and 14 A respectively, the PSU shut itself down long before we even got close to those numbers. At combined loads as low as 270 W, the voltages on the 3.3 V and 12 V rails were already falling below values required by the ATX specification.

Nevertheless, we conducted our tests at 50% of the maximum rated load (250 W), in order to generate some test scores. Efficiency was around 69% at this load level, and it decreased rapidly at higher loads. The efficiency reached its maximium value of 81% at a load of only 20% (100 W), suggesting that this is probably not a 500 W unit, but rather a relabeled 250 or 300 W unit. Ripple and noise measurements resulted in 3.3 V values that were already well out of the ATX specification at a 50% load, and most likely much worse so at higher loads. By the end of our testing, we could clearly smell the overloaded PSU components, and the unit was noticeably hot in some spots.

The Sutai Double Power P4-500W PSU does not have anything at all in common with a modern, reliable, and efficient power supply. The build quality is abysmal, the packaging questionable, and the results are, to put it gently, a good reason to stay away from this brand altogether. Let's take a look at its second unit.

3. Sutai Double Power P4-750W

With its 750 W ceiling, the Sutai Double Power P4-750W is basically a big brother to the P4-500W. While it has one more SATA connector (two total), it has one less Molex connector. Aside from that, the two power supplies are almost like peas in a pod.

The Sutai Double Power P4-750W did not even last as long as the 500 W model. Alarmed by the weaknesses of the 500 W unit, we began our tests at very low loads. Again, the results suggest a PSU originally designed for a maximum load of 300 W. Looking at the test results at low loads, this becomes rather evident.

It reaches an efficiency of 82% at loads as low as 85 W, a value that more expensive and powerful models would be happy with at this low level. But expecting the $25 Sutai Double Power P4-750W to continue delivering efficiency values like that might be a tad optimistic, even at 300 and 400 W, which is output that should be easy for a 750 W PSU. We were not extremely surprised when the PSU suddenly died of a heat stroke at a load of just 350 W, making further testing impossible.

Compared to the 500 W model, the Sutai Double Power P4-750W is an even bigger shame. While its little brother shut down at low loads, at least it'd fire back up after resting a bit. The 750 W unit overheated after just a short period of testing, and was completely dead after that.

4. LPK-225

When it came time to test the PSU LPK-225, the comparatively low power rating of 420 W, the presence of a cardboard box, and an actual bundled AC power cable gave us hope again. Not for long, mind you. But first things first.

We could already tell that this 420 W PSU is not among the most modern products we have tested when we unpacked it. Though small 80 mm fans would still be acceptable, tiny ventilation holes made us sense danger. The unit’s sharp edges are evidence of poor build quality, but perhaps that is to be expected from a PSU worth roughly $15.

Assume, then, that you get what you pay for. In our case, that would be a 400 W PSU according to the auction we bought out. But instead we got what looked like a 420 W unit. Typically we wouldn't complain about getting more than we paid for; in this case, though, it was hard to ignore the obvious fact that we got scammed. The 420 W output was only mentioned in two places: on a small sticker on the housing and (this takes serious cojones) written by hand on the cardboard box.

There is not much to say about the test results of this PSU. We again started out cautiously and were actually able to identify its efficiency levels at 25, 50, and 85 W loads, reaching a maximum efficiency value of 79% at 85 W. Further testing was not possible because, during a short test with a 300 W load, the PSU suddenly bode us farewell with several explosions and sparks flying everywhere. Once more, we didn’t even have our camera ready. We simply were unprepared for a unit to totally give up so far below its rated output ceiling.

In short: if you buy this PSU, you should not be dependent on your computer, and perhaps invest in a fire extinguisher as a precaution.

5. Conclusion

To be honest, we didn't expect much from these cheap PSUs. We believe it’s pretty much impossible to manufacture a complex piece of hardware like a 500+ W power supply at the cost of a rib eye steak. As a general rule, you should definitely be wary when a deal sounds too good to be true, undercutting common price points by such a significant level.

Typically, hope dies last, which was also the case in this review. At least we thought we would be able to run the cheapo power supplies reliably for a couple of hours, even if not at maximum loads. With two of the three units dying quickly and none of them coming close to reaching their specified maximum loads, you have got to start wondering.

The retailers’ marketing of these products is more fraud than funny, and understandingly we must caution everyone about purchasing such products. If you are looking for a really cheap PSU, at least we recommend to go for a branded unit. Quality products start at $40 and it’s easily possible to spend five times more money, which is only necessary for the hardcore users out there.

Everyone else: make sure you're purchasing a quality PSU so that you can rest easy knowing your other components (and ultimately, your data) are safe. This test demonstrates one thing: if you buy too cheap, you will probably end up buying twice.