Contents
Infinity ATX P4-460W
I have previously reviewed a power supply which was branded Infinity. However, I don’t believe that this power supply is from the same company. I have seen this exact model for sale on eBay for around the $30 mark, so it’s unlikely to be a high quality unit.
_
5V heavy again. Seriously, it’s alarming just how many units are still being sold and produced that are better suited to a Pentium III computer than anything newer.
Load Testing
Test 1 (116.16W Load)
| Rail | Load | Voltage | Ripple |
| 12V | 4.7A | 12.16V | 42.0mV |
| 5V | 5.04A | 5.04V | 22.2mV |
| 3.3V | 10.09A | 3.33V | 16.8mV |
| −12V | 0A | −12.17V | 34.2mV |
| 5Vsb | 0A | 5.05V | 37.4mV |
| AC Power | 151.7W | ||
| Efficiency | 76.57% | ||
| Power Factor | 0.6 | ||
.
Test 2 (201.66W Load)
| Rail | Load | Voltage | Ripple |
| 12V | 9.8A | 12.19V | 62.4mV |
| 5V | 9.94A | 4.97V | 33.4mV |
| 3.3V | 9.97A | 3.29V | 17.8mV |
| −12V | 0.1A | −12.37V | 70.6mV |
| 5Vsb | 0.99A | 4.97V | 39.2mV |
| AC Power | 261.8W | ||
| Efficiency | 77.03% | ||
| Power Factor | 0.58 | ||
.
Test 3 (246.65W Load)
| Rail | Load | Voltage | Ripple |
| 12V | 13.8A | 11.96V | 77.4mV |
| 5V | 9.94A | 4.97V | 39.0mV |
| 3.3V | 9.88A | 3.26V | 19.0mV |
| −12V | 0.11A | −12.6V | 88.8mV |
| 5Vsb | 0.99A | 4.94V | 44.2mV |
| AC Power | 332.3W | ||
| Efficiency | 74.23% | ||
| Power Factor | 0.58 | ||
.
The 12V rail had maximum and minimum values of 12.19 and 11.96V respectively, which equates to 0.19V (1.58%) Regulation, and a variation 0f 0.23V, or 1.92%. The 5V was at 5.04V during test 1 and 4.97V during the other 2 tests, which gives us 0.8% regulation, and a 1.4% drop. The 3.3V rail started at 3.33V and dropped to 3.26V, which works out to be 1.21% regulation, and a 2.12% drop. Overall, this is actually a pretty good result.
As usual with these old designs, the efficiency was bad, and never managed 80%. I attempted to load this power supply to about 300W for a fourth test, but the 12V rail voltage dropped to near the minimum allowed in ATX spec, the efficiency dropped back to around 68%, and the power supply failed after a few minutes. The 12V rectifier is what failed, so there was no explosion, unfortunately.
| Rail | Test 2 (201.66W) | Test 3 (246.65W) |
| 12V |  |
 |
| 5V |  |
 |
| 3.3V |  |
 |
| −12V |  |
 |
| 5Vsb |  |
 |
.
The ripple did stay in spec, but it was a little on the high side. I like to see it stay below half the maximum allowed at all times, which didn’t happen. The 5vsb rail’s ripple was getting close to the maximum allowed during Test 3.
Disassembly
_
The input filtering consists of just two Y capacitors, which is completely inadequate. There is PCB space for more parts, but the manufacturer decided to cut costs by not installing them. There is no bridge rectifier present – just four 2A diodes. The primary capacitors are rated at 470uF and are made by Zhila. However, they do look very small physically for 470µF parts, so I do wonder if they could really be 330µF parts wth 470µF sleeves, which is surprisingly common in cheap power supplies. The switching transistors are Huashan KSH13009s rated at 12A.
The 12V rail uses an MUR1620CT ultrafast recovery rectifier rated at 16A. No wonder it failed when I tried to load the 12V rail to 19A for Test 4. The other two rails use 20A MOSPEC S20C40C schottky rectifiers. These rectifiers are theoretically insufficient, as both of these rails are claimed to be capable of more than 20A. The capacitors on the secondary side are mostly from Zhila, except for the one on 12V rail which is from Nichicon – a high quality Japanese brand. Why they would bundle one decent part here with other low quality parts elsewhere is beyond me, especially when it’s the 5Vsb rail that usually kills its caps first.
_
And here’s Mary’s little Wam again. The fan was beginning to fail, and making a loud grinding sound for the first few minutes after being powered on. By the time the power supply failed, however, the noise was gone and it was pumping out a lot of air, so it’s unlikely to have played a major role in the failure. Needless to say, the lubricant in the bearings was insufficient. The heatsinks have reasonable surface to air contact, but they are quite thin.
Specifications and Conclusions
| Real Wattage | 250W |
| OEM | Unknown |
| PFC | None |
| Price | $30 |
| ATX Connector type | 20+4 pin |
| Worst-case voltage regulation (12v, 5v, 3.3v) | 1.6%, 1.9%, 0.8% |
| Worst-case ripple (12v, 5v, 3.3v) | 77.4mV, 39.0mV, 19.0mV |
| Worst-case efficiency | 74.23% |
| Input filtering | Inadequate |
| CPU Connector | ATX (4 pin) |
| PCIe Connectors | None |
| Molex (Peripheral) Connectors | 2 |
| FDD Power connectors | 1 |
| SATA Power connectors | 2 |
.
Pros: Good Voltage Regulation
Cons: Can’t deliver labelled rating (−2), Low quality capacitors (−2), Low quality Fan (−2), Hardly any input filtering (−1), Inefficient (−1) Mediocre ripple suppression (−0.5), Old 5V-Heavy design (−0.5)
Score: 1/10
Loading ...