SUPPLIED BY: InWin
PRODUCT: GreenMe 650W
PROD LINK: InWin’s Current Offerings
PRICE: $74.99 @ NewEgg
Price is at the time of testing!
|GreenMe 650W – Overshoot Transient Tests|
|VSB On||VSB to 100%, 12V||Off to 100%, 12V|
So… what do we see in the above scope shots? How about three very well controlled power on spikes? Yes indeed, Inwin did their homework here. These look very good to me… not the best ever, but just about good enough to be in the textbook about doing power on transient spike suppression right.
Time for that old hot box treatment now. This is going to be a mite tricky, this time, as this unit is being loaded by both load testers and only the top one gets its heat funneled into the modified ATX case I use for a hot box.
I’m just going to turn off all the fans except for the one in the duct that helps pull the hot air into the box. Once test five rolls around, I’ll kick on the big 120V exhaust fan on the top SunMoon – that will dump all its accumulated heat into the box at once, and hopefully drive the temperature up to the forty degrees mark that this unit is rated for.
|GreenMe 650W – Hot Load Tests|
|Test #||+3.3V||+5V||+12V1||+12V2||+12V3||+12V4||DC W/
|Progressive Load Tests|
Well lookie here… we got all the way up to thirty-eight degrees, and the unit had no problem with that. It seems reasonable to expect this unit to do the whole forty degrees, though the exhaust temperature was starting to ramp up by then.
Fortunately, the heat did not have too much effect on the efficiency… we’re still doing Bronze. Not that Bronze is a hard target to hit in 2012, mind you.
Now, before we move on, I want to look at the PFC numbers. I didn’t address this in the cold tests, because it’s not really that important. And it’s not that important here, either, because you only need a 0.90 figure at 50% load to pass Bronze. But… the box promised us we’d see 0.99. On this unit, only one single test level actually meets this, and only if you round up. That’s test five. While I’m going to give Inwin a pass on this, because it did technically hit the magic number, I want to take this time to caution other companies about making promises like this on the box. If I’d seen test five hit 0.984, I’d be scoring against it. Maybe it would be best to promise 0.9 instead of 0.99. It’s a much easier target to hit, and it still satisfies the 80 Plus people. Indeed, just dropping that number nine off the end gets this unit to pass muster on every single test. Not just on test five. Much better, no? Just something to think about for the marketing department.
Moving on, let’s do some calculigrating on the voltage readings again. Going from left to right in the above chart, I see 2.1%, 2.0%, 1.9%, 1.1%, 1.4%, and 1.4%. Wow, the heat really had some effect on these numbers, didn’t it? Again, the 12V1 rail suffered from some issues due to the ATX connector, as did the minor rails, presumably due to the effects of heat inside the power supply housing. Averaged, I get a number of 1.7%. Unfortunately, this is the number I score on, and this number is enough of a downslide to drop this unit out of the “excellent” slot and into the upper end of the “very good” slot. That means a half-point difference for the performance scoring.
Speaking of the scoring, we need to take a look at the most important aspect of performance yet before we can come up with a score there… the ripple control.
Oh, nuts. Things were going so well with this unit, too. Now, because this unit’s design only uses a single 12V output for all four rails, 12V1 is the best indicator of what we should see on all four 12V rails. But, this unit isn’t able to do that on all four, for some reason. The further we go up in numbers, the worse the ripple gets. By the time we’re at 12V4, ripple is a good 25mV above what we see on 12V1. I must admit, I don’t know when I last saw such a large difference like this. Back when I got the second SunMoon, I went out of my way to match up the values on the ATX specification mandated capacitors between the units. I broke out the DMM and measured them all, then picked the closest ones in value.
And on most units, the two SunMoons are in perfect agreement with each other on these scope shots as a result. But not this time. Indeed, even flipping between each 12V load bank on each SunMoon gives me a different result on this unit. Just look at 12V1 and 12V2 up there… see the extra little jaggies getting into 12V2 that weren’t in 12V1? Odd. Very odd indeed. There must be something going on with this unit that’s causing this because nothing else in the whole house is doing it. Thus, I have to call it as I see it. And what I see is 12V4 hitting about 75mV at times. In my book, excellent would be 25mV or less, very good would be around 50mV. This one, I’d call strictly average.
But that’s just the 12V side of things… what’s going on with the minor rails? Oh, about 30mV on the 3.3V rail at worst, and 45mV on the 5V rail at worst. Though both are above the 25mV mark I require for excellent status, the 3.3V rail is close enough I’ll allow a pass there in scoring. But the 5V rail isn’t getting off so easily. It’s just about at spec in test CL2.