First Look
It’s unusual to find a PSU that is actually branded Channel Well, since they usually just make PSUs for other companies. As far as quality is concerned, this unit should be similar to the PSUs they make for other companies. The fan grille on this unit is similar the ones often seen on ThermalTake ToughPower PSUs, which are also made by CWT. The problem with these is that they have thick metal and fewer holes than the more commonly used wire grilles, which means they don’t let as much air through.
It is worth noting that, according to the label, the unit has two 12V rails. We will find out later whether this is really the case, but since this isn’t a low-end $20 unit, it will be treated as a dual rail unit during the tests.
Load Testing
Test 1 (116W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 12V1 | 2.43A | 12.15V | 44.4mV |
| 12V2 | 2.43A | 12.13V | 39.4mV |
| 5V | 5.04A | 5.04V | 58.1mV |
| 3.3V | 9.79A | 3.23V | 20.0mV |
| -12V | 0A | -11.91V | 28.1mV |
| 5Vsb | 0A | 5.08V | 16.3mV |
| AC Power | 145.4W | ||
| Efficiency | 79.78% | ||
| Power Factor | 0.73 | ||
Test 2 (207W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 12V1 | 2.43A | 12.16V | 46.9mV |
| 12V2 | 2.43A | 12.14V | 43.1mV |
| 5V | 5.04A | 5.04V | 52.5mV |
| 3.3V | 9.79A | 3.23V | 17.5mv |
| -12V | 0.1A | -12.00V | 20.0mV |
| 5Vsb | 1A | 5.05V | 15.6mV |
| AC Power | 252W | ||
| Efficiency | 82.14% | ||
| Power Factor | 0.73 | ||
Test 3 (262W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 12V1 | 4.86A | 12.22V | 60.0mV |
| 12V2 | 7.26A | 12.18V | 50.6mV |
| 5V | 15.12A | 5.03V | 48.1mV |
| 3.3V | 9.79A | 3.23V | 25.6mV |
| -12V | 0.1A | -12.21V | 38.8mV |
| 5Vsb | 1.01A | 5.04V | 16.9mV |
| AC Power | 321W | ||
| Efficiency | 81.62% | ||
| Power Factor | 0.73 | ||
Test 4 (319W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 12V1 | 7.36A | 12.27V | 58.8mV |
| 12V2 | 7.35A | 12.25V | 57.5mV |
| 5V | 20.08A | 5.02V | 53.1mV |
| 3.3V | 9.79A | 3.23V | 22.5mV |
| -12V | 0.1A | -12.4V | 43.8mV |
| 5Vsb | 1.01A | 5.03V | 21.9mV |
| AC Power | 393.5W | ||
| Efficiency | 81.07% | ||
| Power Factor | 0.72 | ||
Test 5 (351W Load – Cold)
| Rail | Load | Voltage | Ripple |
| 121V | 7.37A | 12.28V | 45.6mV |
| 12V2 | 7.36A | 12.26V | 43.8mV |
| 5V | 10.08 | 5.02V | 56.3mV |
| 3.3V | 19.58 | 3.23V | 23.8mV |
| -12V | 0.1A | -12.41V | 62.5mV |
| 5Vsb | 1A | 5.01V | 21.9mV |
| AC Power | 434.3W | ||
| Efficiency | 80.82% | ||
| Power Factor | 0.71 | ||
Test 6 (402W Load – Hot)
| Rail | Load | Voltage | Ripple |
| 12V1 | 9.73A | 12.16V | 50.6mV |
| 12V2 | 9.72A | 12.15V | 51.3mV |
| 5V | 19.88A | 4.97V | 51.9mV |
| 3.3V | 19.21A | 3.17V | 34.4mV |
| -12V | 0.11A | -12.6V | 53.1mV |
| 5Vsb | 0.99A | 4.95V | 28.7mV |
| AC Power | 518.4W | ||
| Efficiency | 77.55% | ||
| Power Factor | 0.71 | ||
Overload Test (540W Load – Hot)
| Rail | Load | Voltage | Ripple |
| 12V1 | 9.73A | 12.14V | 63.8mV |
| 12V2 | 9.72A | 12.10V | 70.0mV |
| 5V | 19.88A | 4.93V | 81.3mV |
| 3.3V | 19.21A | 3.14V | 33.7mV |
| -12V | 0.11A | -12.99V | 85.0mV |
| 5Vsb | 0.99A | 4.91V | 28.7mV |
| AC Power | 729.5W | ||
| Efficiency | 74.02% | ||
| Power Factor | 0.70 | ||
The voltage regulation in these tests was excellent until we hit 400W. While still in spec, the voltages were starting to drop, suggesting that the unit was reaching its limits. The fact that the voltage on the molex connectors was slightly different to the voltage on the CPU connector suggests that this unit really does have two 12V rails. The efficiency was also very good. 74.02% in the overload test was as low as it got, but even that is still pretty good considering CWT doesn’t make any big claims about the efficiency. The power factor was better than units with no PFC, but the passive PFC used in this unit is still no match for the active PFC used in higher end units. The ripple was a different story. On the 5v rail, it was above the maximum 50mV in all but test 3, although it stayed within spec on all of the other rails. The capacitors on the 5V rail are probably starting to go bad (as we will find out later). When we attempted to overload the unit to 540W, the ripple on the 5V rail increased significantly and was way too high and the voltage on the 3.3V rail dropped to near the minimum allowed in ATX Specifications. After running at 540W for about 1 minute and 30 seconds, it got too hot and the over temperature protection stepped in and shut it down. It powered back on again after allowing it to cool, so it hadn’t failed.
