Contents
Sun Pro ATX-550W
Some of you may be familiar with a video from Corsair where they tested some cheap PSUs, and in each case, the switching transistors exploded under 335W load. Those power supplies were actually made by Sun Pro, like this one. Therefore, I have every reason to believe that we will finally see a Sydney Harbour new year’s eve fireworks display at around 320 Watts (which will be test 4). I’ll be double checking to make sure I have the camera ready for this one. In the way of connectors, we have 4 Molex connectors, two floppy connectors, a 20+4 pin ATX Connector, and a 4 pin ATX12V connector. All wires are 20AWG, which is thinner than the minimum recommended 18AWG.
_
Come on guys, it’s not 2001. We need more on the 12V rail.
Load Testing
Test 1 (117.67W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 12V | 4.84A | 12.09V | 30.0mV |
| 5V | 5.06A | 5.06V | 11.9mV |
| 3.3V | 10.09A | 3.33V | 15.0mV |
| −12V | 0A | −11.96V | 12.5mV |
| 5Vsb | 0A | 5.25V | 21.9mV |
| AC Power | 145.3W | ||
| Efficiency | 80.99% | ||
| Power Factor | 0.62 | ||
Test 2 (207.76W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 12V | 9.7A | 12.13V | 51.3mV |
| 5V | 10.02A | 5.01V | 15.6mV |
| 3.3V | 10.03A | 3.31V | 18.1mV |
| −12V | 0.1A | −12.2V | 25.6mV |
| 5Vsb | 1.04A | 5.2V | 16.3mV |
| AC Power | 254.1W | ||
| Efficiency | 81.76% | ||
| Power Factor | 0.56 | ||
Test 3 (262.32W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 12V | 12.21A | 12.21V | 68.1mV |
| 5V | 14.85A | 4.95V | 15.0mV |
| 3.3V | 10.0A | 3.30V | 29.4mV |
| −12V | 0.1A | −12.57V | 27.5mV |
| 5Vsb | 1.04A | 5.2V | 16.3mV |
| AC Power | 325.7W | ||
| Efficiency | 80.54% | ||
| Power Factor | 0.59 | ||
Test 4 (325.25W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 12V | 14.76A | 12.3V | 68.8mV |
| 5V | 19.44A | 4.86V | 26.3mV |
| 3.3V | 9.91A | 3.27V | 41.2mV |
| −12V | 0.11A | −13.13V | 38.8mV |
| 5Vsb | 1.04A | 5.18V | 17.5mV |
| AC Power | 411.79W | ||
| Efficiency | 76.55% | ||
| Power Factor | 0.61 | ||
Again, for a cheaper unit, the performance wasn’t too bad below 300W. Like the last 2 units, it did manage to keep the voltages and ripple in check, with reasonable efficiency. It was above 80% most of the time, but only just. Bear in mind, though, that a similar unit used in the US with a lower input voltage wouldn’t achieve this efficiency. However, things changed once I asked for more than 300W. The efficiency took a nose dive and started out at around 76% and gradually dropped as the unit warmed up. It held up long enough to read the voltages and ripple, but my prediction was correct. A couple of minutes after reading the parameters, it exploded – at only 60% of it’s labelled 550W.
| Rail | Test 3 (262.32W) | Test 4 (325.25W) |
| 12V |  |
 |
| 5V |  |
 |
| 3.3V |  |
 |
| −12V |  |
 |
| 5Vsb |  |
 |
A Look Inside
_
Starting at the input filtering, this unit actually has even more components than required. It has two coils, two X-capacitors, three ceramic capacitors and three MOVs. It’s not too bad for a cheap unit, but the ceramic capacitors are not safety rated, so a power spike could cause them to fail short-circuit. The two input capacitors are 470μF parts from the infamous JEE. In the way of switching transistors, two Sino Microelectronics D304Xs are used (one of which is split in half now), which are rated for 12A at 25°C. The four primary rectifying diodes are only rated at 2A, so (again de-rating to allow for the spiky current waveform) this unit can only pull 368W from a 230v input, which equates to about 294W output assuming 80% efficiency. For a 550W unit, this is insufficient. A bridge rectifier rated for at least 4A should have been used.
On the secondary side, the silicon is also undersized. The 12V rail uses a MOSPEC F16C20C fast recovery rectifier rated at 16A. You would probably get away with pulling 17A from it, but a 20A part should have been used. The 5V rail uses a MOSPEC S30D40C rated at 30A. I can’t see that doing 38A without cooking. The 3.3V rail uses a Diodes Inc. SBL2040CT schottky rectifier rated at 20A. Grrrrrr. Use a 30A part, guys. The capacitors are a mix of JEE and GoldLink. The former is known for extremely low quality parts, and the latter isn’t a whole lot better.
_
Te Bao is the name of Sun Pro’s factory, so like most of their power supplies, this one uses a Te Bao Fan. Unfortunately, I often find Te Bao fans with locked up bearings. The fact that this unit lacks a temperature controller doesn’t help either.
Actual specifications
| Real Wattage | 260W |
| OEM | Sun Pro |
| PFC | None |
| Price | $30 (AUD) |
| ATX Connector type | 20+4 pin |
| Worst-case voltage regulation (12v, 5v, 3.3v) | 2.5%, 2.8%, 0.9% |
| Worst-case ripple (12v, 5v, 3.3v) | 68.8mV, 26.3mV, 41.2mV |
| Worst-case efficiency | 76.55% |
| Input filtering | Adequate |
| CPU Connector | ATX12v (4 pin) |
| PCIe Connectors | None |
| Molex (Peripheral) Connectors | 4 |
| FDD Power connectors | 2 |
| SATA Power connectors | None |
Conclusions
Pros: Good voltage regulation, good ripple suppression, good input filtering
Cons: Explodes at only 320W, not enough connectors, low quality capacitors and fan, noisy
Bottom Line: It does OK below 300W, but it’s loud and lacks SATA connectors.
Score: 4/10
Loading ...