The 2013 El-Cheapo Power Supply Round-up

Zumax ZU400W

Zumax is a budget division of Topower – a company which I haven’t seen much of over the last few years. The last Topower unit I got my hands on was the Global Win SAF450 – an old but reasonably well performing unit. Let’s see if this power supply is at least as good as it was.

zu400-label_zu400

Again, we have a label which incorrectly claims two 12V rails. This power supply really only has one. Like most other power supplies around the $20 mark, it is plain grey in colour.

Load Testing

Test 1 (116.05W Load)

Rail Load Voltage Ripple
12V 4.6A 12.29V 41.8mV
5V 5.03A 5.03V 19.2mV
3.3V 10.18A 3.36V 27.0mV
−12V 0A −12.32V 26.2mV
5Vsb 0A 5.05V 32.2mV
AC Power 151.1W
Efficiency 76.80%
Power Factor 0.57

.

Test 2 (199.09W Load)

Rail Load Voltage Ripple
12V 9.2A 12.19V 68.2mV
5V 9.78A 4.89V 31.6mV
3.3V 10.0A 3.30V 28.4mV
−12V 0.11A −12.66V 61.8mV
5Vsb 0.98A 4.89V 42.4mV
AC Power 265.0W
Efficiency 75.13%
Power Factor 0.55

.

Test 3 (251.2W Load)

Rail Load Voltage Ripple
12V 13.6A 12.02V 86.0mV
5V 9.88A 4.94V 38.8mV
3.3V 9.97A 3.29V 28.8mV
−12V 0.11A −12.95V 78.0mV
5Vsb 0.97A 4.86V 47.4mV
AC Power 341.1
Efficiency 73.64%
Power Factor 0.56

.

Test 4 (298.95W Load)

Rail Load Voltage Ripple
12V 17.8A 11.86V 108.8mV
5V 9.94A 4.97V 47.4mV
3.3V 9.94A 3.28V 29.8mV
−12V 0.11A −13.43V 104.2mV
5Vsb 0.97A 4.83V 50.4mV
AC Power 420.2W
Efficiency 71.22%
Power Factor 0.54

.

The 12V rail started at 12.29 and finished at 11.86V, which gives us 0.29V (2.42%) Regulation and a 0.43V (3.58%) drop. On the 5V, we saw between 5.03V and 4.89V, which equates to 0.11V or 2.2% regulation and 0.14V (2.8%) variation. The 3.3V rail was at 3.36V during Test 1 and 3.28V during Test 4, giving us 0.06V (1.82%) regulation and a 0.08V (2.42%) drop. This is not a great result on any of the rails.

The efficiency was woeful. 76.8% was as good as it got, and during Test 4, it was down to just 71.22%. 300W was also as much as I was able to squeeze from this power supply. When I asked for 350W, it blew both switching transistors.

Rail Test 3 (251.2W) Test 4 (298.95W)
12V  zu400-test3-12v  zu400-test4-12v
5V  zu400-test3-5v  zu400-test4-5v
3.3V  zu400-test3-3.3v  zu400-test4-3.3v
−12V  zu400-test3--12v  zu400-test4--12v
5Vsb  zu400-test3-5vsb  zu400-test4-5vsb

.

The ripple suppression was only borderline passable. By Test 4, it was very close to the maximum allowable limit in the 12V and 5V rail, and was slightly above on the 5vsb rail (not that your hardware will notice the 0.4mV difference). If the ripple suppression is borderline now, it won’t take long before the capacitors wear enough to let the ripple get out of spec.

Disassembly

zu400-primary_zu400-secondary

The Zumax’s nput filtering consists of nothing but three Y-Capacitors (including the one after the rectifier), which is nowhere near enough components. The other key components are a 4A bridge rectifier, 470µF Sun branded primary capacitors, and rather ‘well done’ 8A MJE13007 switching transistors – which are completely incapable of sustaining a 400W Load.

The 12V rail uses an F20C20CT rectifier – a Fast Recovery Rectifier rated for 20A. It would be better to use a Schottky rectifier instead, as they offer better efficiency. I couldn’t identify it’s manufacturer, but in all likelihood, its specifications are identical to those of the MOSPEC F20C20C. The other two rails use a MBR2045CT rectifiers, which are also rated at 20A, although they are the more efficient Schottky rectifiers. Again, I couldn’t identify their real manufacturer, but the specs are, in all likelihood, the same as the ON Semiconductors MBR2045CT. The capacitors on the secondary side are all from Asia’X, which are actually made by the dreaded Fuhjyyu – some of the most unreliable capacitors on the planet.

zu400-fans_zu400-internals

The fans are both branded Zumax. The 92mm fan is actually made by EverCool, and, while I can’t confirm this, I suspect that the 80mm fan is too. They are temperature controlled, but they were not quiet. I could hear them during test 1, and they were getting very loud during Test 2. It wasn’t quite as loud as the Aozhong, but still loud enough to get annoying. The 80mm fan did not have sufficient lubricant in the bearing – just a small amount of thick grease. The 92mm fan didn’t have a removable plug on the bearing, so I couldn’t check its lubricant. I wouldn’t be surprised if it’s just as insufficient, however. The heat sinks have a decent amount of surface to air contact, but are too thin to effectively conduct heat away from the silicon components.

Specifications and Conclusions

Real Wattage 300W
OEM Topower
PFC None
Price $20 (USD)
ATX Connector type 20+4 pin
Worst-case voltage regulation (12v, 5v, 3.3v) 2.4%, 2.2%, 1.8%
Worst-case ripple (12v, 5v, 3.3v) 108.8mV, 47.4mV, 29.8mV
Worst-case efficiency 71.22%
Input filtering Inadequate
CPU Connector ATX12V (4 pin)
PCIe Connectors None
Molex (Peripheral) Connectors 4
FDD Power connectors 1
SATA Power connectors 2

.

Pros: None

Cons: Can’t deliver labelled rating (−2), Low quality capacitors (−3), Mediocre ripple suppression (−1), (Mediocre Voltage Regulation (−1), Low quality Fans (−1), Very inefficient (−1), Loud (−1)

Score: 0/10

HWIFail

Pages: 1 2 3 4 5 6 7 8 9 10 11 12