| Rail | Test 6 (402W) | Overload Test (540W) |
| 12V1 |  |
 |
| 12V2 |  |
 |
| 5V |  |
 |
| 3.3V |  |
 |
| -12V |  |
 |
| 5Vsb |  |
 |
A look inside
_
The overall build quality of this unit is very good, which is what we generally expect from Channel Well. Starting at the input filter, the unit has two coils, two X-caps (one of which is attached to the AC receptacle), and four Y-Caps (two of which are attached to the AC receptacle). Not only acceptable, but even better than the minimum recommended input filtering. The primary switchers are a pair of STP14NK50ZFP N-Channel Power MOSFETS. They are rated for 14A at 25°C and 7.6A at 100°C. It was clearly enough to allow this unit to deliver the 400W promised by the label. The one problem with the primary side of this unit is in the PFC circuit. Keen eyed observers may notice the discolouration around the resistors near where the PFC inductor goes on the primary side. This indicates a badly designed circuit where more current is being drawn through the resistors than they can handle, causing them to get extremely hot. It’s also a bad idea to put capacitors there, since the hot resistors will cook them. Thankfully, though, the circuit they are part of only boosts the voltage to the switching transistors, which improves their efficiency but the power supply will still run without it. The two primary capacitors are 820uF, which is a good size for a 400 Watter. Too bad they are made by Fuhjyyu.
Turning the PCB around to the secondary side and de-soldering the silicon there reveals that this unit is built for a modern system which draws from the 12V rail. It also reveals that the unit really does have two 12V rails (the wires are in separate groups and there are two OCP lines). The 5v and 3.3v rails each have an STPS3045CW Schottky rectifier rated at 30A and the two 12V rails share a pair of STPS4045CW Schottly rectifiers rated at 40A each. That gives us a total of 80A capacity between the two 12V rails, which is excellent. However, like the primary side, the secondary also has a major let down. Every single filter capacitor is made by Fuhjyyu. When this unit was removed from the PC, the two 1000uF capacitors on the 5Vsb rail were bulging and had to be replaced before the unit would run for the tests. This only confirms my theory that the capacitors on the 5V rail could also be going bad and allowing the ripple to go out of spec.
_
Uh, oh, we have a death ADDA. Better call in the snake handlers.
Specifications
| Real Wattage | 400W |
| OEM | Channel Well |
| PFC | Passive |
| Price | Unknown |
| ATX Connector type | 20 + 4 pin |
| CPU Connector | EPS12V(8 Pin) and ATX12v (4 pin) |
| PCI-E Connectors | None |
| Molex (Peripheral) Connectors | 5 |
| FDD Power connectors | 1 |
| SATA Power connectors | 2 |
Conclusions
Pros: Can deliver it’s labelled rating, Quiet, PFC, Efficient, Big heat sinks and capacitors, Input filtering, 80A available between the two 12V rails, Over temperature protection
Cons: Low quality Fuhjyyu capacitors, Badly designed voltage boost circuit gets hot
Bottom Line: It’s certainly no total piece of junk, but watch the capacitors.
Score: 7/10
Discuss this review in our forums here.
Review sample source: Removed from PC
Loading ...